From vikas venera.co.in Tue Mar 1 11:51:17 2005 From: vikas venera.co.in (Vikas Singhal) Date: Tue Mar 1 15:10:11 2005 Subject: [Mp4-tech] MP4 encoder/decoder for live video conference vs.streaming References: Message-ID: <005401c51e26$e14456f0$680ba8c0@MAYA> Hi Loren, Usually the VO, VOS, VOL information is transmited by out-of-band means (RTSP, SDP etc.) but this information may also be the part of the stream. If you send it as part of the stream then it will be inefficient if you send this for all VOPs. If you send it only once then also also there is a problem if the initial packets are not received by receiver (due to late joining or packet losses). The ideal solution in that scenario will be to send this information at a particular frequency (lets say once every minute or so... You may decide the actual value). This strategy will provide a random access point into the stream. Hope this helps Vikas ----- Original Message ----- From: "Loren Rogers" To: Sent: Friday, February 25, 2024 10:11 PM Subject: [Mp4-tech] MP4 encoder/decoder for live video conference vs.streaming > Hello, > > How does one get around the problems of transporting MPEG4-2 video streams > for live video conferencing vs. streaming. The reason I'm asking is because > there is information that is left out of the VOP and VOP header, like frame > size (QCIF, CIF, etc.). I'm not sure what's normally done for live video > conferencing. Whether the VOL, or VO, or both pieces of information is sent > just once in the beginning of the call, or for every single VOP. Can > someone please advise. > > Thanks. > > /Loren > > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php > From guogr bjnec.nec.com.cn Tue Mar 1 14:16:39 2005 From: guogr bjnec.nec.com.cn (guogr@bjnec.nec.com.cn) Date: Tue Mar 1 15:13:05 2005 Subject: [Mp4-tech] H.263 to 3Gp Message-ID: Hello all, There is a problem with 3GP file. I have added the fileheader to the frames of H.263 with the ISO base media file format,but the file can't be played. the attached file is the *.3gp file. (See attached file: bitstream.3gp) how can I convert the bitstream of H.263 to a 3GP file? I wanna know the way. I would appreciate very much If someone could help me. Thanks! Cheery Guo NEC IC Design Beijing Co., Ltd. Device Solution Development Project Group Quantum Plaza, Floor 7, No.27, Zhichunlu Haidian District, Beijing 100083, China TEL: +86-10-82351155 ex.6211 FAX: +86-10-82357690 E-mail: guogr@bjnec.nec.com.cn -------------- next part -------------- A non-text attachment was scrubbed... Name: bitstream.3gp Type: application/octet-stream Size: 11746 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20050301/1d5285a6/bitstream-0001.obj From Thammi_Reddy Satyam.com Tue Mar 1 11:55:59 2005 From: Thammi_Reddy Satyam.com (Thammi_Reddy) Date: Tue Mar 1 15:14:35 2005 Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Message-ID: <3D9FDCA910DD4445896A0BE5ECF009D4F5C0DB@bla.satyam.com> Dear Gary, Thanks you very much for the reply. I have following doubt from earlier question/answer 2. The following text is from slice header syntax table. num_ref_idx_active_override_flag if( num_ref_idx_active_override_flag ) num_ref_idx_l0_active_minus1 The syntax element num_ref_idx_active_override_flag can be high or low(I think this is true for all the 3 profiles). In case this flag=0, do I still receive the syntax element num_ref_idx_l0_active_minus1? Or this received and dropped? In case FLAG=0 and num_ref_idx_l0_active_minus1 not present, can I drive a highZ on thse lines? Similarly depending on the value of memory_management_control_operation the following elements are present or not present in the encoded bit stream. Please explain. difference_of_pic_nums_minus1 long_term_pic_num long_term_frame_idx max_long_term_frame_idx_plus1 This is a new question: The bit stuffing for seq parameter set , picture parameter set syntax is done by the encoder or decoder for making byte alignment.? Regards, Thammi Reddy Regards, Thammi Reddy Question2 :In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? Answer2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. _____ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Gary Sullivan Sent: Saturday, February 26, 2024 1:37 AM To: Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts Thammi et al, Question 1: The number of bits used by those two syntax elements is specified by other syntax elements. If you read the semantics of those syntax elements, you will see a specification of how to determine their length. For example, the length of frame_num is log2_max_frame_num_minus4+4 bits. Question 2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. Question 3: The syntax category is used for data partitioning. See the definition of "category" and the introduction section of subclause 7.2 and the syntax and semantics sections for slice data partition A, slice data partition B, and slice data partition C. Data partitioning is not used in the Baseline profile. Question 4: If a decoder receives a value of some syntax element that is prohibited by the standard, I believe a decoder can do whatever it wants to do, for as long as it wants to do it. As far as the standard is concerned, I believe it can halt and catch on fire (although I would not personally recommend it). I think that also applies to value of syntax elements that are not prohibited by the standard but are not supported by the decoder conformance requirements of the profile that the decoder claims to implement. Best Regards, Gary Sullivan _____ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Thammi_Reddy Sent: Thursday, February 24, 2024 11:01 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Dear experts, I have following doubts in H.264 decoder baseline profile. Please answer these questions. Thanks in advance. 1. Ref: chapter 7.3 of the spec. frame_num, pic_order_cnt_lsb syntax elements are marked with descriptor are given as u(v). So the number of bits taken by these syntax elements vary from current slice header and next slice header depending on the magnitude of the actual values sent by the encoder. If this is true how do I know the LSB if data transfer is MSB to LSB? 2. In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? 3. What is the importance of CATEGORY of the syntax elements provided in chapter 7.3(syntax in tabular form)? Please explain. 4. If the values are fixed as per the annexure A to If incorrect values are received for baseline profile (e.g. receive entropy_coding_mode_flag =1, slice_type != (0 OR 2 OR 5 OR 7 )... etc) what processing I have to do? From which stage I have to resume in this case? Regards, Thammi Reddy ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050301/3fdca20b/attachment-0001.html From rohan teneoris.com Tue Mar 1 07:25:06 2005 From: rohan teneoris.com (Rohan Kangralkar) Date: Tue Mar 1 15:15:35 2005 Subject: [Mp4-tech] how to implement buffer intercept method Message-ID: <1109659337.1675.1425.camel@rohan.teneoris.com> hi i have to test a decoder for MPEG-4 video compliance. i have a MoMusys reference decoder. how do i implement the buffer intercept method. how do i do a static and dynamic test. -- Rohan Kangralkar Teneoris Networks PVT LTD From garysull windows.microsoft.com Tue Mar 1 00:47:11 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Tue Mar 1 15:16:43 2005 Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D566D51@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> num_ref_idx_l0_active_minus1 is not received in the slice header when num_ref_idx_active_override_flag is equal to 1. That's what the "if" statement means. It might be useful for you to read the syntax table description at the beginning of clause 7. I don't know what a highZ is. Yes, the presence or absence of the syntax elements that you listed depends on the value of memory_management_control_operation. I'm not sure what further explanation you seek -- perhaps you should read the semantics (subclause 7.4) and decoding processes (clause 8) sections relating to those syntax elements. I'm not sure what byte-alignment bit stuffing you're referring to. Perhaps it is rbsp_trailing_bits()? If so, then making sure that is in the bitstream is the encoder's responsibility. It is the encoder's responsibility for all syntax to match what is specified in clause 7. Best Regards, Gary Sullivan ________________________________ From: Thammi_Reddy [mailto:Thammi_Reddy@Satyam.com] Sent: Monday, February 28, 2024 10:26 PM To: Gary Sullivan; Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts Dear Gary, Thanks you very much for the reply. I have following doubt from earlier question/answer 2. The following text is from slice header syntax table. num_ref_idx_active_override_flag if( num_ref_idx_active_override_flag ) num_ref_idx_l0_active_minus1 The syntax element num_ref_idx_active_override_flag can be high or low(I think this is true for all the 3 profiles). In case this flag=0, do I still receive the syntax element num_ref_idx_l0_active_minus1? Or this received and dropped? In case FLAG=0 and num_ref_idx_l0_active_minus1 not present, can I drive a highZ on thse lines? Similarly depending on the value of memory_management_control_operation the following elements are present or not present in the encoded bit stream. Please explain. difference_of_pic_nums_minus1 long_term_pic_num long_term_frame_idx max_long_term_frame_idx_plus1 This is a new question: The bit stuffing for seq parameter set , picture parameter set syntax is done by the encoder or decoder for making byte alignment.? Regards, Thammi Reddy Regards, Thammi Reddy Question2 :In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? Answer2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Gary Sullivan Sent: Saturday, February 26, 2024 1:37 AM To: Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts Thammi et al, Question 1: The number of bits used by those two syntax elements is specified by other syntax elements. If you read the semantics of those syntax elements, you will see a specification of how to determine their length. For example, the length of frame_num is log2_max_frame_num_minus4+4 bits. Question 2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. Question 3: The syntax category is used for data partitioning. See the definition of "category" and the introduction section of subclause 7.2 and the syntax and semantics sections for slice data partition A, slice data partition B, and slice data partition C. Data partitioning is not used in the Baseline profile. Question 4: If a decoder receives a value of some syntax element that is prohibited by the standard, I believe a decoder can do whatever it wants to do, for as long as it wants to do it. As far as the standard is concerned, I believe it can halt and catch on fire (although I would not personally recommend it). I think that also applies to value of syntax elements that are not prohibited by the standard but are not supported by the decoder conformance requirements of the profile that the decoder claims to implement. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Thammi_Reddy Sent: Thursday, February 24, 2024 11:01 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Dear experts, I have following doubts in H.264 decoder baseline profile. Please answer these questions. Thanks in advance. 1. Ref: chapter 7.3 of the spec. frame_num, pic_order_cnt_lsb syntax elements are marked with descriptor are given as u(v). So the number of bits taken by these syntax elements vary from current slice header and next slice header depending on the magnitude of the actual values sent by the encoder. If this is true how do I know the LSB if data transfer is MSB to LSB? 2. In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? 3. What is the importance of CATEGORY of the syntax elements provided in chapter 7.3(syntax in tabular form)? Please explain. 4. If the values are fixed as per the annexure A to If incorrect values are received for baseline profile (e.g. receive entropy_coding_mode_flag =1, slice_type != (0 OR 2 OR 5 OR 7 )... etc) what processing I have to do? From which stage I have to resume in this case? Regards, Thammi Reddy ************************************************************************ ** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************ ** ************************************************************************ ** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************ ** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050301/3ce3f527/attachment-0001.html From nitthilan fastmail.fm Tue Mar 1 01:27:14 2005 From: nitthilan fastmail.fm (Nitthilan Kannappan Jayakodi) Date: Tue Mar 1 15:18:18 2005 Subject: [Mp4-tech] Reference Lists Message-ID: <1109669234.18051.216263934@webmail.messagingengine.com> Dear experts, In the h.264 standard, two reference lists(0 and 1) are used. Why is this so? because, 1) the reference pictures are present in both the lists 2) the forward , backward and bi-prediction can be done from a single list further if u say that the two reference lists the ordering is such that the neighbouring pictures( the immediate fw. and backward picture) may have a less reference index ( mostly zero) and thus produce compression, why to repeat the pictures in the two lists. We can have List(1) for all backward pictures and List(0) for all fw. pictures. Is there any special reason for this kind of arrangement? Pls. clarify. expecting ur reply, K.J.Nitthilan -- Nitthilan Kannappan Jayakodi nitthilan@fastmail.fm From avinash_s citilindia.com Tue Mar 1 17:12:29 2005 From: avinash_s citilindia.com (Avinash Sawant) Date: Tue Mar 1 15:19:26 2005 Subject: [Mp4-tech] VLD look up table Message-ID: <001b01c51e53$bf734a90$1201a8c0@citpl.com> Skipped content of type multipart/alternative-------------- next part -------------- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.300 / Virus Database: 266.5.1 - Release Date: 2/27/2005 From Thammi_Reddy Satyam.com Tue Mar 1 18:01:59 2005 From: Thammi_Reddy Satyam.com (Thammi_Reddy) Date: Tue Mar 1 15:20:19 2005 Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Message-ID: <3D9FDCA910DD4445896A0BE5ECF009D4F5C0E2@bla.satyam.com> Gary, thanks for the info. It is clear to me now. Thammi Reddy _____ From: Gary Sullivan [mailto:garysull@windows.microsoft.com] Sent: Tuesday, March 01, 2024 2:17 PM To: Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts num_ref_idx_l0_active_minus1 is not received in the slice header when num_ref_idx_active_override_flag is equal to 1. That's what the "if" statement means. It might be useful for you to read the syntax table description at the beginning of clause 7. I don't know what a highZ is. Yes, the presence or absence of the syntax elements that you listed depends on the value of memory_management_control_operation. I'm not sure what further explanation you seek -- perhaps you should read the semantics (subclause 7.4) and decoding processes (clause 8) sections relating to those syntax elements. I'm not sure what byte-alignment bit stuffing you're referring to. Perhaps it is rbsp_trailing_bits()? If so, then making sure that is in the bitstream is the encoder's responsibility. It is the encoder's responsibility for all syntax to match what is specified in clause 7. Best Regards, Gary Sullivan _____ From: Thammi_Reddy [mailto:Thammi_Reddy@Satyam.com] Sent: Monday, February 28, 2024 10:26 PM To: Gary Sullivan; Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts Dear Gary, Thanks you very much for the reply. I have following doubt from earlier question/answer 2. The following text is from slice header syntax table. num_ref_idx_active_override_flag if( num_ref_idx_active_override_flag ) num_ref_idx_l0_active_minus1 The syntax element num_ref_idx_active_override_flag can be high or low(I think this is true for all the 3 profiles). In case this flag=0, do I still receive the syntax element num_ref_idx_l0_active_minus1? Or this received and dropped? In case FLAG=0 and num_ref_idx_l0_active_minus1 not present, can I drive a highZ on thse lines? Similarly depending on the value of memory_management_control_operation the following elements are present or not present in the encoded bit stream. Please explain. difference_of_pic_nums_minus1 long_term_pic_num long_term_frame_idx max_long_term_frame_idx_plus1 This is a new question: The bit stuffing for seq parameter set , picture parameter set syntax is done by the encoder or decoder for making byte alignment.? Regards, Thammi Reddy Regards, Thammi Reddy Question2 :In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? Answer2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. _____ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Gary Sullivan Sent: Saturday, February 26, 2024 1:37 AM To: Thammi_Reddy; mp4-tech@lists.mpegif.org Subject: RE: [Mp4-tech] H.264/decoder/Baseline profile/doubts Thammi et al, Question 1: The number of bits used by those two syntax elements is specified by other syntax elements. If you read the semantics of those syntax elements, you will see a specification of how to determine their length. For example, the length of frame_num is log2_max_frame_num_minus4+4 bits. Question 2: When a profile specifies that a syntax element shall have a particular value, that does not mean that the syntax element is not sent. It just means that when it is sent it must have that specfied value. It appears to me that pic_init_qs_minus26 can have any value (within the specified range) within a Baseline profile bitstream. However, its value will not be used in the decoding process in that case because there will not be any SP or SI slices in a Baseline profile bitstream, and those are the only slice types in which the decoding process would depend on the value of that syntax element. Question 3: The syntax category is used for data partitioning. See the definition of "category" and the introduction section of subclause 7.2 and the syntax and semantics sections for slice data partition A, slice data partition B, and slice data partition C. Data partitioning is not used in the Baseline profile. Question 4: If a decoder receives a value of some syntax element that is prohibited by the standard, I believe a decoder can do whatever it wants to do, for as long as it wants to do it. As far as the standard is concerned, I believe it can halt and catch on fire (although I would not personally recommend it). I think that also applies to value of syntax elements that are not prohibited by the standard but are not supported by the decoder conformance requirements of the profile that the decoder claims to implement. Best Regards, Gary Sullivan _____ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Thammi_Reddy Sent: Thursday, February 24, 2024 11:01 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] H.264/decoder/Baseline profile/doubts Dear experts, I have following doubts in H.264 decoder baseline profile. Please answer these questions. Thanks in advance. 1. Ref: chapter 7.3 of the spec. frame_num, pic_order_cnt_lsb syntax elements are marked with descriptor are given as u(v). So the number of bits taken by these syntax elements vary from current slice header and next slice header depending on the magnitude of the actual values sent by the encoder. If this is true how do I know the LSB if data transfer is MSB to LSB? 2. In sequence parameter set RBSP frame_mbs_only_flag=0 for baseline profile as only frames and no fields are allowed. Do I still receive this flag or encoder will not send it at all? On the similar lines do I receuve pic_init_qs_minus26 even for baseline profile? In general some syntax elements are not received OR all are received but I have to drop? 3. What is the importance of CATEGORY of the syntax elements provided in chapter 7.3(syntax in tabular form)? Please explain. 4. If the values are fixed as per the annexure A to If incorrect values are received for baseline profile (e.g. receive entropy_coding_mode_flag =1, slice_type != (0 OR 2 OR 5 OR 7 )... etc) what processing I have to do? From which stage I have to resume in this case? Regards, Thammi Reddy ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050301/86b23abf/attachment-0001.html From SassanP vbrick.com Tue Mar 1 09:57:28 2005 From: SassanP vbrick.com (Sassan Pejhan) Date: Wed Mar 2 06:32:23 2005 Subject: [Mp4-tech] MP4 encoder/decoder for live video conference vs.streaming Message-ID: Another approach that is sometimes used for multicast sessions - where users may join in midstream - is to transmit the VOS/VO/VOL headers at regular intervals as part of the stream itself. This is useful in cases where SAP is not used. Sassan > -----Original Message----- > From: Herbert Thoma [mailto:tma@iis.fhg.de] > Sent: Monday, February 28, 2024 4:18 AM > To: Loren Rogers > Cc: mp4-tech@lists.mpegif.org > Subject: Re: [Mp4-tech] MP4 encoder/decoder for live video conference > vs.streaming > > > Hi Loren, > > usually the VS, VO and VOL headers are transmitted in the startup > phase of the connection. These headers are typically transmitted > over a reliable channel (e.g. RTSP on TCP in ISMA streaming) > whereas the VOPs are transmitted over a unreliable channel > (eg. RTP on UDP in ISMA streaming). > > Herbert. > > Loren Rogers schrieb: > > Hello, > > > > How does one get around the problems of transporting MPEG4-2 video > > streams for live video conferencing vs. streaming. The reason I?m > > asking is because there is information that is left out of > the VOP and > > VOP header, like frame size (QCIF, CIF, etc?). I?m not sure what?s > > normally done for live video conferencing. Whether the > VOL, or VO, or > > both pieces of information is sent just once in the > beginning of the > > call, or for every single VOP. Can someone please advise. > > > > Thanks. > > > > /Loren > > > > > > _______________________________________________ > > NOTE: Please use clear subject lines for your posts. > Include [audio, > > [video], [systems], [general] or another apppropriate identifier to > > indicate the type of question you have. > > > > Note: Conduct on the mailing list is subject to the > Antitrust guidelines > > found at > > > http://www.mpegif.org/public/documents/vault/mp-out-30042-Anti trust.php > -- Herbert Thoma Group Manager Video Multimedia Realtime Systems Department Fraunhofer IIS Am Wolfsmantel 33, 91058 Erlangen, Germany Phone: +49-9131-776-323 Fax: +49-9131-776-399 email: tma@iis.fhg.de www: http://www.iis.fhg.de/ _______________________________________________ NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php From magnus.hoem popwire.com Tue Mar 1 16:25:19 2005 From: magnus.hoem popwire.com (Magnus Hoem) Date: Wed Mar 2 06:32:28 2005 Subject: [Mp4-tech] H.263 to 3Gp In-Reply-To: References: Message-ID: Hi! There is an error in the file. The size of the mdat-box is 4 bytes to small. The moov-box is now not where it should be. Also, the size of the moov-box seems incorrect. Best regards, Magnus On Mar 1, 2005, at 7:16, guogr@bjnec.nec.com.cn wrote: > > Hello all, > There is a problem with 3GP file. I have added the fileheader to the > frames > of H.263 with the ISO base media file format,but the file can't be > played. > the attached file is the *.3gp file. > (See attached file: bitstream.3gp) > how can I convert the bitstream of H.263 to a 3GP file? I wanna > know the > way. > I would appreciate very much If someone could help me. > Thanks! > > Cheery Guo > > NEC IC Design Beijing Co., Ltd. > Device Solution Development Project Group > Quantum Plaza, Floor 7, No.27, Zhichunlu Haidian District, > Beijing > 100083, China > TEL: +86-10-82351155 ex.6211 FAX: +86-10-82357690 > E-mail: guogr@bjnec.nec.com.cn > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, > [video], [systems], [general] or another apppropriate identifier to > indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust > guidelines found at > http://www.mpegif.org/public/documents/vault/mp-out-30042- > Antitrust.php ********************************************** Popwire Technology Magnus Hoem Senior Research Engineer magnus.hoem@popwire.com ?rsta?ngsv?gen 19 B SE-117 94 Stockholm, Sweden Phone: +46 8 579 116 00 Direct: +46 8 579 112 12 Mobile: +46 733 25 44 34 http://www.popwire.com *********************************************** This message, including any attachments may contain confidential and privileged material; it is intended only for the person to whom it is addressed. Its contents do not constitute a commitment by Popwire except where provided for in a written and undersigned agreement. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/enriched Size: 2033 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20050301/d5e05059/attachment.bin From ravimpeg4video yahoo.co.in Tue Mar 1 16:36:37 2005 From: ravimpeg4video yahoo.co.in (ravi kumar) Date: Wed Mar 2 06:32:32 2005 Subject: [Mp4-tech] optimum fixed point code for pow( 10, x) In-Reply-To: <200503011527.j21FQgXm026757@lists1.magma.ca> Message-ID: <20050301163637.13514.qmail@web8305.mail.in.yahoo.com> Hi all Does any one had optimum fixed point code for pow( 10,x) Regards Ravi --------------------------------- Do you Yahoo!? Read only the mail you want - Yahoo! Mail SpamGuard. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050301/080567cf/attachment.html From singer apple.com Tue Mar 1 17:24:40 2005 From: singer apple.com (Dave Singer) Date: Wed Mar 2 06:32:37 2005 Subject: [Mp4-tech] H.263 to 3Gp In-Reply-To: References: Message-ID: At 2:16 PM +0800 3/1/05, guogr@bjnec.nec.com.cn wrote: >Hello all, >There is a problem with 3GP file. I have added the fileheader to the frames >of H.263 with the ISO base media file format,but the file can't be played. >the attached file is the *.3gp file. >(See attached file: bitstream.3gp) >how can I convert the bitstream of H.263 to a 3GP file? I wanna know the >way. >I would appreciate very much If someone could help me. >Thanks! Hm. My validation software doesn't find the moov atom. You have an ftyp atom at 0, length 14, then an mdat of length 2b24. But if I look at the file at 2b24+14 (hex), I think I am four bytes short of the start of the moov atom. If I fix the length of the mdat to 2b28, I find the moov atom, but my validator then crashes. Are other length fields also wrong? Do you have access to the reference software for the file format? >Cheery Guo > > NEC IC Design Beijing Co., Ltd. > Device Solution Development Project Group > Quantum Plaza, Floor 7, No.27, Zhichunlu Haidian District, Beijing > 100083, China > TEL: +86-10-82351155 ex.6211 FAX: +86-10-82357690 > E-mail: guogr@bjnec.nec.com.cn > >Attachment converted: DaveG49:bitstream.3gp ( / ) (000222B8) >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, >[video], [systems], [general] or another apppropriate identifier to >indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust >guidelines found at >http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -- David Singer Apple Computer/QuickTime From lcheng62 yahoo.com Tue Mar 1 11:44:43 2005 From: lcheng62 yahoo.com (Liang Cheng) Date: Wed Mar 2 06:32:42 2005 Subject: [Mp4-tech] FGS server:A quick question Message-ID: <20050301194443.21636.qmail@web54003.mail.yahoo.com> Hi, all, It seems that the FGS server is implemented very simply. From the ISO MPEG-4 reference code, I found the FGS server. It's main role is just to count the bits. I also read the the MPEG-4 streaming video profile, looks like the syntax of FGS is not simple. It has the motion info, header, marker, etc. Can anyone please tell me if the FGS enhancement layer can be really truncated like what the FGS_server does? One thing for sure is that the truncated fgs enhancement layer stream can be decoded using the standard decoder. Thank you. Liang __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From orho leadtek.com.tw Wed Mar 2 09:47:46 2005 From: orho leadtek.com.tw (=?big5?B?T3JobyCvzqxGpbA=?=) Date: Wed Mar 2 06:32:47 2005 Subject: [Mp4-tech] Where can get the RateControl VM5 & VM8 series documents ? Message-ID: Dear All experts, I am the freshman to touch the rate-comtrol. May you show me where I can download the relational rate-control documents about VM5 & VM8 series? Sincerely, Orho -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050302/8060687d/attachment.html From katayen yahoo.com.tw Wed Mar 2 14:28:20 2005 From: katayen yahoo.com.tw (Heng-Lin Yen) Date: Wed Mar 2 14:57:53 2005 Subject: [Mp4-tech] [H.264]Doubt about the slice_type syntax element Message-ID: <20050302062820.89013.qmail@web17608.mail.tpe.yahoo.com> Dear Experts, I have a doubt about the "slice_type" syntax element when study H.264 recommendation. can anyone explain it for me? Thanks in advance. In Table 7-3, the definition of values of "slice_type" syntax element is specified. But I don't understand why each slice_type has to assign two values simultaneously? For example, P(P slice) is assigned to "0" and "5", and B(B slice) is assigned to "1" and "6" etc. What is its intention? And under this definition, can we arbitrary select one among these two values when encoding the "slice_type" syntax element? Regards, Kata Yen _______________________________________________________________________ Yahoo!集筿獺絚 禣甧秖250MB獺ンぃ┤ http://tw.promo.yahoo.com/mail_new/index.html From dipankar.mitra lgsoftindia.com Wed Mar 2 14:50:59 2005 From: dipankar.mitra lgsoftindia.com (Dipankar Mitra) Date: Wed Mar 2 15:01:33 2005 Subject: [Mp4-tech] H.263 to 3Gp Message-ID: Cheery, Adding to what Dave and Magnus have pointed out.., the last stco offset falls inside the moov box! Seems like the file was not formatted properly. Regards, Dipankar ============================================= 5th Floor, Golf View Homes Wind Tunnel Road, Murugeshpalya, Bangalore -560017, INDIA Ph: +91 - 80- 25275711 extn 161 -----Original Message----- From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org]On Behalf Of Dave Singer Sent: Tuesday, March 01, 2024 10:55 PM To: guogr@bjnec.nec.com.cn; mp4-tech@lists.mpegif.org Subject: Re: [Mp4-tech] H.263 to 3Gp At 2:16 PM +0800 3/1/05, guogr@bjnec.nec.com.cn wrote: >Hello all, >There is a problem with 3GP file. I have added the fileheader to the frames >of H.263 with the ISO base media file format,but the file can't be played. >the attached file is the *.3gp file. >(See attached file: bitstream.3gp) >how can I convert the bitstream of H.263 to a 3GP file? I wanna know the >way. >I would appreciate very much If someone could help me. >Thanks! Hm. My validation software doesn't find the moov atom. You have an ftyp atom at 0, length 14, then an mdat of length 2b24. But if I look at the file at 2b24+14 (hex), I think I am four bytes short of the start of the moov atom. If I fix the length of the mdat to 2b28, I find the moov atom, but my validator then crashes. Are other length fields also wrong? Do you have access to the reference software for the file format? >Cheery Guo > > NEC IC Design Beijing Co., Ltd. > Device Solution Development Project Group > Quantum Plaza, Floor 7, No.27, Zhichunlu Haidian District, Beijing > 100083, China > TEL: +86-10-82351155 ex.6211 FAX: +86-10-82357690 > E-mail: guogr@bjnec.nec.com.cn > >Attachment converted: DaveG49:bitstream.3gp ( / ) (000222B8) >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, >[video], [systems], [general] or another apppropriate identifier to >indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust >guidelines found at >http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -- David Singer Apple Computer/QuickTime _______________________________________________ NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php ********************************************************************************************************************************************************** This email message is for the sole use of the intended recipient(s)and may contain CONFIDENTIAL and PRIVILEGED information. LG Soft India will not be responsible for any viruses or defects or any forwarded attachments emanating either from within LG Soft India or outside. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender By reply email and destroy all copies of the original message. ********************************************************************************************************************************************************** From arunm motechsoftware.com Wed Mar 2 17:47:26 2005 From: arunm motechsoftware.com (Arun Menon) Date: Wed Mar 2 15:03:04 2005 Subject: [Mp4-tech] Disable ac_pred_flag in MPEG-4 simple profile Message-ID: Hi Experts, I want to test my MPEG-4 encoded stream for simple profile with the momusys reference encoder output. I would like to know how to disable AC DC PREDICTION (ac_pred_flag). Can anyone please advice me on this matter. Regards, Arun Menon ------------------------------------------------------------------------------------------------------------------------------- DISCLAIMER: This message (Including any attachments) is confidential and is intended only for the addressee. All MoTech Software mails and attachments are scanned for all known viruses at the time of transmission. However, please scan the mail and attachments for viruses before opening or downloading from this transmission. We will not accept any liability for damages through virus. If you are not the intended recipient or received this message by mistake, please notify the sender by E-mail & delete the message from the system. Any unauthorized use or dissemination of the message in whole or part is strictly prohibited. -------------- next part -------------- A non-text attachment was scrubbed... Name: winmail.dat Type: application/ms-tnef Size: 1636 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20050302/1a724662/winmail-0001.bin From Hee-yong.Kim cinea.com Wed Mar 2 11:10:58 2005 From: Hee-yong.Kim cinea.com (Kim, Hee-yong) Date: Thu Mar 3 20:15:55 2005 Subject: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Message-ID: <4999AF9D8147074283CCE2330986362E2F60E2@iris.dolby.net> Dear JV Reference SW experts, Suppose I would like to code video in IBBP pattern and GOP size=12. Intra frame is forced to be IDR. 1. If I continue to GOP pattern using reference encoder, the coding order of second GOP is: I_12 - B_10 - B_11 - P15 - But since I_12 is IDR, B_10 and B_11 are not bidirectional any more. They can have only backward prection from I_12, even if they are declared as B pictures. Is this a normal practice? 2. JV reference encoder is not flexible enough to do many different kind of picture sequencing. JV encoder Limitation: It always assume that Intra picture is located among equally spaced picture which are presumambly I or P pictures. Because in the configuraion file the Intra period is in terms of the primary picture (I/P) unit. For exmple, there is no way the reference encoder can encode the following sequence: (I_0, P_3,B_1,B_2), (I_4,P_7,B_5,B_6), and so on. 3. What is a normal practice to encode video ? 1 or 2 above or else? I appreciate your answer. Thanks. ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050302/6cc19fde/attachment.html From demirelhakan yahoo.com Wed Mar 2 09:36:43 2005 From: demirelhakan yahoo.com (Hakan Demirel) Date: Thu Mar 3 20:19:01 2005 Subject: [Mp4-tech] [video] ffmpeg h263 decoding problem Message-ID: <20050302173644.57181.qmail@web20926.mail.yahoo.com> Hi, I am trying to convert a .3gp file to a .mp4 file using ffmpeg library. The .3gp file has a H263 encoded video stream and an AMR_NB encoded audio stream.. I want the output .mp4 file to have an MPEG4 encoded video stream and an AAC encoded audio stream.. When I run the following command: ffmpeg -i input.3gp -ar 4000 -ab 24 output.mp4 the resulting video stream is MPEG4 encoded.. But I noticed that many of the frames seem to be dissmissed.. So, the length of the video is smaller.. I tried to increase the frame rate.. It worked but the video stream is flowing really faster according to the original file. I am using ffmeg version 0.4.9pre1 Any idea? :) __________________________________ Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web http://birthday.yahoo.com/netrospective/ From cpark via-telecom.com Wed Mar 2 10:20:36 2005 From: cpark via-telecom.com (Charles Park) Date: Thu Mar 3 20:20:29 2005 Subject: [Mp4-tech] AAC+ test vector Message-ID: <78532FB72FAFBE48A17E2EC19E54B6647BBDA0@exchsd1> Hi all, I can't find any test vector for AAC+ from website. Does anybody have AAC+ encoded test vector? Thanks, Charles(YoungHo) Park From garysull windows.microsoft.com Wed Mar 2 15:47:24 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Thu Mar 3 20:22:13 2005 Subject: [Mp4-tech] Reference Lists Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D5C7B10@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> K.J. Nitthilan et al, See response in-line below: +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of +> Nitthilan Kannappan Jayakodi +> Sent: Tuesday, March 01, 2024 1:27 AM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] Reference Lists +> +> Dear experts, +> In the h.264 standard, two reference lists(0 and 1) +> are used. Why +> is this so? because, +> 1) the reference pictures are present in both the lists They can be present in both lists, but they do not necessarily need to be. That is up to the discretion of an encoder to decide. (Also note that there is little meaningful difference between being at the end of a list without being referred to by reference index and not being in the list at all.) +> 2) the forward , backward and bi-prediction can be done from a single +> list Hypothetically, that might be true in some design. Whether some other design along those lines would be as good as the design we actually have in this standard is different question. More detail about the alternative design would need to be provided and studied. +> +> further if u say that the two reference lists the ordering +> is such that +> the neighbouring pictures( the immediate fw. and backward +> picture) may +> have a less reference index ( mostly zero) and thus produce +> compression, +> why to repeat the pictures in the two lists. The encoder determines what is in the final lists. If it wants to remove some picture(s) from some list(s), it can. Also, if the pictures at the end of a list are never used, it basically makes approximately no functional difference whether they are in the list or not. However, sometimes using multiple references into the same reference picture can provide a coding efficiency benefit. See, for example, the papers and the book published by Markus Flierl, Thomas Wiegand, and/or Bern Girod on the subject of "multi-hypothesis" motion compensation. Additionally, even if the encoder is not using the same picture for both references, there may be other reasons for wanting some picture to be in both lists. For example, if each picture is only allowed to be in one list and there are two pictures that you want to use for bi-prediction, but you have placed both of them in the same list, then you would not be able to do what you want. There is also weighted prediction to consider. And when a list contains more than two reference pictures, there is no loss of coding efficiency involved in making the list longer. And there is no significant increase in complexity that is required to support the idea of letting pictures be in both lists. +> We can have List(1) for all backward pictures and List(0) +> for all fw. +> pictures. That sounds like it would result in just old MPEG-2-style operation. Allowing more than two reference pictures has a coding efficiency advantage. Note that the uses of the words "backward" and "forward" are not really necessarily applicable anymore. Best Regards, Gary Sullivan +> +> +> Is there any special reason for this kind of arrangement? +> Pls. clarify. +> +> expecting ur reply, +> K.J.Nitthilan +> -- +> Nitthilan Kannappan Jayakodi +> nitthilan@fastmail.fm +> +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From andrew astri.org Thu Mar 3 09:18:26 2005 From: andrew astri.org (Wat Yiu Wing Andrew) Date: Thu Mar 3 20:23:52 2005 Subject: [Mp4-tech] telecine/2-3 pulldown in H264 Message-ID: Dear all, I noticed Picture Coding Extension in MPEG2 which tells the decoder to display at what aspect ratio and what field to be repeated. Do H264 have the equivalent Picture Coding Extension as MPEG2? I am particularly interested in the equivilent of the 2-3 Pulldown and telecine achieved by setting these flags REPEAT_FIRST_FIELD$B!"(JTOP_FIELD_FIRST in MPEG-2 streams. Thanks Andrew -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/0f00b026/attachment.html From gbmallikarjunarao yahoo.com Thu Mar 3 03:46:50 2005 From: gbmallikarjunarao yahoo.com (mallikarjun rao) Date: Thu Mar 3 20:24:57 2005 Subject: [Mp4-tech] MPEG4 AAC-LC Decoder Error Calc Message-ID: <20050303034651.83411.qmail@web52005.mail.yahoo.com> Hi, Can any one help me , regarding error calc of MPEG4 AAC-LC Decoder . Calculation of RMS error is only for AL00 test vector or for all test vectors(AL00,AL01,AL02..)? Please send me a reply asap. Thanks Mallikarjun Send instant messages to your online friends http://uk.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/9be2340d/attachment.html From kexu ee.cuhk.edu.hk Thu Mar 3 14:37:24 2005 From: kexu ee.cuhk.edu.hk (Xu Ke) Date: Thu Mar 3 20:26:08 2005 Subject: [Mp4-tech] [H.264]Doubt about the slice_type syntax element Message-ID: <200503030637.j236bKwW023121@cuees8.ee.cuhk.edu.hk> Hi, My understanding is: 1)each slice_type has only one value. 2)when slice_type = 5,other slices of the current coded picture shall have a value of slice_type equals to 5 or 0.That means other slices should have the same slice type (I slice) 3)when slice_type = 0,other slices of the current coded picture can have arbitary slice_type. That means other slices can be P slice instead of I slice only. Correct me if I am wrong. Best regards, XU Ke ======= 2024-03-02 14:28:20 ================= >Dear Experts, > >I have a doubt about the "slice_type" syntax element >when study H.264 recommendation. can anyone explain it >for me? Thanks in advance. > >In Table 7-3, the definition of values of "slice_type" >syntax element is specified. But I don't understand >why each slice_type has to assign two values >simultaneously? For example, P(P slice) is assigned to >"0" and "5", and B(B slice) is assigned to "1" and "6" >etc. What is its intention? > >And under this definition, can we arbitrary select one >among these two values when encoding the "slice_type" >syntax element? > >Regards, >Kata Yen > >_______________________________________________________________________ >Yahoo!奇摩電子信箱 >免費容量250MB，信件在多也不怕 >http://tw.promo.yahoo.com/mail_new/index.html >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php > = = = = = = = = = = = = = = = = = = = = From nittalapavan rediffmail.com Thu Mar 3 08:48:21 2005 From: nittalapavan rediffmail.com (pavan kumar nittala) Date: Thu Mar 3 20:27:31 2005 Subject: [Mp4-tech] [H.264]Doubt about the slice_type syntax element Message-ID: <20050303084949.16985.qmail@webmail18.rediffmail.com> ? Kata Yen ... U can seee in the Reference decoder that while decoding, it is using some thing like if(slice_type> 5) slice_type= slice_type-5; such that the values 7 and 6 will get truncated to 5 and 1 rly ... i donno why they r doing like that ... but i understood like this ... tell me if i have answered to ur query .. pavan On Wed, 02 Mar 2024 Heng-Lin Yen wrote : >Dear Experts, > >I have a doubt about the "slice_type" syntax element >when study H.264 recommendation. can anyone explain it >for me? Thanks in advance. > >In Table 7-3, the definition of values of "slice_type" >syntax element is specified. But I don't understand >why each slice_type has to assign two values >simultaneously? For example, P(P slice) is assigned to >"0" and "5", and B(B slice) is assigned to "1" and "6" >etc. What is its intention? > >And under this definition, can we arbitrary select one >among these two values when encoding the "slice_type" >syntax element? > >Regards, >Kata Yen > >_______________________________________________________________________ >Yahoo!?_???q?l?H?c >?K?O?e?q250MB?A?H???b?h?]???? >http://tw.promo.yahoo.com/mail_new/index.html >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/54243ee5/attachment-0001.html From sathishn teneoris.com Thu Mar 3 10:38:44 2005 From: sathishn teneoris.com (Sathish N) Date: Thu Mar 3 20:29:54 2005 Subject: [Mp4-tech] Direct prediction in 1 MV (16x16) mode In-Reply-To: References: Message-ID: <1109846434.4396.126.camel@sathish> Hello, This question about Direct prediction is with regards to Sec 7.6.9.5 of ISO/IEC 14496-2. Is Direct mode Motion vector decoding (& sample Interpolation) always required to be performed on 8x8 blocks (i.e 4MV mode)? OR is it that it has to be performed on 16x16 macroblock (i.e 1 MV mode) if the co-located MB is coded in 1MV mode & on 8x8 blocks if the co-located MB is coded in 4MV mode ? Would some one please clarify ? Thanks, Sathish From magarwal neomagic.com Thu Mar 3 16:45:25 2005 From: magarwal neomagic.com (Mohit Agarwal) Date: Thu Mar 3 20:32:00 2005 Subject: [Mp4-tech] Enquiry Message-ID: <4226F1CD.2020803@neomagic.com> hi can anybody clear my doubts regarding "vop_rounding_type" as defined in standard (ISO\IEC 14496 -2:2003) 6.3.5 refers to rounding_control which tells about pixel value interpolation in motion compensation.How is this interpolation used in motion compensation as it can be used only in motion estimation. I suppose... From rohan teneoris.com Thu Mar 3 12:11:52 2005 From: rohan teneoris.com (Rohan Kangralkar) Date: Thu Mar 3 20:32:06 2005 Subject: [Mp4-tech] Direct prediction in 1 MV (16x16) mode Message-ID: <1109852668.8190.3.camel@rohan.teneoris.com> Hello, This question about Direct prediction is with regards to Sec 7.6.9.5 of ISO/IEC 14496-2. Is Direct mode Motion vector decoding (& sample Interpolation) always required to be performed on 8x8 blocks (i.e 4MV mode)? OR is it that it has to be performed on 16x16 macroblock (i.e 1 MV mode) if the co-located MB is coded in 1MV mode & on 8x8 blocks if the co-located MB is coded in 4MV mode ? Would someone please clarify ? Thanks, Rohan -- Rohan Kangralkar Teneoris Networks PVT LTD -------------- next part -------------- An embedded message was scrubbed... From: Sathish N Subject: [Fwd: Direct prediction in 1 MV (16x16) mode] Date: 03 Mar 2024 17:40:53 +0530 Size: 2007 Url: /pipermail/mp4-tech/attachments/20050303/de14d476/attachment.mht From Philippe.deneve intec.ugent.be Thu Mar 3 17:14:50 2005 From: Philippe.deneve intec.ugent.be (Philippe De Neve) Date: Thu Mar 3 20:32:10 2005 Subject: [Mp4-tech] H.263 v2/v3 Message-ID: <024f01c5200c$20b9c8b0$bbd6c19d@King> Hi folks, maybe this is not the correct place to post this message but I do not know where to start looking elsewhere. Does anybody have any idea where I can find a h.263 v2 or h.263 v3 (aka h.263 +/++) codec? Or where I can look for it? tnx for all replies, regards, Philippe. From surajt_baba yahoo.co.in Thu Mar 3 16:37:00 2005 From: surajt_baba yahoo.co.in (suraj kumar) Date: Thu Mar 3 20:32:15 2005 Subject: [Mp4-tech] h.264 codec(decoding of 4x4 intraprediction mode) JM9.2(FRExt) Message-ID: <20050303163700.50622.qmail@web8501.mail.in.yahoo.com> hello experts could anyone give a clear explanation on the following variables and why array is been used only for pix_a PixelPos pix_a[4]; PixelPos pix_b, pix_c, pix_d; best regards suraj kumar ________________________________________________________________________ Yahoo! India Matrimony: Find your life partner online Go to: http://yahoo.shaadi.com/india-matrimony From rickyfyg hotmail.com Thu Mar 3 17:07:11 2005 From: rickyfyg hotmail.com (rikcy fei) Date: Thu Mar 3 20:32:19 2005 Subject: [Mp4-tech] CABAC Message-ID: Hello, experts: I don't know how the initialization tables in CABAC are made, i.e how the values of m and n in(m,n) are gotten, can anyone give me a detailed explaination. Thanks so much. paranoia rickyfyg@hotmail.com _________________________________________________________________ 享用世界上最大的电子邮件系统— MSN Hotmail。 http://www.hotmail.com From alexismt comcast.net Thu Mar 3 20:17:00 2005 From: alexismt comcast.net (Alexis Michael Tourapis) Date: Fri Mar 4 16:02:05 2005 Subject: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Message-ID: <030320052017.26920.422770BA0002F7BA0000692822007354469B039C07980A040E@comcast.net> Dear Hee-yong, Regarding Q1, your problem is that you are using IDR frames. IDR frames enforce all references in the buffer to be discarded and therefore resulting in a "closed" GOP structure. However, for most applications, IDR frames might not be necessary. Instead, what you may wish to do is to carefully manage your reference buffer either through MMCO commands or careful reordering/ reference restrictions. Then I coded pictures need not be IDRs and still enable trick modes or whatever other features you may want. To make it more clear, what you could do is to use MMCO commands to discard all prior to I references, only after the "ending" B coded pictures have been encoded. Another method you could do is to restrict the following pictures not to refer to any of the preceeding pictures with clever management in the encoder. Again this would have the same effect. Now regarding your claim in Q2, this is actually not true. I would strongly suggest studying the parameters "PyramidCoding" and "ExplicitPyramidFormat". Setting the first one to 2 would allow you to create rather flexible coding orders or GOPs. Although it may appear that it has some limitations, (i.e. seems to require a periodic formulation) it is possible with simple modifications to even provide the coding structure of the entire sequence with these parameters. Finally, for Q3, I guess there is no real answer here since it all depends on your requirements/application/goal. I.e. you may choose not to use I coded pictures if you wish, have a specific, adaptive, or arbitrary coding order, use an adaptive mix of slice types etc. However i guess you can first try my suggestions in Q1. Best regards, Alexis > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], > [systems], [general] or another apppropriate identifier to indicate the type of > question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An embedded message was scrubbed... From: "Kim, Hee-yong" Subject: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Date: Thu, 3 Mar 2024 19:51:43 +0000 Size: 6045 Url: /pipermail/mp4-tech/attachments/20050303/b613b5cf/attachment.mht From surajt_baba yahoo.co.in Thu Mar 3 20:30:06 2005 From: surajt_baba yahoo.co.in (suraj kumar) Date: Fri Mar 4 16:02:10 2005 Subject: [Mp4-tech] h.264 codec(intraprediction (4x4)luma) jvt 9.2(FRExt) Message-ID: <20050303203006.47721.qmail@web8508.mail.in.yahoo.com> hello experts, i am working on (JVT)JM9.2(RFExt) decoder.I wonder if some one could clarify my doubt mentioned below: what actually variables pos_x and pos_y represent when they are used in intraprdiction(4x4)mode and how can we discriminate them with x and y. typedef struct pix_pos { int available; //availability of neighbouring block int mb_addr; // address of neighbouring macroblock int x; // pixel location wrt x int y; // pixel location wrt y int pos_x; // ? int pos_y; // ? } PixelPos; best regards suraj kumar Yahoo! India Matrimony: Find your life partneronline. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/265f6112/attachment.html From garysull windows.microsoft.com Thu Mar 3 12:34:37 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Fri Mar 4 16:02:16 2005 Subject: [Mp4-tech] telecine/2-3 pulldown in H264 Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D632A1C@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Yes. For example, look in the spec for pic_struct, fixed_frame_rate_flag, and aspect_ratio_idc. The same kind of functionality is supported, but it is in a somewhat different, more flexible, form. And since different system environments may have different ways to handle some of these things, the related syntax is not necessarily always require to be present within the video elementary bitstream under all circumstances in all application environments. Best Regards, -Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Wat Yiu Wing Andrew Sent: Wednesday, March 02, 2024 5:18 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] telecine/2-3 pulldown in H264 Dear all, I noticed Picture Coding Extension in MPEG2 which tells the decoder to display at what aspect ratio and what field to be repeated. Do H264 have the equivalent Picture Coding Extension as MPEG2? I am particularly interested in the equivilent of the 2-3 Pulldown and telecine achieved by setting these flags REPEAT_FIRST_FIELD?TOP_FIELD_FIRST in MPEG-2 streams. Thanks Andrew -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/7250949c/attachment-0001.html From garysull windows.microsoft.com Thu Mar 3 12:42:05 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Fri Mar 4 16:02:21 2005 Subject: [Mp4-tech] [H.264]Doubt about the slice_type syntax element Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D632A3B@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> The only difference between a slice_type value in the range of 0..4 and in the range of 5..9 is the distinction that is expressed in the statement just below the table that you referred to (which is now Table 7-6 in the newer versions): "slice_type values in the range 5..9 specify, in addition to the coding type of the current slice, that all other slices of the current coded picture shall have a value of slice_type equal to the current value of slice_type or equal to the current value of slice_type - 5." For example, if the picture contains some slice that has a slice_type equal to 0, it cannot also contain a slice that has a slice_type in the range of 6 to 9 (and vice versa). Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of pavan kumar nittala Sent: Thursday, March 03, 2024 12:50 AM To: Heng-Lin Yen Cc: mp4-tech@lists.mpegif.org Subject: Re: [Mp4-tech] [H.264]Doubt about the slice_type syntax element Kata Yen ... U can seee in the Reference decoder that while decoding, it is using some thing like if(slice_type> 5) slice_type= slice_type-5; such that the values 7 and 6 will get truncated to 5 and 1 rly ... i donno why they r doing like that ... but i understood like this ... tell me if i have answered to ur query .. pavan On Wed, 02 Mar 2024 Heng-Lin Yen wrote : >Dear Experts, > >I have a doubt about the "slice_type" syntax element >when study H.264 recommendation. can anyone explain it >for me? Thanks in advance. > >In Table 7-3, the definition of values of "slice_type" >syntax element is specified. But I don't understand >why each slice_type has to assign two values >simultaneously? For example, P(P slice) is assigned to >"0" and "5", and B(B slice) is assigned to "1" and "6" >etc. What is its intention? > >And under this definition, can we arbitrary select one >among these two values when encoding the "slice_type" >syntax element? > >Regards, >Kata Yen > >_______________________________________________________________________ >Yahoo!?_???q?l?H?c >?K?O?e?q250MB?A?H???b?h?]???? >http://tw.promo.yahoo.com/mail_new/index.html >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050303/bd9e1e92/attachment.html From changhsu yahoo.com Thu Mar 3 15:11:34 2005 From: changhsu yahoo.com (changhsu@yahoo.com) Date: Fri Mar 4 16:02:28 2005 Subject: [Mp4-tech] [h.264]Why is there a constraint on max number of bins...? Message-ID: <20050303231135.27862.qmail@web53802.mail.yahoo.com> Will anyone please explain the underlying reason for this constrain? [quote from the standard doc] When entropy_coding_mode_flag is equal to 1, the number of bins resulting from decoding the contents of all VCL NAL units of a coded picture shall not exceed ( 32 ÷ 3 ) * NumBytesInVclNALunits + 96 * PicSizeInMbs. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From alexismt comcast.net Thu Mar 3 23:29:50 2005 From: alexismt comcast.net (Alexis Michael Tourapis) Date: Fri Mar 4 16:02:33 2005 Subject: [Mp4-tech] h.264 codec(decoding of 4x4 intraprediction mode) JM9.2(FRExt) Message-ID: <030320052329.11486.42279DED000D96C100002CDE22007621949B039C07980A040E@comcast.net> Dear Suraj, Karsten answered the same question a few days back. pix_a corresponds to pixels on the left, while pix_b, pix_c, pix_d correspond to different pixel positions above your current one. In MBAFF mode it is possible for a macroblock to be in either field or frame mode. This creates an "issue" at macroblock edges for intra prediction since it is possible for the adjacent blocks to be in a different coding mode than the current one (i.e. current is frame while adjacent is in field and vice versa). The basic idea behind this is that pixels need to be predicted horizontally from exactly the same "parity" adjacent pixels and not from opposite parity ones. Although this is not really an issue for prediction pixels coming from MBs above your current position (only one block line needs to be read which corresponds to the same field parity), you would need to properly select the positions of the prediction pixels coming from the left depending on the coding modes of the current and left neighboring MBs. For example if current is frame and left is in field, pix_a[0] and pix[2] should correspond to even parity fields and therefore should be taken from the top MB of the adjacent MB pair. pix_a[1] and pix[3] however correspond to odd parity and therefore should be taken from the bottom MB instead. Somehow similar rules exist if current is in field mode and adjacent is frame. I think you should probably check section 6.4.9.2 for more information. Best regards, Alexis > hello experts > > could anyone give a clear explanation on the following > variables and why array is been used only for pix_a > PixelPos pix_a[4]; > PixelPos pix_b, pix_c, pix_d; > > best regards > suraj kumar > > > ________________________________________________________________________ > Yahoo! India Matrimony: Find your life partner online > Go to: http://yahoo.shaadi.com/india-matrimony > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], > [systems], [general] or another apppropriate identifier to indicate the type of > question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found > at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php From rickyfyg hotmail.com Fri Mar 4 04:34:25 2005 From: rickyfyg hotmail.com (rikcy fei) Date: Fri Mar 4 16:02:37 2005 Subject: [Mp4-tech] CABAC in H.264/AVC Message-ID: Hello, experts: > I don't know how the initialization tables in CABAC are made, i.e how > the values of m and n in(m,n) are gotten, can anyone give me a detailed > explaination. > Thanks so much. > paranoia rickyfyg@hotmail.com > _________________________________________________________________ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn From rickyfyg hotmail.com Fri Mar 4 04:38:25 2005 From: rickyfyg hotmail.com (rikcy fei) Date: Fri Mar 4 16:02:42 2005 Subject: [Mp4-tech] [H.264/AVC] CABAC Message-ID: Hello, experts: I don't know how the initialization tables in CABAC are made, i.e how the values of m and n in(m,n) are gotten, can anyone give me a detailed explanation. Thanks so much. paranoia rickyfyg@hotmail.com _________________________________________________________________ 免费下载 MSN Explorer: http://explorer.msn.com/lccn/ From Hee-yong.Kim cinea.com Fri Mar 4 11:00:18 2005 From: Hee-yong.Kim cinea.com (Kim, Hee-yong) Date: Sat Mar 5 00:17:19 2005 Subject: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Message-ID: <4999AF9D8147074283CCE2330986362E2F60E6@iris.dolby.net> Dear Alexis, Thank you for your kind explanation. But I have not seen any MMCO option in the JV reference encoder. And my understanding is that JV encoder is using only sliding window buffer memory management. Please correcr me if I am wrong. Thanks. Hee-Yong hee-yong.kim@cinea.com -----Original Message----- From: Alexis Michael Tourapis [mailto:alexismt@comcast.net] Sent: Thu 3/3/2024 3:17 PM To: Kim, Hee-yong; mp4-tech@lists.mpegif.org Cc: Mercier, Guillaume; Kim, Hee-yong; Levine, Scott Subject: Re: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Dear Hee-yong, Regarding Q1, your problem is that you are using IDR frames. IDR frames enforce all references in the buffer to be discarded and therefore resulting in a "closed" GOP structure. However, for most applications, IDR frames might not be necessary. Instead, what you may wish to do is to carefully manage your reference buffer either through MMCO commands or careful reordering/ reference restrictions. Then I coded pictures need not be IDRs and still enable trick modes or whatever other features you may want. To make it more clear, what you could do is to use MMCO commands to discard all prior to I references, only after the "ending" B coded pictures have been encoded. Another method you could do is to restrict the following pictures not to refer to any of the preceeding pictures with clever management in the encoder. Again this would have the same effect. Now regarding your claim in Q2, this is actually not true. I would strongly suggest studying the parameters "PyramidCoding" and "ExplicitPyramidFormat". Setting the first one to 2 would allow you to create rather flexible coding orders or GOPs. Although it may appear that it has some limitations, (i.e. seems to require a periodic formulation) it is possible with simple modifications to even provide the coding structure of the entire sequence with these parameters. Finally, for Q3, I guess there is no real answer here since it all depends on your requirements/application/goal. I.e. you may choose not to use I coded pictures if you wish, have a specific, adaptive, or arbitrary coding order, use an adaptive mix of slice types etc. However i guess you can first try my suggestions in Q1. Best regards, Alexis > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], > [systems], [general] or another apppropriate identifier to indicate the type of > question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited. From changhsu yahoo.com Fri Mar 4 09:59:43 2005 From: changhsu yahoo.com (changhsu@yahoo.com) Date: Sat Mar 5 00:20:59 2005 Subject: [Mp4-tech] [H.264]about the constraint of max number of bins for a coded picture Message-ID: <20050304175943.77680.qmail@web53803.mail.yahoo.com> Hello all, Will anyone explain the underlying reason for this constraint? It looks like the JM does not use the same equation as it is in the standard document. Is there a mismatch between JM and the standard? [quote from h.264 document] 7.4.2.10 RBSP slice trailing bits semantics cabac_zero_word is a byte-aligned sequence of two bytes equal to 0x0000. Let NumBytesInVclNALunits be the sum of the values of NumBytesInNALunit for all VCL NAL units of a coded picture. When entropy_coding_mode_flag is equal to 1, the number of bins resulting from decoding the contents of all VCL NAL units of a coded picture shall not exceed ( 32 ÷ 3 ) * NumBytesInVclNALunits + 96 * PicSizeInMbs. NOTE – The constraint on the maximum number of bins resulting from decoding the contents of the slice layer NAL units can be met by inserting a number of cabac_zero_word syntax elements to increase the value of NumBytesInVclNALunits. Each cabac_zero_word is represented in a NAL unit by the three-byte sequence 0x000003 (as a result of the constraints on NAL unit contents that result in requiring inclusion of an emulation_prevention_three_byte for each cabac_zero_word). [quote JM8.0] /*! ************************************************************************ * \brief * Terminates the arithmetic codeword, writes stop bit and stuffing bytes (if any) ************************************************************************ */ void arienco_done_encoding(EncodingEnvironmentPtr eep) { put_one_bit_plus_outstanding((Elow >> (B_BITS-1)) & 1); put_one_bit((Elow >> (B_BITS-2))&1); put_one_bit(1); stat->bit_use_stuffingBits[img->type]+=(8-Ebits_to_go); while (Ebits_to_go != 8) put_one_bit(0); eep->E= eep->E*8 + eep->C; // no of processed bins eep->B= (*Ecodestrm_len - eep->B); // no of written bytes //initializing: eep->B = *code_len; eep->E -= (img->current_mb_nr-img->currentSlice->start_mb_nr); // nr=number eep->E = (eep->E + 31)>>5; // eep->E now contains the minimum number of bytes for the NAL unit } __________________________________ Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web http://birthday.yahoo.com/netrospective/ From alexismt comcast.net Fri Mar 4 10:10:55 2005 From: alexismt comcast.net (Alexis Michael Tourapis) Date: Sat Mar 5 00:23:04 2005 Subject: [Mp4-tech] IDR related questions in H.264 JV Reference encoder In-Reply-To: <4999AF9D8147074283CCE2330986362E2F60E6@iris.dolby.net> Message-ID: <200503041811.j24IBRSG012134@lists1.magma.ca> Dear Hee-Yong, Check the two parameters in the encoder named "PyramidRefReorder" and "PocMemoryManagement". These are used as examples for reordering and MMCO respectively. We have added these relatively recently to better support the pyramid structures. Unfortunately the codec does not provide any more flexible setting of these functionalities, but I think if you see these examples you may figure out what to do. i.e. you could add similar functions but for creating an open GOP structure. You may also want to consider how the num_ref_idx_lX_active parameters are set. Best regards, Alexis -----Original Message----- From: Kim, Hee-yong [mailto:Hee-yong.Kim@cinea.com] Sent: Friday, March 04, 2024 8:00 AM To: Alexis Michael Tourapis; mp4-tech@lists.mpegif.org Cc: Mercier, Guillaume; Levine, Scott Subject: RE: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Dear Alexis, Thank you for your kind explanation. But I have not seen any MMCO option in the JV reference encoder. And my understanding is that JV encoder is using only sliding window buffer memory management. Please correcr me if I am wrong. Thanks. Hee-Yong hee-yong.kim@cinea.com -----Original Message----- From: Alexis Michael Tourapis [mailto:alexismt@comcast.net] Sent: Thu 3/3/2024 3:17 PM To: Kim, Hee-yong; mp4-tech@lists.mpegif.org Cc: Mercier, Guillaume; Kim, Hee-yong; Levine, Scott Subject: Re: [Mp4-tech] IDR related questions in H.264 JV Reference encoder Dear Hee-yong, Regarding Q1, your problem is that you are using IDR frames. IDR frames enforce all references in the buffer to be discarded and therefore resulting in a "closed" GOP structure. However, for most applications, IDR frames might not be necessary. Instead, what you may wish to do is to carefully manage your reference buffer either through MMCO commands or careful reordering/ reference restrictions. Then I coded pictures need not be IDRs and still enable trick modes or whatever other features you may want. To make it more clear, what you could do is to use MMCO commands to discard all prior to I references, only after the "ending" B coded pictures have been encoded. Another method you could do is to restrict the following pictures not to refer to any of the preceeding pictures with clever management in the encoder. Again this would have the same effect. Now regarding your claim in Q2, this is actually not true. I would strongly suggest studying the parameters "PyramidCoding" and "ExplicitPyramidFormat". Setting the first one to 2 would allow you to create rather flexible coding orders or GOPs. Although it may appear that it has some limitations, (i.e. seems to require a periodic formulation) it is possible with simple modifications to even provide the coding structure of the entire sequence with these parameters. Finally, for Q3, I guess there is no real answer here since it all depends on your requirements/application/goal. I.e. you may choose not to use I coded pictures if you wish, have a specific, adaptive, or arbitrary coding order, use an adaptive mix of slice types etc. However i guess you can first try my suggestions in Q1. Best regards, Alexis > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], > [systems], [general] or another apppropriate identifier to indicate the type of > question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited. From garysull windows.microsoft.com Fri Mar 4 11:52:02 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Sat Mar 5 00:25:15 2005 Subject: [Mp4-tech] [h.264]Why is there a constraint on max number ofbins...? Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D699FC8@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> In a sense, there was a fear that the compression performance could be too good. We had to stop that from happening, for job security reasons. :-) Actually... The issue (discovered by Frank Bossen of NTT DoCoMo) is that in certain highly-artificial circumstances (such as the deliberate creation of "evil" bitstreams), it might be possible for the intent of imposing practicality by placing a constraint on the maximum bit rate to be circumvented by a huge explosion in the number of decoded symbols produced per coded bit by the CABAC parsing process. If not corrected, this situation would have allowed the creation of a bitstream that was technically conforming but that could not be decoded properly by ordinary practical decoder implementations. So an explicit bound was placed (at the picture level) on the amount of compression that could take place between the bin representation (which is like an ordinary VLC representation) and the output of the arithmetic encoder. If the bound is exceeded, the encoder must insert some "filler bits" to keep the total under the bound. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of +> changhsu@yahoo.com +> Sent: Thursday, March 03, 2024 3:12 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] [h.264]Why is there a constraint on max +> number ofbins...? +> +> Will anyone please explain the underlying reason for +> this constrain? +> +> [quote from the standard doc] +> When entropy_coding_mode_flag is equal to 1, the +> number of bins resulting from decoding the contents of +> all VCL NAL units of a coded picture shall not exceed +> ( 32 ?? 3 ) * NumBytesInVclNALunits + 96 * +> PicSizeInMbs. +> +> +> __________________________________________________ +> Do You Yahoo!? +> Tired of spam? Yahoo! Mail has the best spam protection around +> http://mail.yahoo.com +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From garysull windows.microsoft.com Fri Mar 4 12:27:06 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Sat Mar 5 00:27:13 2005 Subject: [Mp4-tech] [H.264/AVC] CABAC Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D69A098@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Ricky, It might be a good idea to read the following paper, which is available at http://iphome.hhi.de/marpe/pub.htm. D. Marpe, H. Schwarz, and T. Wiegand: Context-Based Adaptive Binary Arithmetic Coding in the H.264 / AVC Video Compression Standard (invited paper), IEEE Transactions on Circuits and Systems for Video Technology, vol. 13, no. 7, pp. 620-636, July 2003. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of rikcy fei +> Sent: Thursday, March 03, 2024 8:38 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] [H.264/AVC] CABAC +> +> Hello, experts: +> I don't know how the initialization tables in CABAC +> are made, i.e how +> the values of m and n in(m,n) are gotten, can anyone give +> me a detailed +> explanation. +> Thanks so much. +> paranoia +> +> +> rickyfyg@hotmail.com +> +> _________________________________________________________________ +> ???? MSN Explorer: http://explorer.msn.com/lccn/ +> +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From garysull windows.microsoft.com Fri Mar 4 13:16:37 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Sat Mar 5 00:29:03 2005 Subject: [Mp4-tech] H.263 v2/v3 Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D69A1AA@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Phillipe, It may be helpful to know whether you are looking for product implementations, reference software for academic algorithm testing, open-source implementation code on the web, etc. A better place to ask about H.263 is in VCEG. You can subscribe to the VCEG email reflector by visiting http://www.imtc.org, clicking on "activity groups" and looking down low on the page where it provides a VCEG email reflector subscription link. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of +> Philippe De Neve +> Sent: Thursday, March 03, 2024 8:15 AM +> To: Mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] H.263 v2/v3 +> +> Hi folks, +> +> maybe this is not the correct place to post this message but +> I do not know +> where to start looking elsewhere. +> +> Does anybody have any idea where I can find a h.263 v2 or +> h.263 v3 (aka +> h.263 +/++) codec? Or where I can look for it? +> +> tnx for all replies, +> regards, Philippe. +> +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From jammahesh gmail.com Sat Mar 5 12:58:42 2005 From: jammahesh gmail.com (arul mahesh) Date: Sun Mar 6 11:33:07 2005 Subject: [Mp4-tech] CABAC in H.264/AVC In-Reply-To: References: Message-ID: <9a1d782d05030503586cee4f14@mail.gmail.com> Hi Ricky, The conext initialisation is done from the file context_ini.c ( function init_contexts( ) ) and the contexts are intialised for motion and texture coding info... you can find two data structures (motion and texture in global header file ), these two data structures hold the context values for each SLICE Type and these contexts are updated for each bin value by the arithmetic coder. For each syntax element, depending on Slice type and model number the corresponding set of data array is selected from ctx_tables.h and assigned into the data structures (motion or texture) by the function init_contexts( ). Then these data structures are used for further calculations. It is generally as follows. Get the Slice type and model number, then the element to be accessed is indexed by the m and n values. you can understand it better if u analyse the function biari_init_context() in biariencode.c Hope, your doubts are cleared... ;-), if not, please let me know.. regards -- Arul On Fri, 04 Mar 2024 04:34:25 +0000, rikcy fei wrote: > Hello, experts: > > I don't know how the initialization tables in CABAC are made, i.e > how > > the values of m and n in(m,n) are gotten, can anyone give me a detailed > > explaination. > > Thanks so much. > > paranoia > > rickyfyg@hotmail.com > > > > _________________________________________________________________ > ?????????????? MSN Messenger: http://messenger.msn.com/cn > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php > From ksuehring web.de Sat Mar 5 13:15:36 2005 From: ksuehring web.de (Karsten Suehring) Date: Sun Mar 6 11:33:13 2005 Subject: [Mp4-tech] [H.264]about the constraint of max number of bins for a coded picture In-Reply-To: <20050304175943.77680.qmail@web53803.mail.yahoo.com> References: <20050304175943.77680.qmail@web53803.mail.yahoo.com> Message-ID: <4229A2E8.5010408@web.de> > It looks like the JM does not use the same equation as > it is in the standard document. Is there a mismatch > between JM and the standard? Yes, this has already been discovered recently. I have a fix in my development branch which will be contained in the next release. Best regards, Karsten From pt_david yahoo.com Sun Mar 6 03:37:56 2005 From: pt_david yahoo.com (Prasanth T David) Date: Mon Mar 7 09:29:46 2005 Subject: [Mp4-tech] Deblocking Filter - redundant code in GetStrength ? Message-ID: <20050306113756.3620.qmail@web53510.mail.yahoo.com> hello experts, In the GetStrength() function in Deblocking filter inside the for idx=1 to 16 loop..the Strength is evaluated for each 16 pixels of an Edge. But in the part where the strength depends on the motion vector, dec_picture->mv[LIST_0] is assigned to list0_mv. similarly for LIST_1. I find that the element dec_picture->mv[LIST_0][subblocky][subblockx][component] is updated in the main decode_macroblock( )function on a per MB basis ie the values for all the MBs are not stored in this array. But the DeblockMB() function which calls the GetStrength() is on a Frame basis. Hence for all the MBs ***list0_mv=dec_picture->mv[LIST_0] is the same for all the MBs.If so whats the use of evaluating these conditions in the GetStrenght() on a MB basis let alone on a pixel basis? Is this a bug or am I missing something ? I would like to know how the information mv[LIST_0][subblocky][subblockx][component] which changes for each MB is taken into the GetStrength() from the JM code I dont see how. I am using JM 8.1 Can anyone help please Thanks regards, David __________________________________ Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web http://birthday.yahoo.com/netrospective/ From Poornima.Narasimhan lntinfotech.com Mon Mar 7 13:46:54 2005 From: Poornima.Narasimhan lntinfotech.com (Poornima Narasimhan) Date: Mon Mar 7 09:32:31 2005 Subject: [Mp4-tech] mp3PRO decoder Message-ID: hi frens, i am working on the development of mp3PRO decoder. I am stuck on writing header information and huffman decodding. Will be very very greatful if some pointers for standard or reference is provided. Poornima Narasimhan, ______________________________________________________________________ -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050307/556f2c6d/attachment.html From satish_babudv yahoo.com Mon Mar 7 06:01:35 2005 From: satish_babudv yahoo.com (satish babu) Date: Tue Mar 8 05:58:37 2005 Subject: [Mp4-tech] regarding H.264 Message-ID: <20050307140135.40405.qmail@web52508.mail.yahoo.com> --------------------------------- Celebrate Yahoo!'s 10th Birthday! Yahoo! Netrospective: 100 Moments of the Web -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050307/dea14c0e/attachment.html From Philippe.deneve intec.ugent.be Mon Mar 7 17:10:18 2005 From: Philippe.deneve intec.ugent.be (Philippe De Neve) Date: Tue Mar 8 06:01:05 2005 Subject: [Mp4-tech] H.263 v2/v3 References: <91D7F2CEE3425A4A9D11311D09FCE2460D69A1AA@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Message-ID: <00a501c52330$27bbf940$bbd6c19d@King> Hi Gary, I'm looking for an open-source implementation or commercial product. I've understand that ffmpeg supports the H.263+ codec. I think this is through the use of libavcodec. I'll. get into this. Thanx for the directions guys, Philippe. ----- Original Message ----- From: "Gary Sullivan" To: "Philippe De Neve" ; Sent: Friday, March 04, 2024 10:16 PM Subject: RE: [Mp4-tech] H.263 v2/v3 > Phillipe, > > It may be helpful to know whether you are looking for product > implementations, reference software for academic algorithm testing, > open-source implementation code on the web, etc. > > A better place to ask about H.263 is in VCEG. You can subscribe to the > VCEG email reflector by visiting http://www.imtc.org, clicking on > "activity groups" and looking down low on the page where it provides a > VCEG email reflector subscription link. > > Best Regards, > > Gary Sullivan > > +> -----Original Message----- > +> From: mp4-tech-bounces@lists.mpegif.org > +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of > +> Philippe De Neve > +> Sent: Thursday, March 03, 2024 8:15 AM > +> To: Mp4-tech@lists.mpegif.org > +> Subject: [Mp4-tech] H.263 v2/v3 > +> > +> Hi folks, > +> > +> maybe this is not the correct place to post this message but > +> I do not know > +> where to start looking elsewhere. > +> > +> Does anybody have any idea where I can find a h.263 v2 or > +> h.263 v3 (aka > +> h.263 +/++) codec? Or where I can look for it? > +> > +> tnx for all replies, > +> regards, Philippe. > +> > +> _______________________________________________ > +> NOTE: Please use clear subject lines for your posts. Include > +> [audio, [video], [systems], [general] or another > +> apppropriate identifier to indicate the type of question you have. > +> > +> Note: Conduct on the mailing list is subject to the > +> Antitrust guidelines found at > +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant > +> itrust.php > +> > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php > From Thammi_Reddy Satyam.com Thu Mar 10 10:30:41 2005 From: Thammi_Reddy Satyam.com (Thammi_Reddy) Date: Thu Mar 10 05:46:11 2005 Subject: [Mp4-tech] H.264/decoder/Baseline/mvd_l0 Message-ID: <3D9FDCA910DD4445896A0BE5ECF009D4F5C0EE@bla.satyam.com> Hi all, This is regarding syntax element mvd_l0[][][]. The descriptor for mvd_l0 is given as se(v) OR ae(v). For baseline profile it is not ae(v). And mvd)l0 is explained only in CABAC chapter(9.3). So, please tell me whether I receive this in the encoded bit-stream for baseline profile decoder? Thank you. Thammi Reddy ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050310/6ccff921/attachment.html From faithful_seagull yahoo.com.au Fri Mar 11 11:46:01 2005 From: faithful_seagull yahoo.com.au (K.) Date: Fri Mar 11 14:19:34 2005 Subject: [Mp4-tech] how does JM decoder handles missing frames and why JM9.3 decoder gave me segmentation fault when trying to decode a bitstream which has gaps in the frame number and the other version of JM decoded the same bitstream without segmentation fault? Message-ID: <20050311114601.82841.qmail@web31306.mail.mud.yahoo.com> Hi Experts, I am trying to understand how JM reference decoder handles gaps in frames. However, before I can try to understand this, I encountered a perculiar problem. I used JM9.3 (and JM9.2 I believe) to try to decode some 264 bitstreams (which has some missing frames), JM9.3/JM9.2 drops out at the missing frame (eg. frames9, 10, 11, 12, 13 are missing, the next available frame is frame 14) and tells me segmentation fault, and the trance file recorded that JM decodes the frame before the missing frame, and the recording stops at the frame after the missing frame which was when the segmentation fault happened. Using the same bitstreams with another older version (which I only have the decoder but do not have the entire software, and hence unable to find out the version of this decoder) of JM decoded the bitstream without segmentation fault. My question is why would one version of JM works and the another gave segmentation fault? Is it a bug in JM or is there some explanation about this? Could any one tell me which version of JM decoder would handle gaps in frames? Thanks a lot in advance, Kathy Send instant messages to your online friends http://uk.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050311/ec1fd62b/attachment.html From garysull windows.microsoft.com Wed Mar 9 21:39:56 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Fri Mar 11 14:22:14 2005 Subject: [Mp4-tech] H.264/decoder/Baseline/mvd_l0 Message-ID: <91D7F2CEE3425A4A9D11311D09FCE24608D0635F@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Yes, of course you receive it, whichever way it is entropy coded and regardless of profile. If something is in the syntax table under a particular set of conditions, you will always get it when those conditions are fulfilled. No exceptions. There is no special explanation for the parsing process for that syntax element because there is just one parsing process for all syntax elements that are represented as se(v). No special discussion is therefore necessary -- se(v) is fully specified. Near the beginning of clause 7 you will find an explanation of how to interpret syntax table entries that have two alternative parsing processes (one of which is ae(v)). Best Regards, -Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org on behalf of Thammi_Reddy Sent: Wed 3/9/2023 9:00 PM To: Mp4-tech@lists.mpegif.org Subject: [Mp4-tech] H.264/decoder/Baseline/mvd_l0 Hi all, This is regarding syntax element mvd_l0[][][]. The descriptor for mvd_l0 is given as se(v) OR ae(v). For baseline profile it is not ae(v). And mvd)l0 is explained only in CABAC chapter(9.3). So, please tell me whether I receive this in the encoded bit-stream for baseline profile decoder? Thank you. Thammi Reddy ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050309/015d6dd1/attachment.html From niz5 caramail.com Thu Mar 10 11:16:21 2005 From: niz5 caramail.com (ET) Date: Fri Mar 11 14:23:27 2005 Subject: [Mp4-tech] error control for real time video Message-ID: <1110453381004081@lycos-europe.com> An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050310/0a1ef5ed/attachment.html From ingo.langmeier thales-ee.com Thu Mar 10 13:57:01 2005 From: ingo.langmeier thales-ee.com (Ingo Langmeier) Date: Fri Mar 11 14:24:33 2005 Subject: [Mp4-tech] parsing of DCT coefficients Message-ID: <44DCF3173FE277498D96DAAA10758B17043CA7@EXCHANGE.winserver.thales-ee.com> Dear all, I am currently developing a bitstream parser for mpeg-4 (simple profile L0-3). I have following question: The procedure VideoObjectLayer (clause 6.2.3 in 14496-2:2001(E)) calls the procedure 'video_plane_with_short_header' (6.2.5.2) when next_bits !=vol_start_code; in this case, the AC/DC coefficients are read in by calling the procedure macroblock and block(i). In normal mode (no short header mode), the procedure VideoObjectPlane (clause 6.2.5) is processed,but in the pseudo-code I do not find the entry-point for calling the procedures macroblock (6.2.6) resp. block (6.2.7), therefore I don't know when to start decoding of mcbpc,cbpy,dct_dc_size_luminance ... Thanks in advance, Ingo -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050310/b94f47e0/attachment.html From mdorlando univ.trieste.it Thu Mar 10 15:21:48 2005 From: mdorlando univ.trieste.it ( Marco D'Orlando) Date: Fri Mar 11 14:25:25 2005 Subject: [Mp4-tech] [AUDIO]MPEG AAC Message-ID: <003601c5257c$7f6dc580$7632698c@Olimpo> Dear all, I'm looking for an error resilient audio codec for some test in an erorr-prone environment. I think that AAC codec handle errors in the stream and support scalability of the bitstream. Is it true?Where can I find source code of the codec? Thanx, Marco -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050310/9affb15e/attachment.html From sdey pace.stpp.soft.net Fri Mar 11 15:13:36 2005 From: sdey pace.stpp.soft.net (Soumen Kumar Dey) Date: Fri Mar 11 14:26:10 2005 Subject: [Mp4-tech] H.264/decoder/Baseline/mvd_l0 References: <3D9FDCA910DD4445896A0BE5ECF009D4F5C0EE@bla.satyam.com> Message-ID: <00c901c5261e$cc7bd8b0$8564a8c0@psil> H.264/decoder/Baseline/mvd_l0For decoding of mvd_l0 in baseline profile, se_v is used and se_v is explained. Just decode it and add with predicted motion vector to get the actual motion vector. regards, Soumen ----- Original Message ----- From: Thammi_Reddy To: Mp4-tech@lists.mpegif.org Sent: Thursday, March 10, 2024 10:30 AM Subject: [Mp4-tech] H.264/decoder/Baseline/mvd_l0 Hi all, This is regarding syntax element mvd_l0[][][]. The descriptor for mvd_l0 is given as se(v) OR ae(v). For baseline profile it is not ae(v). And mvd)l0 is explained only in CABAC chapter(9.3). So, please tell me whether I receive this in the encoded bit-stream for baseline profile decoder? Thank you. Thammi Reddy ************************************************************************** This email (including any attachments) is intended for the sole use of the intended recipient/s and may contain material that is CONFIDENTIAL AND PRIVATE COMPANY INFORMATION. Any review or reliance by others or copying or distribution or forwarding of any or all of the contents in this message is STRICTLY PROHIBITED. If you are not the intended recipient, please contact the sender by email and delete all copies; your cooperation in this regard is appreciated. ************************************************************************** ------------------------------------------------------------------------------ _______________________________________________ NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050311/804219cc/attachment-0001.html From rlei ati.com Fri Mar 11 10:01:22 2005 From: rlei ati.com (Ryan Lei) Date: Fri Mar 11 16:05:38 2005 Subject: [Mp4-tech] H263/mpeg4 decoder Message-ID: Hi, Does any know some comercial mpeg4/h263 video decoder or play that can save the decoded yuv data? Ryan Lei, Software Engineer, Video Codec Handheld Products Department | ATI Technology Inc. 1.905.882.2600x2712 | www.ati.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050311/abfc1248/attachment.html From Yubin.Zhu esstech.com Fri Mar 11 09:10:30 2005 From: Yubin.Zhu esstech.com (Yu-Bin Zhu) Date: Sat Mar 12 12:22:18 2005 Subject: [Mp4-tech] Latest official H.264 spec Message-ID: <89976087EE37FA4D8684C71F318DD7D901E34941@ess2kmail.essnet.com> Hi experts, I'd like to know where I can find the latest official H.264 spec(I have JVT-G050r1). By the way, can I get the latest WMV9(VC-1) spec for free? thanks, Yubin -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050311/90ad2429/attachment.html From ralph.sperschneider iis.fraunhofer.de Fri Mar 11 18:34:01 2005 From: ralph.sperschneider iis.fraunhofer.de (Ralph Sperschneider) Date: Sat Mar 12 12:26:20 2005 Subject: [Mp4-tech] Re: Error calculation MPEG4 AAC LC Decoder output PCM samples In-Reply-To: <20050208053205.42144.qmail@web52006.mail.yahoo.com> References: <20050208053205.42144.qmail@web52006.mail.yahoo.com> Message-ID: <4231D689.5030003@iis.fraunhofer.de> mallikarjun rao wrote: > Hi Everybody, > > > > This is Mallikarjun. I downloaded some test vectors from the site: > http://www.iso.org/iso/en/ittf/PubliclyAvailableStandards/ISO_IEC_14496-4_2000_Conformance_Testing/CONFORMANCE_BITSTREAMS_CD2/AUDIO/GA/AAC_LC/ > The data at that place are outdated. Instead, use what you find here: ftp://mpaudconf:adif2mp4@ftp.iis.fraunhofer.de/ Ralph -- Dipl.-Ing. Ralph Sperschneider | Phone: +49 9131 776 344 Fraunhofer IIS | Fax: +49 9131 776 67 344 Am Wolfsmantel 33 | mailto:ralph.sperschneider@iis.fraunhofer.de D 91058 Erlangen | http://www.iis.fraunhofer.de/amm/ From ralph.sperschneider iis.fraunhofer.de Fri Mar 11 18:52:44 2005 From: ralph.sperschneider iis.fraunhofer.de (Ralph Sperschneider) Date: Sat Mar 12 12:27:45 2005 Subject: [Mp4-tech] Re: Enhanced AACPlus [AUDIO] In-Reply-To: References: Message-ID: <4231DAEC.8040806@iis.fraunhofer.de> Andreas Schneider wrote: > Yes, it is possible to implement the functionality these two libraries > provide on your own. > There's currently a 3G-meeting ongoing. If I'm not mistaken, one expected > output of this meeting is a way for people to obtain source code for the > ISO media library. > Source code for the audio file libary will not be available. You may use > any other library that provides the same functionality. > > Regards, > > Andreas > Source code for the AFsp library is available here: ftp://ftp.tsp.ece.mcgill.ca/pub/AFsp/ Ralph -- Dipl.-Ing. Ralph Sperschneider | Phone: +49 9131 776 344 Fraunhofer IIS | Fax: +49 9131 776 67 344 Am Wolfsmantel 33 | mailto:ralph.sperschneider@iis.fraunhofer.de D 91058 Erlangen | http://www.iis.fraunhofer.de/amm/ From ralph.sperschneider iis.fraunhofer.de Fri Mar 11 18:57:58 2005 From: ralph.sperschneider iis.fraunhofer.de (Ralph Sperschneider) Date: Sat Mar 12 12:29:30 2005 Subject: [Mp4-tech] Re: Non-sbr streams in HE-AAC decoder In-Reply-To: References: Message-ID: <4231DC26.1060304@iis.fraunhofer.de> Andreas Schneider wrote: > Hello, > > >>According to HE-AAC spec, the max AAC sampling frequency for non-SBR >>streams (AAC only streams) for levels 2/3 decoder is 48kHz . >>(Table 1.8A of N5570 doc) >> >>1) Is my understanding right? Why is such a restriction imposed? > > > Yes, your understanding is correct. This restriction has been imposed to > limit the requirements of such a decoder. > Decoding an AAC bitstream with a sampling rate of 96 kHz requires > (approx.) twice the computational ressources than decoding a 48 kHz AAC > bitstream. So by restricting what the decoder has to support to 48 kHz, a > decoder manufacturer may use a slower and potentially cheaper chip while > still being compliant to HE-AAC Profile and AAC Profile Level 2. > > >>2) If an HEAAC decoder, decodes AAC-only streams with sampling freq >>greater than 48kHz, will the decoder be considered not compliant to >>HEAAC spec? > > > No, the decoder will still be compliant to Level 2 of the HE-AAC Profile. > The standard has the following statement: > "[...]Hence, a decoder supporting the High Efficiency AAC Profile at a > given level can decode an AAC Profile stream of the same or a lower level. > " ... but might decode anything else in addition ... It is not in the spirit of MPEG to limit decoders by any means. There are only requirements what a decoder *must* do, but no one telling anything what in addition a decoder *must* *not* do. So your HE-AAC decoder might as well decode (e.g.) AAC scalable and will still be a compliant HE-AAC decoder. Ralph -- Dipl.-Ing. Ralph Sperschneider | Phone: +49 9131 776 344 Fraunhofer IIS | Fax: +49 9131 776 67 344 Am Wolfsmantel 33 | mailto:ralph.sperschneider@iis.fraunhofer.de D 91058 Erlangen | http://www.iis.fraunhofer.de/amm/ From garysull windows.microsoft.com Fri Mar 11 12:30:19 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Tue Mar 15 14:11:36 2005 Subject: [Mp4-tech] how does JM decoder handles missing frames and why JM9.3decoder gave me segmentation fault when trying to decode abitstream which has gaps in the frame number and the otherversion of JM decoded the same bitstream without segmentation fault? Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D87C168@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Karsten, Alexis, Kathy et al, Possible decoder reference software bug report below. Some forms of "missing frames" are allowed in the standard and some are not. For example, if gaps_in_frame_num_value_allowed_flag is equal to 0 or if fixed_frame_rate_flag is equal to 1, or if some of the "non-missing frames" refer to some of the "missing frames" for inter prediction, then I think you're probably talking about a non-conforming bitstream and the decoder is not obliged to be graceful in handling that (although it would be nice to get a good informative error message and perhaps good concealment behavior rather than just a segmentation fault). Someone may need a copy of an example such bitstream to debug. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of K. Sent: Friday, March 11, 2024 3:46 AM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] how does JM decoder handles missing frames and why JM9.3decoder gave me segmentation fault when trying to decode abitstream which has gaps in the frame number and the otherversion of JM decoded the same bitstream without segmentation fault? Hi Experts, I am trying to understand how JM reference decoder handles gaps in frames. However, before I can try to understand this, I encountered a perculiar problem. I used JM9.3 (and JM9.2 I believe) to try to decode some 264 bitstreams (which has some missing frames), JM9.3/JM9.2 drops out at the missing frame (eg. frames9, 10, 11, 12, 13 are missing, the next available frame is frame 14) and tells me segmentation fault, and the trance file recorded that JM decodes the frame before the missing frame, and the recording stops at the frame after the missing frame which was when the segmentation fault happened. Using the same bitstreams with another older version (which I only have the decoder but do not have the entire software, and hence unable to find out the version of this decoder) of JM decoded the bitstream without segmentation fault. My question is why would one version of JM works and the another gave segmentation fault? Is it a bug in JM or is there some explanation ! about this? Could any one tell me which version of JM decoder would handle gaps in frames? Thanks a lot in advance, Kathy -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050311/92fdff0c/attachment-0001.html From garysull windows.microsoft.com Sat Mar 12 11:03:14 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Tue Mar 15 14:16:20 2005 Subject: [Mp4-tech] Latest official H.264 spec Message-ID: <91D7F2CEE3425A4A9D11311D09FCE24608D06364@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Yubin et al, The ITU-T and ISO/IEC both have "version 2" published texts of H.264/MPEG-4 AVC, which is one major iteration beyond what you have. There has not been any official published integrated text of H.264/AVC since version 2, although the ITU will publish the JVT's "version 4" soon (exactly how soon, I don't know and am trying to find out - but the approval process has been completed). If you are a member of MPEG or the ITU, you can get a more recent integrated editor's draft, but if you're not, then you have two choices: 1) wait 2) piece together the current state of the document from the revision-list documents that are available on the JVT site. This would require adding two lists of changes to the version 2 text (outputs of the July 2004 and January 2005 meetings). To get a copy of the SMPTE VC-1 spec, you should join SMPTE. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org on behalf of Yu-Bin Zhu Sent: Fri 3/11/2023 9:10 AM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] Latest official H.264 spec Hi experts, I'd like to know where I can find the latest official H.264 spec(I have JVT-G050r1). By the way, can I get the latest WMV9(VC-1) spec for free? thanks, Yubin -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050312/5dc70618/attachment.html From oldpisces hotmail.com Sun Mar 13 09:51:23 2005 From: oldpisces hotmail.com (=?gb2312?B?s8Ig4bA=?=) Date: Tue Mar 15 14:24:54 2005 Subject: [Mp4-tech] where can I find the document about Tmn8 Message-ID: hi ,all ITU-T/SG15, Video Codec Test Model, Near-Term, TMN 8, ITU Study Group 16, Video Coding Experts Group, Document Q15-A-59, Portland, USA, June 1997. I can't find this document and the other information about it.Can you help me? Thank you very much! _________________________________________________________________ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn From highfliergal2000 yahoo.com Sun Mar 13 13:44:06 2005 From: highfliergal2000 yahoo.com (sharmili arumugam) Date: Tue Mar 15 14:49:38 2005 Subject: [Mp4-tech] FRAME DROP Message-ID: <20050313134406.61040.qmail@web30502.mail.mud.yahoo.com> Dear experts, I am working on video encoding and decoding for my project. I noticed that, when I decode a video sequence after processing it in my system, there is a frame drop about 2-3 frames. Does anyone knows how to explain this situation? Thanks in advance! ________________________________________________________________________ Yahoo! Messenger - Communicate instantly..."Ping" your friends today! Download Messenger Now http://uk.messenger.yahoo.com/download/index.html From guogr bjnec.nec.com.cn Mon Mar 14 14:27:37 2005 From: guogr bjnec.nec.com.cn (guogr@bjnec.nec.com.cn) Date: Tue Mar 15 14:49:44 2005 Subject: [Mp4-tech] 3gp/H.263 Message-ID: Hello all, I'm an freshman about the 3gp. played the 3gp file by the player, only I frames can be decoded, all P frames will be jumped. I don't know why. I think there is something wrong with the file format.the attached file is the *.3gp file. I will appreciate if anyone can help me. Thanks. (See attached file: temp.3gp) Cheery Guo -------------- next part -------------- A non-text attachment was scrubbed... Name: temp.3gp Type: application/octet-stream Size: 37648 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20050314/9043422b/temp-0001.obj From sathishn teneoris.com Mon Mar 14 13:30:21 2005 From: sathishn teneoris.com (Sathish N) Date: Tue Mar 15 14:49:49 2005 Subject: [Mp4-tech] MPEG2 Free/Commercial Encoder Message-ID: <1110807085.31144.31.camel@sathish> Experts, I am in search of an MPEG2 encoder which will allow the user to 1) control the picture type of each field of a picture e.g. To encode both Top & bottom fields of a picture as I Fields [ Most of the encoders I've tried encode (only) the Top field as I & the bottom field as P ] 2) bypass 16x8 and Dual Prime prediction for Field prediction. Please let me know if any free/commercial encoder is available which offers the above controls to the user. Thanks Sathish From faithful_seagull yahoo.com.au Tue Mar 15 01:36:16 2005 From: faithful_seagull yahoo.com.au (K.) Date: Tue Mar 15 14:49:54 2005 Subject: [Mp4-tech] How to do bit exact test for a conformance bitstream which has missing frames? Message-ID: <20050315013616.84151.qmail@web31301.mail.mud.yahoo.com> Hi Experts, I have been trying to understand how JM reference decoder handles conformance bitstreams such as MR3_TANDBERG_B.264 with reference to the H.264 standard. I would really appreciate it if any one can give me some opinions about this. What I manage to understand so far is that JM reference decoder generates some "non-existing" frames through "alloc_storable_picture(...)" and then sets some parameters to ensure conformance, but I couldn't see how the sample values are created, did JM just use some existing frames in the decoded picture buffer? >From subclause 8.2.5.2, it says that "the sample values of the generated frames may be set to any value." But if the sample values can be set to any value, then HOW could we test for BIT EXACTNESS? since our decoder and JM decoder may not use the same sample values to generate the non-existing frames for a conformance bitstream such as MR3_TANDBERG_B.264. I'd be grateful if you could give me some help here. Thank you in advance, Kathy Send instant messages to your online friends http://uk.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050315/a2c8dccf/attachment.html From delvecch cefriel.it Tue Mar 15 14:56:22 2005 From: delvecch cefriel.it (Luca Del Vecchio) Date: Tue Mar 15 14:49:58 2005 Subject: [Mp4-tech] H.264 motion vectors Message-ID: <9F1223587A755D47805AECAB7ECAF5D9022A603F@swing.cefriel.it> Hi all, I'm implementing an H.264 decoder and it seems that JM9.2 encoder may generate motion vectors referring to pixels outside the picture boundaries. >From my understanding of the code, in these cases the borderline pixels are repeated as necessary to get the complete reference. Is it correct? I've tried to verify what observed before looking into the standard, (my copy is an old one: JVT-G050r1.doc), but I was not able to find something about that. Could you help me? Best regards. Luca Del Vecchio From garysull windows.microsoft.com Tue Mar 15 20:08:44 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Wed Mar 16 08:14:05 2005 Subject: [Mp4-tech] where can I find the document about Tmn8 Message-ID: <91D7F2CEE3425A4A9D11311D09FCE24608D06372@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> That document, along with the other documents from that meeting, can be found at: http://ftp3.itu.ch/av-arch/video-site/9706_Por/. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org on behalf of ? ? Sent: Sat 3/12/2023 5:51 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] where can I find the document about Tmn8 hi ,all ITU-T/SG15, Video Codec Test Model, Near-Term, TMN 8, ITU Study Group 16, Video Coding Experts Group, Document Q15-A-59, Portland, USA, June 1997. I can't find this document and the other information about it.Can you help me? Thank you very much! _________________________________________________________________ ??????????,??? MSN Messenger: http://messenger.msn.com/cn _______________________________________________ NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050315/6e677d94/attachment.html From garysull windows.microsoft.com Tue Mar 15 20:12:10 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Wed Mar 16 08:14:10 2005 Subject: [Mp4-tech] How to do bit exact test for a conformance bitstreamwhich has missing frames? Message-ID: <91D7F2CEE3425A4A9D11311D09FCE24608D06374@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Those samples can be set to any value because those "phantom pictures" are never output by the decoder and they are never referred to in the inter prediction process for the decoding of subsequent pictures in the bitstream. In fact the decoder doesn't need to actually generate picture arrays and fill them with anything - it just needs to act as if such pictures existed. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org on behalf of K. Sent: Mon 3/14/2005 5:36 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] How to do bit exact test for a conformance bitstreamwhich has missing frames? Hi Experts, I have been trying to understand how JM reference decoder handles conformance bitstreams such as MR3_TANDBERG_B.264 with reference to the H.264 standard. I would really appreciate it if any one can give me some opinions about this. What I manage to understand so far is that JM reference decoder generates some "non-existing" frames through "alloc_storable_picture(...)" and then sets some parameters to ensure conformance, but I couldn't see how the sample values are created, did JM just use some existing frames in the decoded picture buffer? >From subclause 8.2.5.2, it says that "the sample values of the generated frames may be set to any value." But if the sample values can be set to any value, then HOW could we test for BIT EXACTNESS? since our decoder and JM decoder may not use the same sample values to generate the non-existing frames for a conformance bitstream such as MR3_TANDBERG_B.264. I'd be grateful if you could give me some help here. Thank you in advance, Kathy Send instant messages to your online friends http://uk.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050315/aeb9103a/attachment.html From ksuehring web.de Wed Mar 16 11:07:22 2005 From: ksuehring web.de (Karsten Suehring) Date: Wed Mar 16 11:14:06 2005 Subject: [Mp4-tech] H.264 motion vectors In-Reply-To: <9F1223587A755D47805AECAB7ECAF5D9022A603F@swing.cefriel.it> References: <9F1223587A755D47805AECAB7ECAF5D9022A603F@swing.cefriel.it> Message-ID: <4238055A.1070406@web.de> Luca, in the luma and chroma sample interpolation the sample locations are clipped to the border positions. This is logically the same like repeating the samples outside the picture. See equation 8-184 for luma samples and 8-206 to 8-213 for chroma samples (JVT-G050r1) Best regards, Karsten Luca Del Vecchio wrote: > Hi all, > > I'm implementing an H.264 decoder and it seems that JM9.2 encoder may > generate motion vectors referring to pixels outside the picture > boundaries. >>From my understanding of the code, in these cases the borderline pixels > are repeated as necessary to get the complete reference. Is it correct? > > I've tried to verify what observed before looking into the standard, (my > copy is an old one: JVT-G050r1.doc), but I was not able to find > something about that. Could you help me? > > Best regards. > > Luca Del Vecchio > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php From faithful_seagull yahoo.com.au Wed Mar 16 12:28:12 2005 From: faithful_seagull yahoo.com.au (K.) Date: Wed Mar 16 13:04:40 2005 Subject: [Mp4-tech] gaps in frame num Message-ID: <20050316122812.35749.qmail@web31305.mail.mud.yahoo.com> Hi Experts, Thank you very much for your help, especially thanks to Gary and Glenn for taking the time to answer my questions and I have found your answers very helpful and am convinced that the "phantom frames" are not being output. However, I am still confused about some aspects of this gaps in frame num business and hope you don't mind to answer a few more questions of mine: :) 1a) In subclause 8.2.5.2, it states that the generated frames shall be marked as "used for short term reference", if these frames are not being used and not being output, why would it be marked as "used for short term reference" for? 1b) Does sliding window process mark these non-existing frames to "unsed for reference"? 2a) Also, if these frames are not being used any where, why would JM reference decoder taking the trouble to generate and store these frames? 2b) In JM reference decoder, the attribute "is_output" of these generated frames are set to '1' , doesn't this mean that these frames are to be output? Am I allow to set it to '0'? 2c) The bitstream "MR3_TANDBERG_B.YUV" has a picture order count type set to 2 (and has missing frames), In JM reference decoder, if picture order count type is not equal to one, the function "decode_poc" would be invoked, but I am not interested in interlaced coding yet the output of "decode_poc" are TopFieldOrderCnt and BottomFieldOrderCnt, how do I use this function for a frame coding bitstream with picture order count type set to 2? (Since these frames are not being used, can I simply ignore the decoding of POC?) 3) Are these generated frames' frame num significant (since they must be generated using equation 7-14 of the standard) ? I mean, is there somewhere in the code (after the generation of these frames) that uses these frame numbers even if they aren't to be output? 4) Are these generated frame nums being used in reference picture marking? 5) Finally, where would I be wrong if I say, "As long as these generated frames are generated using equation 7-14 and marked as 'non-existing' , the rest can be anything I want" ? Thank you all very much, Kathy Send instant messages to your online friends http://uk.messenger.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050316/45be6be4/attachment-0001.html From garysull windows.microsoft.com Wed Mar 16 11:02:47 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Thu Mar 17 00:10:03 2005 Subject: [Mp4-tech] gaps in frame num Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D9C9354@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> See below. Best Regards, Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of K. Sent: Wednesday, March 16, 2024 4:28 AM To: mp4-tech@lists.mpegif.org; Kathy Sia Subject: [Mp4-tech] gaps in frame num Hi Experts, Thank you very much for your help, especially thanks to Gary and Glenn for taking the time to answer my questions and I have found your answers very helpful and am convinced that the "phantom frames" are not being output. However, I am still confused about some aspects of this gaps in frame num business and hope you don't mind to answer a few more questions of mine: :) 1a) In subclause 8.2.5.2, it states that the generated frames shall be marked as "used for short term reference", if these frames are not being used and not being output, why would it be marked as "used for short term reference" for? [GJS: ] This is just to make the DPB handling process act as if the frames were there. It might be a good idea for you to study the "sub-sequence"-related SEI messages and the old contribution documents by Miska Hannuksela and Y-Kui Wang to get a better understanding of how this can be useful. 1b) Does sliding window process mark these non-existing frames to "unused for reference"? [GJS: ] Yes, when the time comes. 2a) Also, if these frames are not being used any where, why would JM reference decoder taking the trouble to generate and store these frames? [GJS: ] I guess one way to make sure that the decoder is acting like the frames are there is to actually create them. But it might not be the fastest way if you're trying to implement a real product-quality decoder. 2b) In JM reference decoder, the attribute "is_output" of these generated frames are set to '1' , doesn't this mean that these frames are to be output? Am I allow to set it to '0'? [GJS: ] I'm not familiar with what the reference software is doing in that regard. All I can tell you is that subclause C.4.2 of the standard says that the "phantom" frames are marked as "not needed for output". 2c) The bitstream "MR3_TANDBERG_B.YUV" has a picture order count type set to 2 (and has missing frames), In JM reference decoder, if picture order count type is not equal to one, the function "decode_poc" would be invoked, but I am not interested in interlaced coding yet the output of "decode_poc" are TopFieldOrderCnt and BottomFieldOrderCnt, how do I use this function for a frame coding bitstream with picture order count type set to 2? (Since these frames are not being used, can I simply ignore the decoding of POC?) [GJS: ] Strictly speaking, POC (officially) is not an interlacing indication. POC is something that influences the values of the decoded pictures. It is about the decoding process, not the capture or display process. You must use POC to build a conforming decoder (at least for profiles other than Baseline). But if you're building an encoder for progressive-scan video, I would probably recommend encoding the video in a way that the Top and Bottom POC values of every frame are equal. I believe that if you use pic_order_cnt_type equal to 2 and always use field_pic_flag equal to 0, Top and Bottom POC values will always be equal. 3) Are these generated frames' frame num significant (since they must be generated using equation 7-14 of the standard) ? I mean, is there somewhere in the code (after the generation of these frames) that uses these frame numbers even if they aren't to be output? [GJS: ] I think they determine the initial reference picture list order for list 0 in P slices. I think they also affect the DPB management (e.g., in the sliding window process and they provide a way to refer to a specific frame when marking something as "unused for reference" or "used for long-term reference"). They also affect POC calculations under some circumstances (such as pic_order_cnt_type equal to 2). 4) Are these generated frame nums being used in reference picture marking? [GJS: ] Yes. 5) Finally, where would I be wrong if I say, "As long as these generated frames are generated using equation 7-14 and marked as 'non-existing' , the rest can be anything I want" ? [GJS: ] You may also need to generate POC and to track the marking of the pictures. For decoding a conforming bitstream, you should not ever actually need to access the sample values in the pictures, so you don't really even need to have a frame store for them in memory. Thank you all very much, Kathy -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050316/aebe2f35/attachment.html From Hee-yong.Kim cinea.com Wed Mar 16 15:05:50 2005 From: Hee-yong.Kim cinea.com (Kim, Hee-yong) Date: Thu Mar 17 00:12:50 2005 Subject: [Mp4-tech] CABAC Related Problem in Reference Coder(JM9.3) Message-ID: <4999AF9D8147074283CCE2330986362E2F60EF@iris.dolby.net> Dear all, I have decoding failure while doing encoding and decoding the following: ------------------------------------------------------------------------ -------------------- Picture size: 720x576 (4:2:2) Video: Mobile Frame Rate: 25 Hz Profile: 122 Level: 40 Picture Structure: IBBPBBPBBPBBI IDRIntraEnable:0 (Intra Picture is not IDR picture except 1st) BreferncePicture:1 (B frame is reference frame) PictureInterlace: 2 (Adaptive Frame/Field) MbInterlace: 2 (Adaptive Frame/Field) SymbolMode: 1 (CABAC) NumberReferenceFrames: 5 ------------------------------------------------------------------------ ------------------------- I did encoding and decoding. During decoding, there was a Segment Violation in the middle by referecing frame number 5 which is NULL pointer and is out of DPB if it is a frame based prediction. I encoded and decoded twice to make sure this and they were same. I used DDD and at least I know where it was, but could not nail down all the way. However, I did not have a problem when I encoded with CAVLC. So, I think that it is probably related to CABAC. I wonder that anyone had the similar problem I had and please let me know. Also when I look at the code, 'read_one_macroblock()' in "image.c", it is hard to follow by H.264 syntax. It is probably equivalent code but I wish it follows the syntax as much as possible. Also, the notation is not the same as H.264. For example, skip_flag in the macroblock seems to be used the other way around. By syntax, skip_flag=1 means that MB is skipped, but the code negates the bits. Also, MB_type is at first assigned to skip_flag when it reads skip_flag , and then later it is overwritten when it really reads MB_type. Regards, Hee-Yong Kim Hee-yong.kim@cinea.com ----------------------------------------- This message (including any attachments) may contain confidential information intended for a specific individual and purpose. If you are not the intended recipient, delete this message. If you are not the intended recipient, disclosing, copying, distributing, or taking any action based on this message is strictly prohibited. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050316/829176d8/attachment.html From alexismt comcast.net Thu Mar 17 00:40:01 2005 From: alexismt comcast.net (Alexis Michael Tourapis) Date: Thu Mar 17 06:00:28 2005 Subject: [Mp4-tech] CABAC Related Problem in Reference Coder(JM9.3) Message-ID: <031720050040.20965.4238D1E0000DAB2B000051E522007511509B039C07980A040E@comcast.net> Skipped content of type multipart/alternative-------------- next part -------------- An embedded message was scrubbed... From: "Kim, Hee-yong" Subject: [Mp4-tech] CABAC Related Problem in Reference Coder(JM9.3) Date: Thu, 17 Mar 2024 00:10:47 +0000 Size: 1009 Url: /pipermail/mp4-tech/attachments/20050317/547a3f24/attachment-0001.mht From katayen yahoo.com.tw Thu Mar 17 14:18:21 2005 From: katayen yahoo.com.tw (Heng-Lin Yen) Date: Thu Mar 17 08:40:59 2005 Subject: [Mp4-tech] [H.264] Complementary field pair In-Reply-To: 6667 Message-ID: <20050317061821.17116.qmail@web17602.mail.tpe.yahoo.com> Dear Experts, I have studied the ISO/IEC 14496-10 for a while, but still confused by some basic definitions. As my understanding, 1. One coded picture is denoted that either "one coded frame" or "one coded field". 2. One coded frame is derived from "only one" access unit. In this case, the top and bottom field are derived together. 3. Two consecutive coded fields, in order of that "top field follows by bottom field", are derived from two consecutive access units individually, and formed a complementary field pair. 4. So in decoded order, within one complementary field pair, one top field is derived from one access unit first, then one bottom field is derived from another access unit. Where am I wrong? Thanks to your correction in advance. Best regards, Kata _______________________________________________________________________ Yahoo!集筿獺絚 禣甧秖250MB獺ンぃ┤ http://tw.promo.yahoo.com/mail_new/index.html From pavan_dsp yahoo.co.in Thu Mar 17 14:39:01 2005 From: pavan_dsp yahoo.co.in (pavan kumar) Date: Thu Mar 17 22:07:23 2005 Subject: [Mp4-tech] [H.264] one or more slice groups Message-ID: <20050317143901.95006.qmail@web8404.mail.in.yahoo.com> Hi can anyone could give me a pointers for the input *.264 file to H.264 decoder which should have pps->num_slice_groups_minus1 not equal to "0".i.e for a Picture(frame) which has one or more slice groups. presently i am working with the foreman( 3 frames,QCIF ) on FMO . Thanks Regards Pavan Yahoo! India Matrimony: Find your life partneronline. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050317/854eea19/attachment.html From pesh northwestern.edu Thu Mar 17 10:48:54 2005 From: pesh northwestern.edu (Peshala V. Pahalawatta) Date: Thu Mar 17 22:10:02 2005 Subject: [Mp4-tech] [Video] Decoding Frame by Frame Message-ID: <4239B4F6.8070502@northwestern.edu> Hi! I'm new to the reference codec and I'm trying to decode one frame at a time from the bitstream using a reference frame or frames taken from an input file. I'm thinking that I could input the reference frames in the init_dpb() function but any help on what to lookout for will be appreciated. Thanks, Pesh -- -- _ ______________________________________ From oamato wanadoo.fr Thu Mar 17 19:51:35 2005 From: oamato wanadoo.fr (Olivier Amato) Date: Thu Mar 17 22:10:56 2005 Subject: [Mp4-tech] Oversized MP4 files with ffmpeg and XviD References: <031720050040.20965.4238D1E0000DAB2B000051E522007511509B039C07980A040E@comcast.net> Message-ID: <003401c52b22$6c053450$0a00000a@tototxoxqlsjoa> I'm getting a rate control issue when using ffmpeg to encode in 2-pass mode with XviD codec : the resulting bitrate is far higher than the target I'm indicating. Here are the 2 command lines I'm using : ffmpeg -i input_file.avi -vcodec xvid -b 1000 -g 125 -pass 1 -passlogfile log_file output_file.mp4 ffmpeg -i input_file.avi -vcodec xvid -b 1000 -g 125 -pass 2 -passlogfile log_file output_file.mp4 The log file is well produced and seems to be valid, but the resulting "output_file.mp4" has an average bitrate of 3655 kbps ( instead of 1000 kbps ) !!?? When I try the same process with mpeg4 codec ( just replacing "-vcodec xvid" by "-vcodec mpeg4" ), the encoded file then has an average bitrate of 1005 kbps as expected. BTW, when I'm encoding in 1 pass only with Xvid, I don't have this rate control issue anymore. The following command line produces a resulting file with an average bitrate of 1018 kbps : ffmpeg -i input_file.avi -vcodec xvid -b 1000 -g 125 output_file.mp4 Would somebody know how to solve this issue ( I've already posted this question to ffmpeg-user-list without feedback ... ) ? Olivier From garysull windows.microsoft.com Thu Mar 17 11:05:00 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Thu Mar 17 22:11:42 2005 Subject: [Mp4-tech] [H.264] Complementary field pair Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460D9CA224@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Kata et al, Sounds basically right, with the following caveats: 1) In a complementary field pair, there is no constraint that the top field will be first. The bottom field can be first. This was also true in MPEG-2. 2) Non-paired fields are also supported. That was not true in MPEG-2. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Heng-Lin Yen +> Sent: Wednesday, March 16, 2024 10:18 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] [H.264] Complementary field pair +> +> Dear Experts, +> +> I have studied the ISO/IEC 14496-10 for a while, +> but still confused by some basic definitions. +> +> As my understanding, +> 1. One coded picture is denoted that either "one coded +> frame" or "one coded field". +> +> 2. One coded frame is derived from "only one" access +> unit. +> In this case, the top and bottom field are derived +> together. +> +> 3. Two consecutive coded fields, in order of that "top +> field follows by bottom field", are derived from two +> consecutive access units individually, and formed a +> complementary field pair. +> +> 4. So in decoded order, within one complementary field +> pair, one top field is derived from one access unit +> first, then one bottom field is derived from another +> access unit. +> +> Where am I wrong? Thanks to your correction in +> advance. +> +> Best regards, +> Kata +> +> _____________________________________________________________ +> __________ +> Yahoo!?????? +> ????250MB???????? +> http://tw.promo.yahoo.com/mail_new/index.html +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From mg-li prolific.com.tw Fri Mar 18 10:08:49 2005 From: mg-li prolific.com.tw (=?big5?B?TGkuTWcgW6f1slHB4l0=?=) Date: Fri Mar 18 07:43:13 2005 Subject: [Mp4-tech] about H.264/AVC standard document Message-ID: Dear experts, I'm also confused about the case of MacroBlock to Slice group of 8.2.2.8 in ITU-T H.264 (2003/05) version. 8.2.2.8: "Otherwise (frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0), the macroblock to slice group map is specified by: MbToSliceGroupMap[ i ] = mapUnitToSliceGroupMap[ ( i / ( 2 * PicWidthInMbs ) ) * PicWidthInMbs + ( i % PicWidthInMbs ) ]" when parameter : frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0 --> it means this is a coded frame with all the MacroBlocks are field MarcoBlocks ? BR, MG Li _____ [Mp4-tech] about H.264/AVC standard document Gary Sullivan garysull windows.microsoft.com Mon Jan 24 09:24:36 ESTEDT 2005 * Previous message: [Mp4-tech] AAC transparency bitrates for sampling rates other than 44.1? * Next message: [Mp4-tech] MPEG4, MPEG2, and MPEG1 * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] _____ With field_pic_flag equal to 0 and mb-aff turned off, I think you're talking about an ordinary frame. BTW you should get ahold of something more recent than G050r1. That's from several iterations ago. -G. ________________________________ From: mp4-tech-bounces lists.mpegif.org [mailto:mp4-tech-bounces lists.mpegif.org ] On Behalf Of ??? Sent: Monday, January 24, 2024 3:01 AM To: mp4-tech lists.mpegif.org Subject: [Mp4-tech] about H.264/AVC standard document hello,experts: In h.264/AVC standard document(JVT-G050r1,Page 86) there is a paragraph about FMO decoding: "Otherwise (frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0), the macroblock to slice group map is specified by: MbToSliceGroupMap[ i ] = mapUnitToSliceGroupMap[ ( i / ( 2 * PicWidthInMbs ) ) * PicWidthInMbs + ( i % PicWidthInMbs ) ] (8-27)" who can tell me what's the picture type(frame or field or others) if "frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0"? thanks! ronggang -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050124/5856f54f/attachment.html _____ * Previous message: [Mp4-tech] AAC transparency bitrates for sampling rates other than 44.1? * Next message: [Mp4-tech] MPEG4, MPEG2, and MPEG1 * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] _____ More information about the Mp4-tech mailing list -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050318/f044441f/attachment-0001.html From mrukant gmail.com Fri Mar 18 11:14:38 2005 From: mrukant gmail.com (MRUKANT POPAT) Date: Fri Mar 18 07:43:19 2005 Subject: [Mp4-tech] need MPEG4 interactive content Message-ID: <23155bfd050317214448fda4da@mail.gmail.com> hi all, I am looking for interactive MPEG-4 content. Can any one tell me if it is available on internet for free download. I would also like to know if there is any freely available MP4 streaming server which supports interactivity. regards, MRUKANT From getsunil_6 yahoo.co.in Fri Mar 18 05:11:51 2005 From: getsunil_6 yahoo.co.in (venkata sunil) Date: Sat Mar 19 14:18:23 2005 Subject: [Mp4-tech] AAN Fast IDCT Algorithm In-Reply-To: <200412071726.iB7HMQ66017992@lists1.magma.ca> Message-ID: <20050318131151.22270.qmail@web8506.mail.in.yahoo.com> I am Looking for a Fast IDCT Algorithm. If any one have the below paper please send me. Y. Arai, T. Agui, and M. Nakajima. A fast DCT-SQ scheme for images. Trans. IEICE, vol. E-71, no. 11, pp. 1095-1097, Nov. 1988. Thanks in Advance, Sunil. Sunil. --------------------------------- Do you Yahoo!? Yahoo! Small Business - Try our new resources site! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050318/7e240f66/attachment.html From Wesley.DeNeve ugent.be Fri Mar 18 20:34:22 2005 From: Wesley.DeNeve ugent.be (Wesley De Neve) Date: Sat Mar 19 14:23:03 2005 Subject: [Mp4-tech] need MPEG4 interactive content References: <23155bfd050317214448fda4da@mail.gmail.com> Message-ID: <00a001c52bf1$7c8aa1a0$0200a8c0@Persephone> Hi, MRUKANT POPAT wrote: > hi all, > > I am looking for interactive MPEG-4 content. Can any one tell me if it > is available on internet for free download. I would also like to know The GPAC project and the IBM Toolkit for MPEG-4 are good starting points for obtaining free interactive MPEG-4 Content. http://gpac.sourceforge.net/ (look for the GPAC Regression Test Suite in the Downloads section) http://sourceforge.net/forum/?group_id=84101 http://www.alphaworks.ibm.com/tech/tk4mpeg4 However, the current license policy seems to impose a heavy burden on the usage of MPEG-4 Systems tools for the creation and distribution of (free) interactive MPEG-4 content. > if there is any freely available MP4 streaming server which supports > interactivity. The Darwin Streaming Server from Apple Computer should allow you to stream interactive scenes, and this in combination with the MP4Box tool from the GPAC project (for the necessary hinting). http://developer.apple.com/darwin/projects/streaming/ http://sourceforge.net/forum/?group_id=84101 Best regards, -Wesley From zhweizh mailst.xjtu.edu.cn Sun Mar 20 15:26:29 2005 From: zhweizh mailst.xjtu.edu.cn (zhweizh) Date: Sun Mar 20 21:32:58 2005 Subject: [Mp4-tech] source coder with chaanel coding Message-ID: <200503200726.j2K7Qkuv004960@lists1.magma.ca> Hi, hello everyone, where can I find the video coding VM with channel coding [RS(n,k) coding]. Because I check video transmission effect in noise channel. many thanks! zhweizh From mpeg4maniac rediffmail.com Mon Mar 21 06:59:29 2005 From: mpeg4maniac rediffmail.com (MPEGFOUR MANIAC) Date: Mon Mar 21 07:22:31 2005 Subject: [Mp4-tech] MP4 STBs in market yet? Message-ID: <20050321070040.29190.qmail@webmail50.rediffmail.com> ? hi friends, I was looking for MPEG-4 STBs in market which are stable and production version. Can any one tell me which company has one which support both DVB and IP. I had heard that DIRECTV is starting MPEG4 broadcast this year and they also have STBs with them. Does anybody know which (company) STB they are planning to use. thank you, MPEG4 MANIAC -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050321/571e768e/attachment.html From satish_babudv yahoo.com Mon Mar 21 01:59:23 2005 From: satish_babudv yahoo.com (satish babu) Date: Mon Mar 21 13:52:08 2005 Subject: [Mp4-tech] Regarding finding "nextMbaddr" of current macroblock in H.264 decoder Message-ID: <20050321095923.74358.qmail@web52510.mail.yahoo.com> hi Experts, Presently I am working on H.264 decoder(Baseline). In finding the next macroblock address why standard is using "MbToslicegroupmap" when each slice is getting "first_mb_in_slice" in its header?can anyone explain me what is the actual use of "MbToslicegroupmap"? Regards, satish babu. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050321/11c6f64e/attachment.html From ravimpeg4video yahoo.co.in Mon Mar 21 14:41:36 2005 From: ravimpeg4video yahoo.co.in (ravi kumar) Date: Mon Mar 21 21:51:48 2005 Subject: [Mp4-tech] Does there any free Video encoder and decoder and its reference code is available on the net In-Reply-To: <200503170605.j2H60uHT003362@lists1.magma.ca> Message-ID: <20050321144136.72249.qmail@web8302.mail.in.yahoo.com> Hi All, Does there any free Video encoder and decoder and its reference code is available on the net? If yes please provide the link Regards Ravi Yahoo! India Matrimony: Find your life partneronline. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050321/180b3d7c/attachment.html From Wesley.DeNeve ugent.be Mon Mar 21 22:08:10 2005 From: Wesley.DeNeve ugent.be (Wesley De Neve) Date: Mon Mar 21 21:51:53 2005 Subject: [Mp4-tech] Regarding finding "nextMbaddr" of current macroblock inH.264 decoder References: <20050321095923.74358.qmail@web52510.mail.yahoo.com> Message-ID: <007d01c52e5a$1643bec0$0200a8c0@Persephone> Hi, satish babu wrote: >> hi Experts, >> >> Presently I am working on H.264 decoder(Baseline). >> >> In finding the next macroblock address why standard is using >> "MbToslicegroupmap" when each slice is getting "first_mb_in_slice" >> in its header?can anyone explain me what is the actual use of >> "MbToslicegroupmap"? The usage of MBToSliceGroupMap is related to a tool called Flexible Macroblock Ordering (FMO). The latter is a tool for error resilience within the H.264/AVC specification. JVT documents JVT-B027, JVT-C048, and JVT-C089 are good starting points for learning more about the concepts as used by FMO. Best regards, Wesley De Neve From kylee astri.org Tue Mar 22 09:40:13 2005 From: kylee astri.org (Lee Ka Yuk) Date: Tue Mar 22 09:30:17 2005 Subject: [Mp4-tech] [MP4 System] MP3 in MP4, DSS seeking Message-ID: Hi all, 1) MP3 in MP4 file standardized? I came across a 14496-3 Amd3 (N5575 in March 2003). It mentioned about MP3 in MP4. I wonder if there is any update on this doc, or is it finalized. 2) Seeking in DSS (or by RTSP in general) If we use RTSP's PLAY command for seeking, will the first sample after seek be an RAP? How does Darwin implement seeking through RTSP? Dose it read the nearest RAP in the hint track? Thanks. From lichu1025 yahoo.com.tw Tue Mar 22 10:43:53 2005 From: lichu1025 yahoo.com.tw (=?big5?q?=C4R=A6=CB?=) Date: Tue Mar 22 09:30:23 2005 Subject: [Mp4-tech] About H.264 data partition Message-ID: <20050322024353.81616.qmail@web17902.mail.tpe.yahoo.com> Dear all: I have some questions about H.264 data partition.We know that H.264 partition the encoded bitstream to three partition:partition A,B and C,and adopt the UEP(unequal error protection) to protect the important data.Is it the best partition mode? How can I find the paper about the data partition thesis??please give me some advises. Thanks a lot. Best Regards. Judy --------------------------------- Yahoo!集筿獺絚 禣甧秖250MB獺ンぃ┤弧 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050322/8d9d072f/attachment.html From gripened gmail.com Tue Mar 22 09:43:13 2005 From: gripened gmail.com (Jayant Chauhan) Date: Tue Mar 22 09:30:28 2005 Subject: [Mp4-tech] MP4 Decoder Variable Frame Size[Video] Message-ID: <5b996acb05032120133c3cb0f@mail.gmail.com> Hey all I have a doubt. I have a MP4 Simple Profile Decoder. Now the problem I am facing is that when I try to play a mp4 file with variable frame size video, my media player crashes, because it seems that the decoder initializes stuff first and as the frame size changes, it is not able to get to the VOLparameters, and hence fails. Does a Simple Profile decoder support variable frame sizes ?!! Could you give me a way out of this problem ?! As in, how can I make these kinds of videos work. with regards Jayant From jammahesh gmail.com Tue Mar 22 10:57:34 2005 From: jammahesh gmail.com (arul mahesh) Date: Tue Mar 22 14:00:43 2005 Subject: [Mp4-tech] Re: cabac In-Reply-To: References: Message-ID: <9a1d782d0503220157548ed344@mail.gmail.com> Dear vamshi, When ctx_idx is equal to 276, (for instant) the initial values associated are state = 63 and MPS = 0, where State = 63 represents a "Non-adapting probability state". Whereas ctx_idx = 276 is associated with the end of slice flag and the bin of mb_type, which specifies the I_PCM macroblock type. You can understand it better if you go through the ISO standard thoroughly. So these contexts are mentioned as CTX_UNUSED. These are used for contexts initialisation to the motion and texture data structures. Then these values are appropriately changed by the Binary arithmetic coder of CABAC, and the same changed contexts are used for the further following bins of a syntax element. And the model number is used for selection of the Slice model. I mean, as you can see from the ctx_tables, I slice has one model and the P or B slices have three model numbers. Depending on the slice type and model number the contexts are initialised. First do understand the values from table 9.11, then proceed on to the tables 9.12, 9.13 ?and so on? Table 9.11 gives you the ctx to be assigned for a syntax element depending on their slice type. Regards, Arul Mahesh On Fri, 11 Mar 2024 10:14:30 +0530, bdr bdr wrote: > hello mahesh , > i have seen u r explanation regarding CABAC in > mp4 dicussion , i would like to clear some doubts regarding > intialaztion of context tables , i dint understand how there are > intialized , y,it is in 3D , it is like that 1st row represnt I frame > , seconfd->for p-picture and third for B-picture is my understanding > is correct and also how the ctx_unsued is used in table , can u clear > me . > > but when i am going through JVTG050 the table(9-11) indictes thats for > i slice the context intilzation is of 3-10 , but when refering code 4 > elemnt is unsed , iam confusing > with this ctx_unused how it is used , > > -------vamshi . > From tomotohara yahoo.com Tue Mar 22 15:29:41 2005 From: tomotohara yahoo.com (Tomo) Date: Wed Mar 23 07:30:19 2005 Subject: [Mp4-tech] H264: conformance test streams etc. Message-ID: <20050322232942.77864.qmail@web50905.mail.yahoo.com> Hello. Can anybody of the mailing list point me the location how to obtain h.264 conformance test streams (decoder, main profile, level 3)? And if some one knows more bit streams (hoping freely available ones) for decoder, I would like to know, too. Regards, Tomo From liuhanjun168 hotmail.com Wed Mar 23 01:33:01 2005 From: liuhanjun168 hotmail.com (liu hanjun) Date: Wed Mar 23 07:30:25 2005 Subject: [Mp4-tech] where can I find JVT019r1 and JVT50V2 ? Message-ID: Hi all, I want to take a look at JVT019r1 and JVT50V2, where can I find it? thank you! soundofsummer _________________________________________________________________ 享用世界上最大的电子邮件系统— MSN Hotmail。 http://www.hotmail.com From reshma sasken.com Wed Mar 23 17:28:44 2005 From: reshma sasken.com (Reshma Prasanna) Date: Wed Mar 23 12:32:30 2005 Subject: [Mp4-tech] [Video][H.264] Decoded Reference Picture Marking Process Message-ID: Dear H.264 Experts, I would really appreciate it if anyone could answer my questions. Qn 1) Decoded reference picture marking. Clause 8 of the H.264 std specifies that the only restriction on invoking the decoded reference picture marking process is that the decoded picture must be a reference picture, there are no constraints on the type of frame i.e. I, P, IDR etc. Can the adaptive memory control decoded reference picture marking process be invoked if the decoded picture is an I frame but not IDR? If yes, then consider the case when there are decoded reference pictures stored in the DPB, the current decoded picture is an I frame(not IDR) and MMCO = 1 is decoded. This means that some short term ref picture with picNum = CurrPicNum - (difference_of_pic_nums_minus_1 + 1) is to be marked as "unused for reference". Since the decoded picture has only I slices, the reference lists have not been constructed and hence picNum = FrameNumWrap has not been calculated for any short tm reference frames w.r.t the current picture's frame_num. Then how will the mmco = 1 operation be properly executed? The same question holds with any other mmco operation that requires comparison to reference lists, if the current decoded picture is an I frame. Qn 2) Consider this case in a H.264 encoder: The DPB is full and a reconstructed(decoded) picture (not IDR) is to be inserted into the DPB. If adaptive_ref_pic_marking_mode_flag = 0, then sliding window process will mark the short tm reference frm with smallest value of FrameNumWrap as "unused". But if say, the encoder wishes to mark the current picture as "used for long term reference", since the picture is not IDR, then adaptive_ref_pic_marking_mode_flag must be set equal to 1 and mmco = 6 must be used. In this case, sliding window process is not invoked and hence none of the frames in the DPB will be marked as "unused for reference". There will not be any space in the DPB for the current frame. My understanding is that the encoder must then find which short term frame has smallest frameNumWrap and send another mmco = 1 to set this short tm reference frame as "unused" so that the current reconstructed frame can be inserted into the DPB. Is this correct? Thanks in advance, Reshma. "SASKEN RATED THE BEST EMPLOYER IN THE COUNTRY by the BUSINESS TODAY Mercer Survey 2004" SASKEN BUSINESS DISCLAIMER This message may contain confidential, proprietary or legally Privileged information. In case you are not the original intended Recipient of the message, you must not, directly or indirectly, use, Disclose, distribute, print, or copy any part of this message and you are requested to delete it and inform the sender. Any views expressed in this message are those of the individual sender unless otherwise stated. Nothing contained in this message shall be construed as an offer or acceptance of any offer by Sasken Communication Technologies Limited ("Sasken") unless sent with that express intent and with due authority of Sasken. Sasken has taken enough precautions to prevent the spread of viruses. However the company accepts no liability for any damage caused by any virus transmitted by this email From ksuehring web.de Wed Mar 23 14:57:55 2005 From: ksuehring web.de (Karsten Suehring) Date: Wed Mar 23 22:47:30 2005 Subject: [Mp4-tech] [Video][H.264] Decoded Reference Picture Marking Process In-Reply-To: References: Message-ID: <424175E3.4050909@web.de> Dear Reshma, please see my comments inline. Reshma Prasanna wrote: > Dear H.264 Experts, > > I would really appreciate it if anyone could answer my questions. > > Qn 1) Decoded reference picture marking. > Clause 8 of the H.264 std specifies that the only restriction on invoking > the decoded reference picture marking process is that the decoded picture > must be a reference picture, there are no constraints on the type of frame > i.e. I, P, IDR etc. > > Can the adaptive memory control decoded reference picture marking process > be invoked if the decoded picture is an I frame but not IDR? Yes, that's possible. > If yes, then consider the case when there are decoded reference pictures > stored in the DPB, the current decoded picture is an I frame(not IDR) and > MMCO = 1 is decoded. This means that some short term ref picture with > picNum = CurrPicNum - (difference_of_pic_nums_minus_1 + 1) is to be marked > as "unused for reference". Since the decoded picture has only I slices, > the reference lists have not been constructed and hence picNum = > FrameNumWrap has not been calculated for any short tm reference frames > w.r.t the current picture's frame_num. Then how will the mmco = 1 > operation be properly executed? The same question holds with any other > mmco operation that requires comparison to reference lists, if the current > decoded picture is an I frame. This problem has already been identified. The text was fixed in the first corrigendum to invoke the FrameNumWrap calculation also for I slices. This change should already be included in the recommendation/standard version currently published by ITU and ISO. > Qn 2) Consider this case in a H.264 encoder: > The DPB is full and a reconstructed(decoded) picture (not IDR) is to be > inserted into the DPB. If adaptive_ref_pic_marking_mode_flag = 0, then > sliding window process will mark the short tm reference frm with smallest > value of FrameNumWrap as "unused". But if say, the encoder wishes to mark > the current picture as "used for long term reference", since the picture > is not IDR, then adaptive_ref_pic_marking_mode_flag must be set equal to 1 > and mmco = 6 must be used. In this case, sliding window process is not > invoked and hence none of the frames in the DPB will be marked as "unused > for reference". There will not be any space in the DPB for the current > frame. The unmarking must be signaled explicitly in that case. > My understanding is that the encoder must then find which short term frame > has smallest frameNumWrap and send another mmco = 1 to set this short tm > reference frame as "unused" so that the current reconstructed frame can be > inserted into the DPB. Is this correct? The encoder is free to decide which picture is marked "unused for reference" (short or long-term) as long as the unmarking is done before marking the current picture. That doesn't need to be the short-term frame with the smallest value of frameNumWrap. The corrigendum text also contains some clarifications which commands are allowed in which order and at which points the maximum number of reference frames constraint is checked. I would suggest reading the updated text on MMCO. Best regards, Karsten From mdellagn ati.com Wed Mar 23 11:40:36 2005 From: mdellagn ati.com (Mark Dell'Agnese) Date: Wed Mar 23 22:47:36 2005 Subject: [Mp4-tech] h264 4x4 transform Message-ID: Hello, I've been looking at the 4x4 transform in the H.264 encoder but I have a question about the inputs to the transform. The 4x4 transform process is split into 2 parts and shows a horizontal transform followed by a vertical transform. This is straightforward enough. However, I don't understand why the inputs to the 4x4 transform are transposed. That is, after the prediction data is subtracted from the current macroblock data for each 4x4 block, the resulting differences are transposed first, then transformed. So if the resulting 4x4 differences are (1st row) [d00, d01, d02, d03], {2nd row) [d10, d11, etc], the input to the 4x4 transform should be the first column, [d00, d10, d20, d30]. Instead it's the first row. Why is it the first row and not the first column? The reconstructed pixels are then transposed again to produce proper reconstruction values. Can we not do away with the transpose at the beginning of the transform and then another transpose at the end of the reconstruction? Thanks, Mark. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050323/cec2632d/attachment.html From kinanea eeng.dcu.ie Wed Mar 23 17:55:52 2005 From: kinanea eeng.dcu.ie (Andrew Kinane) Date: Wed Mar 23 22:47:41 2005 Subject: [Mp4-tech] [Video] MPEG-4 ACE Profile In-Reply-To: Message-ID: Hello All, I was wondering if anybody knows of a document (MPEG document, conference paper or otherwise) where a complexity analysis is done for MPEG-4 Advanced Coding Efficiency (ACE) profile. In particular I am interested in the % requirements for the SA-DCT function.... Any help would be much appreciated. kind regards - Andrew From garysull windows.microsoft.com Wed Mar 23 17:50:16 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Thu Mar 24 20:56:59 2005 Subject: [Mp4-tech] where can I find JVT019r1 and JVT50V2 ? Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460DBF0C0A@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Those do not appear to be correct JVT document numbers. JVT documents can be found at http://ftp3.itu.int/av-arch/jvt-site. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of liu hanjun +> Sent: Tuesday, March 22, 2024 5:33 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] where can I find JVT019r1 and JVT50V2 ? +> +> Hi all, +> +> I want to take a look at JVT019r1 and JVT50V2, where can +> I find it? +> +> thank you! +> +> soundofsummer +> +> _________________________________________________________________ +> ??????????????? MSN Hotmail? http://www.hotmail.com +> +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From garysull windows.microsoft.com Wed Mar 23 17:52:35 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Thu Mar 24 21:00:48 2005 Subject: [Mp4-tech] H264: conformance test streams etc. Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460DBF0C19@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Tomo et al, You should probably start at http://ftp3.itu.ch/av-arch/jvt-site/draft_conformance/. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Tomo +> Sent: Tuesday, March 22, 2024 3:30 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] H264: conformance test streams etc. +> +> Hello. +> +> Can anybody of the mailing list point me the location how to +> obtain h.264 conformance test streams (decoder, main +> profile, level 3)? +> And if some one knows more bit streams (hoping freely available ones) +> for decoder, I would like to know, too. +> +> Regards, +> +> +> Tomo +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From reshma sasken.com Thu Mar 24 11:21:56 2005 From: reshma sasken.com (Reshma Prasanna) Date: Thu Mar 24 21:01:55 2005 Subject: [Mp4-tech] [Video][H.264] Decoded Reference Pic Marking (JM 9.4) In-Reply-To: <424175E3.4050909@web.de> Message-ID: Dear Karsten, With reference to your reply below, I have the following question about the reference software JM version 9.4: In the file mbuffer.c, in the function init_lists(), if the current frame is an I frame, the updation of picNum for short term reference frames is performed and after that the code if ((currSliceType == I_SLICE)||(currSliceType == SI_SLICE)) { listXsize[0] = 0; listXsize[1] = 0; return; } causes the function to return. The updation of the long term picNum i.e. dpb.fs_ltref[i]->frame->long_term_pic_num = dpb.fs_ltref[i]->frame->long_term_frame_idx; is done only if the current frame is a P frame. Hence, if the previous frame was marked as a long term frame and the current frame is an I frame and mmco operations contain mmco = 2 (longTermPicNum = 0), then what should be done, is to set any long term picture with longTermPicNum = 0 as unused for reference. But in this case, the frame previous to the current I frame is also set as unused for reference because the longTermPicNum has not been set equal to longTermFrameIdx(since the current frame frame is I), and the default value for longTermPicNum = 0. I would think that the assignment of longTermPicIdx to longTermFramePicNum for all long term frames in the DPB should also be done if the current frame is an I frame. Is this correct? Thanks & Best Regards, Reshma. On Wed, 23 Mar 2005, Karsten Suehring wrote: > Dear Reshma, > > please see my comments inline. > > Reshma Prasanna wrote: > > Dear H.264 Experts, > > > > I would really appreciate it if anyone could answer my questions. > > > > Qn 1) Decoded reference picture marking. > > Clause 8 of the H.264 std specifies that the only restriction on invoking > > the decoded reference picture marking process is that the decoded picture > > must be a reference picture, there are no constraints on the type of frame > > i.e. I, P, IDR etc. > > > > Can the adaptive memory control decoded reference picture marking process > > be invoked if the decoded picture is an I frame but not IDR? > > Yes, that's possible. > > > If yes, then consider the case when there are decoded reference pictures > > stored in the DPB, the current decoded picture is an I frame(not IDR) and > > MMCO = 1 is decoded. This means that some short term ref picture with > > picNum = CurrPicNum - (difference_of_pic_nums_minus_1 + 1) is to be marked > > as "unused for reference". Since the decoded picture has only I slices, > > the reference lists have not been constructed and hence picNum = > > FrameNumWrap has not been calculated for any short tm reference frames > > w.r.t the current picture's frame_num. Then how will the mmco = 1 > > operation be properly executed? The same question holds with any other > > mmco operation that requires comparison to reference lists, if the current > > decoded picture is an I frame. > > This problem has already been identified. The text was fixed in the > first corrigendum to invoke the FrameNumWrap calculation also for I slices. > > This change should already be included in the recommendation/standard > version currently published by ITU and ISO. > > > Qn 2) Consider this case in a H.264 encoder: > > The DPB is full and a reconstructed(decoded) picture (not IDR) is to be > > inserted into the DPB. If adaptive_ref_pic_marking_mode_flag = 0, then > > sliding window process will mark the short tm reference frm with smallest > > value of FrameNumWrap as "unused". But if say, the encoder wishes to mark > > the current picture as "used for long term reference", since the picture > > is not IDR, then adaptive_ref_pic_marking_mode_flag must be set equal to 1 > > and mmco = 6 must be used. In this case, sliding window process is not > > invoked and hence none of the frames in the DPB will be marked as "unused > > for reference". There will not be any space in the DPB for the current > > frame. > > The unmarking must be signaled explicitly in that case. > > > My understanding is that the encoder must then find which short term frame > > has smallest frameNumWrap and send another mmco = 1 to set this short tm > > reference frame as "unused" so that the current reconstructed frame can be > > inserted into the DPB. Is this correct? > > The encoder is free to decide which picture is marked "unused for > reference" (short or long-term) as long as the unmarking is done before > marking the current picture. That doesn't need to be the short-term > frame with the smallest value of frameNumWrap. > > The corrigendum text also contains some clarifications which commands > are allowed in which order and at which points the maximum number of > reference frames constraint is checked. > > I would suggest reading the updated text on MMCO. > > Best regards, > Karsten > "SASKEN RATED THE BEST EMPLOYER IN THE COUNTRY by the BUSINESS TODAY Mercer Survey 2004" SASKEN BUSINESS DISCLAIMER This message may contain confidential, proprietary or legally Privileged information. In case you are not the original intended Recipient of the message, you must not, directly or indirectly, use, Disclose, distribute, print, or copy any part of this message and you are requested to delete it and inform the sender. Any views expressed in this message are those of the individual sender unless otherwise stated. Nothing contained in this message shall be construed as an offer or acceptance of any offer by Sasken Communication Technologies Limited ("Sasken") unless sent with that express intent and with due authority of Sasken. Sasken has taken enough precautions to prevent the spread of viruses. However the company accepts no liability for any damage caused by any virus transmitted by this email From jm_claud yahoo.com Thu Mar 24 06:56:33 2005 From: jm_claud yahoo.com (jim claud) Date: Thu Mar 24 21:02:35 2005 Subject: [Mp4-tech] help on audio codec In-Reply-To: 6667 Message-ID: <20050324145633.97053.qmail@web42110.mail.yahoo.com> Hi how much MIPS does TI55X takes for AMR -WM encoder and decoder in which company have actual implementation expertise? regs Jim --------------------------------- Do you Yahoo!? Yahoo! Small Business - Try our new resources site! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050324/2bb56010/attachment.html From jm_claud yahoo.com Thu Mar 24 07:07:44 2005 From: jm_claud yahoo.com (jim claud) Date: Thu Mar 24 21:03:30 2005 Subject: [Mp4-tech] doubts on audio codec Message-ID: <20050324150745.1335.qmail@web42106.mail.yahoo.com> Hi how much MIPS does TI55X takes for AMR -WM encoder and decoder in which company have actual implementation expertise? regs Jim --------------------------------- Do you Yahoo!? Yahoo! Small Business - Try our new resources site! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050324/849d79a4/attachment.html From highfliergal2000 yahoo.com Fri Mar 25 05:56:41 2005 From: highfliergal2000 yahoo.com (sharmili arumugam) Date: Fri Mar 25 12:14:47 2005 Subject: [Mp4-tech] Frame Drop Message-ID: <20050325055641.15326.qmail@web30502.mail.mud.yahoo.com> Dear Experts, I am doing some encoding and decoding of a mpeg4 video sequence and I noticed that there are a few frame drops. The original sequence is 50 frames while the decoded one only consists of 47-48 frames. Is there any possible way for me to identify which frame is dropped? Or any specific software to do so? Because I need that to calculate the exact MSE and PSNR.Please do help and thank you in advance. Ms.Sharmili Arumugam B.Eng(Hons)Electronics Majoring In Telecommunications Multimedia University Cyberjaya,Malaysia. ________________________________________________________________________ Yahoo! Messenger - Communicate instantly..."Ping" your friends today! Download Messenger Now http://uk.messenger.yahoo.com/download/index.html From stephen5_hay yahoo.com Fri Mar 25 04:49:08 2005 From: stephen5_hay yahoo.com (stephen Hay) Date: Fri Mar 25 20:01:47 2005 Subject: [Mp4-tech] Re:ans for MPEG HE+ In-Reply-To: <200503241704.j2OH15II029406@lists1.magma.ca> Message-ID: <20050325124908.9024.qmail@web42109.mail.yahoo.com> Hi Andrew For MPEG AAC LTP, HE+ DSP impl. cycle you can get good approximation from www.agenttech.org, log on to that site they work on TI processor,.. since for AMR-WB I had few implementation issues I visited that website and they gave higly reliable answers, while implementing I realised that company answer was perfect Regs Stephen Message: 3 Date: Wed, 23 Mar 2024 17:55:52 -0000 From: "Andrew Kinane" Subject: [Mp4-tech] [Video] MPEG-4 ACE Profile To: Message-ID: Content-Type: text/plain; charset="iso-8859-1" Hello All, I was wondering if anybody knows of a document (MPEG document, conference paper or otherwise) where a complexity analysis is done for MPEG-4 Advanced Coding Efficiency (ACE) profile. In particular I am interested in the % requirements for the SA-DCT function.... Any help would be much appreciated. kind regards - Andrew mp4-tech-request@lists.mpegif.org wrote: Send Mp4-tech mailing list submissions to mp4-tech@lists.mpegif.org To subscribe or unsubscribe via the World Wide Web, visit http://lists.mpegif.org/mailman/listinfo/mp4-tech or, via email, send a message with subject or body 'help' to mp4-tech-request@lists.mpegif.org You can reach the person managing the list at mp4-tech-owner@lists.mpegif.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Mp4-tech digest..." Today's Topics: 1. Re: [Video][H.264] Decoded Reference Picture Marking Process (Karsten Suehring) 2. h264 4x4 transform (Mark Dell'Agnese) 3. [Video] MPEG-4 ACE Profile (Andrew Kinane) ---------------------------------------------------------------------- Message: 1 Date: Wed, 23 Mar 2024 14:57:55 +0100 From: Karsten Suehring Subject: Re: [Mp4-tech] [Video][H.264] Decoded Reference Picture Marking Process To: Reshma Prasanna Cc: Mp4-tech@lists.mpegif.org Message-ID: <424175E3.4050909@web.de> Content-Type: text/plain; charset=ISO-8859-1 Dear Reshma, please see my comments inline. Reshma Prasanna wrote: > Dear H.264 Experts, > > I would really appreciate it if anyone could answer my questions. > > Qn 1) Decoded reference picture marking. > Clause 8 of the H.264 std specifies that the only restriction on invoking > the decoded reference picture marking process is that the decoded picture > must be a reference picture, there are no constraints on the type of frame > i.e. I, P, IDR etc. > > Can the adaptive memory control decoded reference picture marking process > be invoked if the decoded picture is an I frame but not IDR? Yes, that's possible. > If yes, then consider the case when there are decoded reference pictures > stored in the DPB, the current decoded picture is an I frame(not IDR) and > MMCO = 1 is decoded. This means that some short term ref picture with > picNum = CurrPicNum - (difference_of_pic_nums_minus_1 + 1) is to be marked > as "unused for reference". Since the decoded picture has only I slices, > the reference lists have not been constructed and hence picNum = > FrameNumWrap has not been calculated for any short tm reference frames > w.r.t the current picture's frame_num. Then how will the mmco = 1 > operation be properly executed? The same question holds with any other > mmco operation that requires comparison to reference lists, if the current > decoded picture is an I frame. This problem has already been identified. The text was fixed in the first corrigendum to invoke the FrameNumWrap calculation also for I slices. This change should already be included in the recommendation/standard version currently published by ITU and ISO. > Qn 2) Consider this case in a H.264 encoder: > The DPB is full and a reconstructed(decoded) picture (not IDR) is to be > inserted into the DPB. If adaptive_ref_pic_marking_mode_flag = 0, then > sliding window process will mark the short tm reference frm with smallest > value of FrameNumWrap as "unused". But if say, the encoder wishes to mark > the current picture as "used for long term reference", since the picture > is not IDR, then adaptive_ref_pic_marking_mode_flag must be set equal to 1 > and mmco = 6 must be used. In this case, sliding window process is not > invoked and hence none of the frames in the DPB will be marked as "unused > for reference". There will not be any space in the DPB for the current > frame. The unmarking must be signaled explicitly in that case. > My understanding is that the encoder must then find which short term frame > has smallest frameNumWrap and send another mmco = 1 to set this short tm > reference frame as "unused" so that the current reconstructed frame can be > inserted into the DPB. Is this correct? The encoder is free to decide which picture is marked "unused for reference" (short or long-term) as long as the unmarking is done before marking the current picture. That doesn't need to be the short-term frame with the smallest value of frameNumWrap. The corrigendum text also contains some clarifications which commands are allowed in which order and at which points the maximum number of reference frames constraint is checked. I would suggest reading the updated text on MMCO. Best regards, Karsten ------------------------------ Message: 2 Date: Wed, 23 Mar 2024 11:40:36 -0500 From: "Mark Dell'Agnese" Subject: [Mp4-tech] h264 4x4 transform To: mp4-tech@lists.mpegif.org Message-ID: Content-Type: text/plain; charset="us-ascii" Hello, I've been looking at the 4x4 transform in the H.264 encoder but I have a question about the inputs to the transform. The 4x4 transform process is split into 2 parts and shows a horizontal transform followed by a vertical transform. This is straightforward enough. However, I don't understand why the inputs to the 4x4 transform are transposed. That is, after the prediction data is subtracted from the current macroblock data for each 4x4 block, the resulting differences are transposed first, then transformed. So if the resulting 4x4 differences are (1st row) [d00, d01, d02, d03], {2nd row) [d10, d11, etc], the input to the 4x4 transform should be the first column, [d00, d10, d20, d30]. Instead it's the first row. Why is it the first row and not the first column? The reconstructed pixels are then transposed again to produce proper reconstruction values. Can we not do away with the transpose at the beginning of the transform and then another transpose at the end of the reconstruction? Thanks, Mark. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050323/cec2632d/attachment-0001.html ------------------------------ Message: 3 Date: Wed, 23 Mar 2024 17:55:52 -0000 From: "Andrew Kinane" Subject: [Mp4-tech] [Video] MPEG-4 ACE Profile To: Message-ID: Content-Type: text/plain; charset="iso-8859-1" Hello All, I was wondering if anybody knows of a document (MPEG document, conference paper or otherwise) where a complexity analysis is done for MPEG-4 Advanced Coding Efficiency (ACE) profile. In particular I am interested in the % requirements for the SA-DCT function.... Any help would be much appreciated. kind regards - Andrew ------------------------------ _______________________________________________ Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php End of Mp4-tech Digest, Vol 20, Issue 27 **************************************** --------------------------------- Do you Yahoo!? Yahoo! Small Business - Try our new resources site! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050325/242ae570/attachment.html From Wesley.DeNeve ugent.be Fri Mar 25 14:53:15 2005 From: Wesley.DeNeve ugent.be (Wesley De Neve) Date: Fri Mar 25 20:04:05 2005 Subject: [Mp4-tech] Frame Drop References: <20050325055641.15326.qmail@web30502.mail.mud.yahoo.com> Message-ID: <00f301c53141$fe13e880$6a01a8c0@Persephone> Hi, sharmili arumugam wrote: > Dear Experts, > > I am doing some encoding and decoding of a mpeg4 video > sequence and I noticed that there are a few frame > drops. The original sequence is 50 frames while the > decoded one only consists of 47-48 frames. Is there > any possible way for me to identify which frame is > dropped? Or any specific software to do so? Because I > need that to calculate the exact MSE and PSNR.Please > do help and thank you in advance. Your best bet is probably a side by side comparison of the decompressed frames. It may be possible that the software has dropped the last two/three frames due to the usage of a specific GOP structure. Some encoders also tend to drop the very first frames of a video sequence in order to be able to store bidirectionally compressed frames in an AVI container (for synchronization purposes). I believe one of the older versions of the DivX encoder did that. Best regards, Wesley De Neve From harish.vasudeva amd.com Fri Mar 25 10:30:27 2005 From: harish.vasudeva amd.com (Vasudeva, Harish) Date: Fri Mar 25 20:05:12 2005 Subject: [Mp4-tech] Microsoft vs MoMuSys Message-ID: <4A77E4BC05A4EF4D9DAA4C8CB5EEE51CBA31DD@SAUSEXMB1.amd.com> Hi Folks, I have ported the Microsoft Ver 2.3 of the reference code to Linux and I am seeing very low performance (may be an issue with g++). So, I just downloaded the MoMuSys-FDIS Ver 1.0 (it just compiled without any issues) and tested the same & the performance looks real good. We have been using the MS Ver 2.3 for a long time now & switching to MoMuSys seems kind of risky if it does NOT have all the features of the MS version. Does anybody know the exact differences between the two? I am looking for "Advanced Simple Profile" support. Best Regards, HARISH V -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050325/ad8d6b5e/attachment.html From Dtrikannad entriq.com Fri Mar 25 12:02:28 2005 From: Dtrikannad entriq.com (Deepak Trikannad) Date: Sat Mar 26 14:16:19 2005 Subject: [Mp4-tech] hinting Message-ID: <0DB23F0E993D514B8709179482166C0AE50EDB@us01ml2> Hey all, I was wondering if someone here could point me to a few good sites to help me learn about how to develop my own hinting application for mp4 streaming? Any help would be greatly appreciated! Thanks Deep -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050325/4e9487f1/attachment-0001.html From ysramkumar yahoo.com Sun Mar 27 01:00:04 2005 From: ysramkumar yahoo.com (Srinivas Ramkumar Yadavalli) Date: Sun Mar 27 23:30:18 2005 Subject: [Mp4-tech] Mux/Transport stream Message-ID: <20050327090004.44573.qmail@web54504.mail.yahoo.com> Hello, I have encoded a yuv frame into a H.264 format, with file extension .264 using the JVT software's lenccod.exe I need to put this on a transport stream format. Please suggest ways to do this. Problem is that, I am windows person and not much of a unix/linux person. Please help. Ramkumar __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/ From dsn2603 rediffmail.com Sun Mar 27 11:14:25 2005 From: dsn2603 rediffmail.com (sakthi narayanan) Date: Sun Mar 27 23:30:24 2005 Subject: [Mp4-tech] hinting Message-ID: <20050327111548.7349.qmail@webmail49.rediffmail.com> ? hi, U can get streaming information & free source code from mpeg4ip.net. with regards, sakthi On Sat, 26 Mar 2024 Deepak Trikannad wrote : >Hey all, I was wondering if someone here could point me to a few good sites >to help me learn about how to develop my own hinting application for mp4 >streaming? > > > >Any help would be greatly appreciated! > > > >Thanks > > > >Deep > >_______________________________________________ >NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > >Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050327/5a5ac436/attachment.html From Wesley.DeNeve ugent.be Sun Mar 27 15:00:23 2005 From: Wesley.DeNeve ugent.be (Wesley De Neve) Date: Sun Mar 27 23:30:29 2005 Subject: [Mp4-tech] hinting References: <0DB23F0E993D514B8709179482166C0AE50EDB@us01ml2> Message-ID: <039e01c532cc$f0a72450$0200a8c0@Persephone> Hi, Deepak Trikannad wrote: >> Hey all, I was wondering if someone here could point me to a few >> good sites >> to help me learn about how to develop my own hinting application for >> mp4 >> streaming? I would recommend to have a look at the source code of MP4Box or MP4Creator in case you'd like to study hinting. MP4Box is part of the GPAC project (http://gpac.sourceforge.net/), while MP4Creator is part of the MPEG4IP project (http://mpeg4ip.net/). The tools in question are quite mature and the projects are also quite active, so you should be able to get some descent help in case you're running into trouble. The following web page contains a discussion of some of the subtle differences between the two applications with respect to the creation of hinted MP4 files: http://gpac.sourceforge.net/auth_mp4box.php. In case you're interested in streaming an sich, it might be a good idea to have a look at the source code of the Darwin Streaming Server (http://developer.apple.com/darwin/projects/streaming/). Hope this helps, Wesley De Neve From hworld21st yahoo.com.cn Mon Mar 28 15:52:28 2005 From: hworld21st yahoo.com.cn (Jason Huang) Date: Mon Mar 28 09:19:12 2005 Subject: [Mp4-tech] [H264]Question about the status of FRext Message-ID: <20050328075228.80944.qmail@web15809.mail.cnb.yahoo.com> Hi, Dear Experts, What's the current status of FRext? I found a great many new tools have been proposed in 2004, but I wonder which of them has been adopted. Does the paper, "The H.264 Advanced Video Coding Standard - Overview and Introduction to the Fidelity Range Extensions", describe the final version? What's the lastest draft? Thanks Jason _________________________________________________________ Do You Yahoo!? 150万曲MP3疯狂搜，带您闯入音乐殿堂 http://music.yisou.com/ 美女明星应有尽有，搜遍美图、艳图和酷图 http://image.yisou.com 1G就是1000兆，雅虎电邮自助扩容！ http://cn.rd.yahoo.com/mail_cn/tag/1g/*http://cn.mail.yahoo.com/event/mail_1g/ From ysramkumar yahoo.com Mon Mar 28 06:08:44 2005 From: ysramkumar yahoo.com (Srinivas Ramkumar Yadavalli) Date: Tue Mar 29 14:39:55 2005 Subject: [Mp4-tech] Video/Transport H.264 In-Reply-To: 6667 Message-ID: <20050328140844.5639.qmail@web54508.mail.yahoo.com> Wesley, Not to confuse the issue, but 1. Is it correct to say that code from MPEG4IP is based on MPEG 4 PART 2? 2. These site do they have any source code or utilities for H.264? Regards, Ramkumar --- Wesley De Neve wrote: > Hi, > > Deepak Trikannad wrote: > >> Hey all, I was wondering if someone here could > point me to a few > >> good sites > >> to help me learn about how to develop my own > hinting application for > >> mp4 > >> streaming? > > I would recommend to have a look at the source code > of MP4Box or MP4Creator > in case you'd like to study hinting. MP4Box is part > of the GPAC project > (http://gpac.sourceforge.net/), while MP4Creator is > part of the MPEG4IP > project (http://mpeg4ip.net/). The tools in question > are quite mature and > the projects are also quite active, so you should be > able to get some > descent help in case you're running into trouble. > The following web page > contains a discussion of some of the subtle > differences between the two > applications with respect to the creation of hinted > MP4 files: > http://gpac.sourceforge.net/auth_mp4box.php. > > In case you're interested in streaming an sich, it > might be a good idea to > have a look at the source code of the Darwin > Streaming Server > (http://developer.apple.com/darwin/projects/streaming/). > > Hope this helps, > Wesley De Neve > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. > Include [audio, [video], [systems], [general] or > another apppropriate identifier to indicate the type > of question you have. > > Note: Conduct on the mailing list is subject to the > Antitrust guidelines found at > http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php > Regards, Ramkumar (Srinivas Ramkumar Yadavalli) __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From garysull windows.microsoft.com Mon Mar 28 10:35:42 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Tue Mar 29 14:40:01 2005 Subject: [Mp4-tech] [H264]Question about the status of FRext Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460DCCB5CB@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> FRExt is a finished project. That set of extensions has completed formal and final approval in both ITU-T and ISO/IEC. Yes, the conference paper that you refer to (available from http://www.fastvdo.com/spie04) describes the final FRExt design. The latest editors' draft of the standard is JVT-N050d1. It is available to members of the JVT, but not (yet) to the general public. Approximately equivalent documents are available, but only in "change-list" format relative to the prior text of the standard, not as "integrated" text. Best Regards, Gary Sullivan +> -----Original Message----- +> From: mp4-tech-bounces@lists.mpegif.org +> [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Jason Huang +> Sent: Sunday, March 27, 2024 11:52 PM +> To: mp4-tech@lists.mpegif.org +> Subject: [Mp4-tech] [H264]Question about the status of FRext +> +> Hi, Dear Experts, +> +> What's the current status of FRext? I found a great +> many new tools have been proposed in 2004, but I +> wonder which of them has been adopted. Does the paper, +> "The H.264 Advanced Video Coding Standard - Overview +> and Introduction to the Fidelity Range Extensions", +> describe the final version? What's the lastest draft? +> +> Thanks +> Jason +> +> +> _________________________________________________________ +> Do You Yahoo!? +> 150??MP3???????????? +> http://music.yisou.com/ +> ??????????????????? +> http://image.yisou.com +> 1G??1000??????????? +> http://cn.rd.yahoo.com/mail_cn/tag/1g/*http://cn.mail.yahoo.c +> om/event/mail_1g/ +> _______________________________________________ +> NOTE: Please use clear subject lines for your posts. Include +> [audio, [video], [systems], [general] or another +> apppropriate identifier to indicate the type of question you have. +> +> Note: Conduct on the mailing list is subject to the +> Antitrust guidelines found at +> http://www.mpegif.org/public/documents/vault/mp-out-30042-Ant +> itrust.php +> From Wesley.DeNeve ugent.be Mon Mar 28 20:39:49 2005 From: Wesley.DeNeve ugent.be (Wesley De Neve) Date: Tue Mar 29 14:40:07 2005 Subject: [Mp4-tech] Video/Transport H.264 References: <20050328140844.5639.qmail@web54508.mail.yahoo.com> Message-ID: <009e01c533c5$863b8a30$0d00a8c0@Persephone> Hi Ramkumar, Srinivas Ramkumar Yadavalli wrote: > Wesley, > > Not to confuse the issue, but > > 1. Is it correct to say that code from MPEG4IP is > based on MPEG 4 PART 2? >From a developer's point of view, one can say that part of their source code for encoding uncompressed video files is based on the reference software for MPEG-4 Part 2. However, the community behind MPEGIP is doing a lot more than that: MPEGIP provides source code that allows you to create (hinted) MP4 files, that allows you to setup a live stream (MP4Live), ... In fact, as it is stated on their website, they are providing the glue that makes it possible to integrate a lot of multimedia related packages. For instance, they are relying on the XviD-encoder to create MPEG-4 Visual elementary streams and on the Darwin Streaming Server for streaming hinted MP4 files. I would suggest to have a look at the following page in order to have a good overview of the features they offer: http://mpeg4ip.sourceforge.net/features/index.php > 2. These site do they have any source code or > utilities for H.264? They have a parser for H.264/AVC bitstreams having the Byte Stream NALU syntax, and MP4Creator is able to store those streams in an MP4 file. Have a look at their site on SourceForge in order to obtain the latest version of the source code of these tools from CVS. Best regards, -Wesley > --- Wesley De Neve wrote: >> Hi, >> >> Deepak Trikannad wrote: >>>> Hey all, I was wondering if someone here could >> point me to a few >>>> good sites >>>> to help me learn about how to develop my own >> hinting application for >>>> mp4 >>>> streaming? >> >> I would recommend to have a look at the source code >> of MP4Box or MP4Creator >> in case you'd like to study hinting. MP4Box is part >> of the GPAC project >> (http://gpac.sourceforge.net/), while MP4Creator is >> part of the MPEG4IP >> project (http://mpeg4ip.net/). The tools in question >> are quite mature and >> the projects are also quite active, so you should be >> able to get some >> descent help in case you're running into trouble. >> The following web page >> contains a discussion of some of the subtle >> differences between the two >> applications with respect to the creation of hinted >> MP4 files: >> http://gpac.sourceforge.net/auth_mp4box.php. >> >> In case you're interested in streaming an sich, it >> might be a good idea to >> have a look at the source code of the Darwin >> Streaming Server >> > (http://developer.apple.com/darwin/projects/streaming/). >> >> Hope this helps, >> Wesley De Neve >> >> _______________________________________________ >> NOTE: Please use clear subject lines for your posts. >> Include [audio, [video], [systems], [general] or >> another apppropriate identifier to indicate the type >> of question you have. >> >> Note: Conduct on the mailing list is subject to the >> Antitrust guidelines found at >> > http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php >> > > Regards, > Ramkumar > (Srinivas Ramkumar Yadavalli) > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com From Dtrikannad entriq.com Mon Mar 28 10:46:27 2005 From: Dtrikannad entriq.com (Deepak Trikannad) Date: Tue Mar 29 14:40:13 2005 Subject: [SPAM] - Re: [Mp4-tech] hinting - Email has different SMTP TO : and MIME TO: fields in the email addresses Message-ID: <0DB23F0E993D514B8709179482166C0AE50EEC@us01ml2> Thanks this is a huge help!! -----Original Message----- From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of Wesley De Neve Sent: Sunday, March 27, 2024 5:00 AM To: Deepak Trikannad; mp4-tech@lists.mpegif.org Subject: [SPAM] - Re: [Mp4-tech] hinting - Email has different SMTP TO: and MIME TO: fields in the email addresses Hi, Deepak Trikannad wrote: >> Hey all, I was wondering if someone here could point me to a few >> good sites >> to help me learn about how to develop my own hinting application for >> mp4 >> streaming? I would recommend to have a look at the source code of MP4Box or MP4Creator in case you'd like to study hinting. MP4Box is part of the GPAC project (http://gpac.sourceforge.net/), while MP4Creator is part of the MPEG4IP project (http://mpeg4ip.net/). The tools in question are quite mature and the projects are also quite active, so you should be able to get some descent help in case you're running into trouble. The following web page contains a discussion of some of the subtle differences between the two applications with respect to the creation of hinted MP4 files: http://gpac.sourceforge.net/auth_mp4box.php. In case you're interested in streaming an sich, it might be a good idea to have a look at the source code of the Darwin Streaming Server (http://developer.apple.com/darwin/projects/streaming/). Hope this helps, Wesley De Neve _______________________________________________ NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php From Stephen.Henry elixent.com Mon Mar 28 14:14:04 2005 From: Stephen.Henry elixent.com (Stephen Henry) Date: Tue Mar 29 14:40:19 2005 Subject: [Mp4-tech] AAC-HE Message-ID: <69D9787BB47F4C4A8AF38D62921F9070051332AC@MI8NYCMAIL01.Mi8.com> Hi all, I feel quite stupid for asking this but, are there any nice introductory papers or any reference source code available for the AAC-HE (or even plain AAC) that briefly describe the principles involved. I've been spoiled by the availability of such information for H.264, and I'm amazed that there doesn't appear to be anything similar for the audio aspects of the system. Thanks, Stephen Henry -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050328/cc47b077/attachment.html From mg-li prolific.com.tw Tue Mar 29 09:47:03 2005 From: mg-li prolific.com.tw (=?big5?B?TGkuTWcgW6f1slHB4l0=?=) Date: Tue Mar 29 14:40:25 2005 Subject: [Mp4-tech] [H.264]inter prediction of field MB in MBAFF coded frame Message-ID: Dear experts, In H.264 standard 8.4.2.2. How to generate (xAL,yAL) when current MB is field MB in MBAFF coded frame ? After checking reference SW JM94, in macroblock.c line 4182 to line 4194 states ¨ if (!curr_mb_field) { vec1_y = j4*4*mv_mul + mv_array[j4][i4][1]; } else --> field MB in MBAFF coded frame { if (mb_nr%2 == 0) vec1_y = (img->block_y * 2 + joff) * mv_mul + mv_array[j4][i4][1]; --> eq (1), top field MB else vec1_y = ((img->block_y-4) * 2 + joff)* mv_mul + mv_array[j4][i4][1]; --> eq (2), bottom field MB } get_block (ref_idx, list, vec1_x, vec1_y, img, tmp_block); ¨ I was confused that in eq(1) and eq(2) , where joff = j *4. Since current MB is field MB, the reference pictureˇs height is half size (one field) of original reference frame. The half effect reflect on img->block_y * 2 instead of *4. But joff is still four times of j. Does this make sense? In eq(2), why img->block_y needs to subtract 4 ? In my poor understanding, both eq(1),and eq(2) shall be Vec1_y = (img->block_y * 2 + joff/2) * mv_mul + mv_array[j4][i4][1]; Please tell me where Iˇm wrong. Thank you so much. BR, MG Li -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050329/dc174081/attachment-0001.html From dengxiongshu yulong.com Tue Mar 29 10:39:41 2005 From: dengxiongshu yulong.com (=?utf-8?B?6YKT6ZuE5Lmm?=) Date: Tue Mar 29 14:40:32 2005 Subject: [Mp4-tech] question on mpeg4 part 2 dquant Message-ID: <002201c53408$8ede6750$1a040180@dengxiongshu> dear expert: I find only When macro type is inter+q and intra+q can write dquant value in mpeg4 part 2 standard,so when macro type is inter4v can't write dquant value because have not inter4v+q macro mode in ISO/IEC 14496-2,that is to say when macro have four motion vector ,it can not change quant value in macro level. So I was question how can implement macro level rate control when a frame have a lot inter4v macro ? Have I misunderstood 14496-2 standard? I was reference ISO/IEC 14496-2 Second edition 2001-12-01.And I find xvid write dquant and decode dquant as above. Can somebody of mpeg4 part 2 expert answer my question? thanks so much xiongshu deng -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050329/c22f4fff/attachment.html From sven- gmx.de Tue Mar 29 11:16:08 2005 From: sven- gmx.de (Sven Wontroba) Date: Thu Mar 31 14:43:31 2005 Subject: [Mp4-tech] scalable formats for PDA, PC, TV In-Reply-To: <20050327090004.44573.qmail@web54504.mail.yahoo.com> References: <20050327090004.44573.qmail@web54504.mail.yahoo.com> Message-ID: <7310370020.20050329111608@gmx.de> List members, I'm a student writing my thesis about scalable formats for PDA's, PC, TV,... So far I read a dif papers about MPEG4,7, SMIL, MPEG21,... and there is one point I don't understand completely. With XMT-O and XMT-A it's posible to describe complex sceens and create some interaction,... within MPEG4 but more precisely in the MPEG21 Idem Adaption, if I understand everything correct I'm able to scale dif. object depending my device, user specification,... The point where I'm stuck is, how do I transform my complex XMT presentation into a MPEG21 presentation. I found some application to create dims at the Klagenfurt University Austria with some other helpful information. But may be there are other information out there. So far I don't understand how to create a complex presentation for one device as an abstract MPEG21 description. Some hints may be to papers(ACM, IEEE, MPEG... ) or other project would be helpful or may be a link to other lists with more information about this problem. Thanks Sven -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050329/c3fa6ce6/attachment.html From katayen yahoo.com.tw Wed Mar 30 15:35:10 2005 From: katayen yahoo.com.tw (Heng-Lin Yen) Date: Thu Mar 31 14:43:37 2005 Subject: [Mp4-tech] [H.264] Doubts about the te(v) Message-ID: <20050330073510.45426.qmail@web17605.mail.tpe.yahoo.com> Dear Experts, As writen in the H.264 in spec 9.1 for truncated EG code "te(v)", when the syntax element is coded as te(v), the range of the syntax element shall be determined first. Does the "range of the syntax element" indicates the range of syntax element vlue? In addition, does anyone kindly show the table of te(v)? Although the derivation for syntax element of te(v) is described in spec 9.1, I can't catch what it means exactly. Thanks in addvance. Best regards, Kata _______________________________________________________________________ Yahoo!集筿獺絚 禣甧秖250MB獺ンぃ┤ http://tw.promo.yahoo.com/mail_new/index.html From mg-li prolific.com.tw Wed Mar 30 17:26:46 2005 From: mg-li prolific.com.tw (=?big5?B?TGkuTWcgW6f1slHB4l0=?=) Date: Thu Mar 31 14:43:42 2005 Subject: [Mp4-tech] about H.264/AVC standard document Message-ID: Dear experts, I'm also confused about the case of MacroBlock to Slice group of 8.2.2.8 in ITU-T H.264 (2003/05) version. 8.2.2.8: "Otherwise (frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0), the macroblock to slice group map is specified by: MbToSliceGroupMap[ i ] = mapUnitToSliceGroupMap[ ( i / ( 2 * PicWidthInMbs ) ) * PicWidthInMbs + ( i % PicWidthInMbs ) ]" when parameter : frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0 --> it means this is a coded frame with all the MacroBlocks are field MarcoBlocks ? BR, MG Li _____ [Mp4-tech] about H.264/AVC standard document Gary Sullivan garysull windows.microsoft.com Mon Jan 24 09:24:36 ESTEDT 2005 * Previous message: [Mp4-tech] AAC transparency bitrates for sampling rates other than 44.1? * Next message: [Mp4-tech] MPEG4, MPEG2, and MPEG1 * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] _____ With field_pic_flag equal to 0 and mb-aff turned off, I think you're talking about an ordinary frame. BTW you should get ahold of something more recent than G050r1. That's from several iterations ago. -G. ________________________________ From: mp4-tech-bounces lists.mpegif.org [mailto:mp4-tech-bounces lists.mpegif.org ] On Behalf Of ??? Sent: Monday, January 24, 2024 3:01 AM To: mp4-tech lists.mpegif.org Subject: [Mp4-tech] about H.264/AVC standard document hello,experts: In h.264/AVC standard document(JVT-G050r1,Page 86) there is a paragraph about FMO decoding: "Otherwise (frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0), the macroblock to slice group map is specified by: MbToSliceGroupMap[ i ] = mapUnitToSliceGroupMap[ ( i / ( 2 * PicWidthInMbs ) ) * PicWidthInMbs + ( i % PicWidthInMbs ) ] (8-27)" who can tell me what's the picture type(frame or field or others) if "frame_mbs_only_flag is equal to 0 and mb_adaptive_frame_field_flag is equal to 0 and field_pic_flag is equal to 0"? thanks! ronggang -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050124/5856f54f/attachment.html _____ * Previous message: [Mp4-tech] AAC transparency bitrates for sampling rates other than 44.1? * Next message: [Mp4-tech] MPEG4, MPEG2, and MPEG1 * Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] _____ More information about the Mp4-tech mailing list -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050330/fb5e4ddf/attachment.html From mg-li prolific.com.tw Wed Mar 30 17:28:59 2005 From: mg-li prolific.com.tw (=?big5?B?TGkuTWcgW6f1slHB4l0=?=) Date: Thu Mar 31 14:43:48 2005 Subject: [Mp4-tech][H.264]inter prediction of field MB in MBAFF coded frame Message-ID: Dear experts, In H.264 standard 8.4.2.2. How to generate (xAL,yAL) when current MB is field MB in MBAFF coded frame ? After checking reference SW JM94, in macroblock.c line 4182 to line 4194 states ¨ if (!curr_mb_field) { vec1_y = j4*4*mv_mul + mv_array[j4][i4][1]; } else --> field MB in MBAFF coded frame { if (mb_nr%2 == 0) vec1_y = (img->block_y * 2 + joff) * mv_mul + mv_array[j4][i4][1]; --> eq (1), top field MB else vec1_y = ((img->block_y-4) * 2 + joff)* mv_mul + mv_array[j4][i4][1]; --> eq (2), bottom field MB } get_block (ref_idx, list, vec1_x, vec1_y, img, tmp_block); ¨ I was confused that in eq(1) and eq(2) , where joff = j *4. Since current MB is field MB, the reference pictureˇs height is half size (one field) of original reference frame. The half effect reflect on img->block_y * 2 instead of *4. But joff is still four times of j. Does this make sense? In eq(2), why img->block_y needs to subtract 4 ? In my poor understanding, both eq(1),and eq(2) shall be Vec1_y = (img->block_y * 2 + joff/2) * mv_mul + mv_array[j4][i4][1]; Please tell me where Iˇm wrong. Thank you so much. BR, MG Li -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050330/d832265a/attachment-0001.html From mrukant gmail.com Wed Mar 30 20:09:46 2005 From: mrukant gmail.com (MRUKANT POPAT) Date: Thu Mar 31 14:43:54 2005 Subject: [Mp4-tech] how does MPEGJ support MPEG-4 IPMP or control over Descrambler Message-ID: <23155bfd050330063933aa6b91@mail.gmail.com> hi, I was just looking at MPEGJ implementation & Architecture. But I could not find how MPEGJ supports MPEG-4 IPMP. MPEGJ gives ability to have control on decoding process. Could you please give me any pointers as to how does it support MPEG-4 IPMP or provides facility to access or control Descramble/Decryptor or Conditional Access module at Client Side (STB) regards, MRUKANT From garysull windows.microsoft.com Thu Mar 31 10:40:59 2005 From: garysull windows.microsoft.com (Gary Sullivan) Date: Fri Apr 1 17:05:37 2005 Subject: [Mp4-tech] question on mpeg4 part 2 dquant Message-ID: <91D7F2CEE3425A4A9D11311D09FCE2460DD8D832@WIN-MSG-10.wingroup.windeploy.ntdev.microsoft.com> Your interpretation is correct. I think I recently mentioned a weakness of the MPEG-4p2 dquant design on email on this reflector recently -- this was what I was alluding to. It can make rate control a little tricky. Basically you must find a way to decide which is more important to you -- using 4mv or using dquant. That's a kind of strange decision to need to make, but that's the way it works. It's probably not a major problem, but it's a little quirk you need to find a way around in your design. That's a little design bug that came into MPEG-4p2 from H.263v1. I recall that we put some kind of tweak into H.263+ to fix it, but somehow the MPEG-4p2 design went ahead with the old design without including the fix. Best Regards, -Gary Sullivan ________________________________ From: mp4-tech-bounces@lists.mpegif.org [mailto:mp4-tech-bounces@lists.mpegif.org] On Behalf Of ??? Sent: Monday, March 28, 2024 6:40 PM To: mp4-tech@lists.mpegif.org Subject: [Mp4-tech] question on mpeg4 part 2 dquant dear expert: I find only When macro type is inter+q and intra+q can write dquant value in mpeg4 part 2 standard,so when macro type is inter4v can't write dquant value because have not inter4v+q macro mode in ISO/IEC 14496-2,that is to say when macro have four motion vector ,it can not change quant value in macro level. So I was question how can implement macro level rate control when a frame have a lot inter4v macro ? Have I misunderstood 14496-2 standard? I was reference ISO/IEC 14496-2 Second edition 2001-12-01.And I find xvid write dquant and decode dquant as above. Can somebody of mpeg4 part 2 expert answer my question? thanks so much xiongshu deng -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050331/922b1d42/attachment.html From tma iis.fhg.de Thu Mar 31 20:15:43 2005 From: tma iis.fhg.de (Herbert Thoma) Date: Fri Apr 1 17:05:48 2005 Subject: [Mp4-tech] [video] question on mpeg4 part 2 dquant In-Reply-To: <002201c53408$8ede6750$1a040180@dengxiongshu> References: <002201c53408$8ede6750$1a040180@dengxiongshu> Message-ID: <424C3E4F.60406@iis.fhg.de> ??? wrote: > > > dear expert: > > I find only When macro type is inter+q and intra+q can write dquant > value in mpeg4 part 2 standard,so when macro type is inter4v can't write > dquant value because have not inter4v+q macro mode in ISO/IEC > 14496-2,that is to say when macro have four motion vector ,it can not > change quant value in macro level. So I was question how can implement > macro level rate control when a frame have a lot inter4v macro ? Have I > misunderstood 14496-2 standard? I was reference ISO/IEC 14496-2 Second > edition 2001-12-01.And I find xvid write dquant and decode dquant as > above. Can somebody of mpeg4 part 2 expert answer my question? You are absolutely right in your understanding of the standard. It is _not_ possible to have dquant and 4v in one macroblock. If you need to change the quantizer in a particular macroblock you can not use 4 vector mode in this macroblock. Kind regards, Herbert. > thanks so much > > xiongshu deng > > > ------------------------------------------------------------------------ > > _______________________________________________ > NOTE: Please use clear subject lines for your posts. Include [audio, [video], [systems], [general] or another apppropriate identifier to indicate the type of question you have. > > Note: Conduct on the mailing list is subject to the Antitrust guidelines found at http://www.mpegif.org/public/documents/vault/mp-out-30042-Antitrust.php -- Herbert Thoma Group Manager Video Multimedia Realtime Systems Department Fraunhofer IIS Am Wolfsmantel 33, 91058 Erlangen, Germany Phone: +49-9131-776-323 Fax: +49-9131-776-399 email: tma@iis.fhg.de www: http://www.iis.fhg.de/ From mpeg2_user yahoo.com Thu Mar 31 19:29:10 2005 From: mpeg2_user yahoo.com (Tom P) Date: Fri Apr 1 17:05:54 2005 Subject: [Mp4-tech] Performance numbers(MIPS) for H.264 decoder on TMS320C55x Message-ID: <20050401032910.36780.qmail@web54701.mail.yahoo.com> Hi All, Does there any H.264 decoder implementation exists on TMS320C55x ? If so, please let me know the details like MIPS etc. Thanks in advance, __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20050331/505722b7/attachment.html