From biswas tataelxsi.co.in Tue Oct 1 14:08:18 2002 From: biswas tataelxsi.co.in (Biswajit Biswas) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] Dynamic Range Control with MP3 stream References: <3D8AF855.99F79643@tataelxsi.co.in> Message-ID: <3D9950EA.553FD501@tataelxsi.co.in> I am putting across below query again. Has anyone implemented DRC with Mp3 encoder ?? rgds/ Bis Biswajit Biswas wrote: > > Hi All, > > Can anyone suggest how to implement Dynamic Range Control with Mp3? > MPEG-1 layer 3 audio encoder doesn't have any implementation for DRC. > But DRC information can be generated at the encoder side and muxed with > mp3 stream and later at the receiving side this can be demuxed and DRC > can be applied in the decoder. But this would be a propriety > implementation. All I am asking is , has anyone done this with mp3 > streams and if yes where can I get more info on the same. > > With regards > Biswajit > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes From guraaf yahoo.co.in Tue Oct 1 21:16:38 2002 From: guraaf yahoo.co.in (Gaurav Aggarwal) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] max of 4 objects in SP/ASP Message-ID: <000701c26959$58f014d0$6440840a@blr.broadcom.com> Hi all, Need to understand the max of 4 objects in SP/ASP. If I want to build a decoder, then should I have the capability to decode 4 objects (coming on different PIDs, I suppose) in real time and then compose them?? Isn't that a lot of decode bandwidth expected? Or is it enough to decode a single object? Thanks in advance, Gaurav ________________________________________________________________________ Want to sell your car? advertise on Yahoo Autos Classifieds. It's Free!! visit http://in.autos.yahoo.com From rkoenen intertrust.com Tue Oct 1 09:10:18 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] max of 4 objects in SP/ASP Message-ID: <3C124172E7FDD511B510000347426D5901324BDE@exchange.epr.com> The total size of the objects is always the same, whether you have one large object or a max of 4 smaller ones. The encoder/author can choose how to distribute resources (Macroblocks) over the different objects. So decode bandwidth is not affected. Rob > -----Original Message----- > From: Gaurav Aggarwal [mailto:guraaf@yahoo.co.in] > Sent: Tuesday, October 01, 2023 7:47 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] max of 4 objects in SP/ASP > > > Hi all, > > Need to understand the max of 4 objects in SP/ASP. If I want > to build a decoder, then should I have the capability to > decode 4 objects (coming on different PIDs, I suppose) in > real time and then compose them?? Isn't that a lot of decode > bandwidth expected? Or is it enough to decode a single object? > > Thanks in advance, > Gaurav > > ______________________________________________________________ > __________ > Want to sell your car? advertise on Yahoo Autos Classifieds. > It's Free!! > visit http://in.autos.yahoo.com > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From amit.klir emblaze.com Tue Oct 1 20:36:49 2002 From: amit.klir emblaze.com (Amit Klir) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] question about the value of fixed_vop_time_increment Message-ID: <980C7B879697C7448A54C9B065AAFAAC41FFBE@embexcil01.emblaze.ent> Hi, We have a question regarding the number of bits for the representation of VOP_time_increment_resolution where its value is 1. In the case of vop_time_increment_resolution == 1 and fixed_vop_rate == 1 the sending of the bit for the fixed_vop_time_increment is redundant. the only possible values of the vop_time_increment and fixed_vop_time_increment are in the range of [0,VOP_time_increment_resolution).In this case the value is 0. If we have the value of the VOP_time_increment_resolution == n , we need n-1 bits for the vop_time_increment representation. In this special case the number of bits for the fixed_vop_time_increment is 0. According the standard, what is the actual number of bits that need to specify the case of VOP_time_increment_resolution=1 0 - since 0 is the minimum between [0 1) ? or 1- since the value 0 is forbidden ? Thanks in advance Leonid & Amit ************************************************************************************************** The contents of this email and any attachments are confidential. It is intended for the named recipient(s) only. If you have received this email in error please notify the system manager or the sender immediately and do not disclose the contents to any one or make copies. ** eSafe scanned this email for viruses, vandals and malicious content ** ************************************************************************************************** From tomotohara yahoo.com Tue Oct 1 10:59:16 2002 From: tomotohara yahoo.com (Tomo) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] Quick question: video binary shape coder software (routine) Message-ID: <20021001165916.67559.qmail@web10508.mail.yahoo.com> Hi. I have been looking for publicly available binary shape coder (and decoder) routine in C or C++, hopefully available of source code. Does anybody know about publicly available shape coder? Thanks ===== Tomo __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com From rkoenen intertrust.com Tue Oct 1 11:11:06 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] Quick question: video binary shape coder sof tware (routine) Message-ID: <3C124172E7FDD511B510000347426D5901324BED@exchange.epr.com> check out the reference SW you can download from ISO. See m4if's resources page for links. Rob > -----Original Message----- > From: Tomo [mailto:tomotohara@yahoo.com] > Sent: Tuesday, October 01, 2023 9:59 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Quick question: video binary shape > coder software (routine) > > > Hi. > > I have been looking for publicly available binary > shape coder (and decoder) routine in C or C++, hopefully > available of source code. > Does anybody know about publicly available shape coder? > > Thanks > > > ===== > Tomo > > __________________________________________________ > Do you Yahoo!? > New DSL Internet Access from SBC & Yahoo! > http://sbc.yahoo.com _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From chl math.uni-bonn.de Wed Oct 2 11:31:58 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] question about the value of fixed_vop_time_increment In-Reply-To: <980C7B879697C7448A54C9B065AAFAAC41FFBE@embexcil01.emblaze.ent> Message-ID: On Tue, 1 Oct 2002, Amit Klir wrote: > If we have the value of the VOP_time_increment_resolution == n , we need n-1 bits for the vop_time_increment representation. > In this special case the number of bits for the fixed_vop_time_increment is 0. > > According the standard, what is the actual number of bits that need to specify the case of VOP_time_increment_resolution=1 > 0 - since 0 is the minimum between [0 1) ? > or > 1- since the value 0 is forbidden ? Since the value 0 is _explicitly forbidden_, the value has to be 1, and a 1-bit value vop-time_increment=0 has to be written to the bitstream for every VOP, even though this value is redundant. The other criterion of vop_time_increment lying in the interval [0,vop_time_increment_resolution) is _impossible_ to fulfill in this case, so I guess this special case was simply overlooked when writing about the interval. Christoph Lampert -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Zi. 14, 53115 Bonn, Germany | nell erstellt und bedarf Tel. (0228) 73-2948 Fax. +49 228 73-7916 | keiner Unterschrift. AZ 27B-6 From chl math.uni-bonn.de Wed Oct 2 17:07:57 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] A different question about vop_time_increment: stuffing? Message-ID: Hi, since I thought about vop_time_increment for a while I have a question to the experts, too. I found the part 7.13.1 about bit stuffing to avoid generating accidential vop-start-codes # 7.13.1 Start codes and bit stuffing A start code is a two-byte code of # the form 0000 0000 00xx xxxx . Several such codes are inserted into the # bitstream for synchronisation purposes. To prevent any wrongful # synchronisation, such codes shall not be emulated by the data. This is # guaranteed by the insertion of a stuffing bit 1 after each # byte-aligned sequence of eight 0 s. The decoder shall skip these # stuffing bits when parsing the bit stream. Now when I think about timecodes, this happens all the time: Take an ordinary vop_time_increment_resolution 60000 (like for 59.94 Hz NTSC). Then first vop_time_increment==0 would have to be a encoded as 16 zero bits. So there will surely be a sequence of 8 zero bits at a byte aligned address. Would a compliant bitstream have to set a 1 bit at the correct position _within_ the numerical value, thus saving e.g. 000|00000000|100000 instead of 000|00000000|00000 where | indicates the byte boundary? If yes, the decoder has to detect if this 1 is a _real_ 1 or just an "extra 1" and decide whether to skip it by counting zero bits since the last byte boundary... I don't remember seing this in MoMuSys reference software. Christoph Lampert -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Zi. 14, 53115 Bonn, Germany | nell erstellt und bedarf Tel. (0228) 73-2948 Fax. +49 228 73-7916 | keiner Unterschrift. AZ 27B-6 From khuber sorenson.com Wed Oct 2 11:15:45 2002 From: khuber sorenson.com (Kris Huber) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] A different question about vop_time_increme nt: stuffing? Message-ID: <70A238C106788B49A1B7B46C050DEDFE9BD2F9@pandora.sorenson.com> Hi Chris, 7.13.1 refers to 3D mesh coding. That paragraph doesn't apply to the start codes in your example (unless maybe vop_time_increment_resolution is part of the 3D mesh coding syntax, in which case a clarification one way or the other might be appropriate). Regards, Kris Huber -----Original Message----- Date: Wed, 2 Oct 2023 16:07:57 +0200 (CEST) From: Christoph Lampert To: technotes@lists.m4if.org Subject: [M4IF Technotes] A different question about vop_time_increment: stuffing? Hi, since I thought about vop_time_increment for a while I have a question to the experts, too. I found the part 7.13.1 about bit stuffing to avoid generating accidential vop-start-codes # 7.13.1 Start codes and bit stuffing A start code is a two-byte code of # the form 0000 0000 00xx xxxx . Several such codes are inserted into the # bitstream for synchronisation purposes. To prevent any wrongful # synchronisation, such codes shall not be emulated by the data. This is # guaranteed by the insertion of a stuffing bit 1 after each # byte-aligned sequence of eight 0 s. The decoder shall skip these # stuffing bits when parsing the bit stream. Now when I think about timecodes, this happens all the time: Take an ordinary vop_time_increment_resolution 60000 (like for 59.94 Hz NTSC). Then first vop_time_increment==0 would have to be a encoded as 16 zero bits. So there will surely be a sequence of 8 zero bits at a byte aligned address. Would a compliant bitstream have to set a 1 bit at the correct position _within_ the numerical value, thus saving e.g. 000|00000000|100000 instead of 000|00000000|00000 where | indicates the byte boundary? If yes, the decoder has to detect if this 1 is a _real_ 1 or just an "extra 1" and decide whether to skip it by counting zero bits since the last byte boundary... I don't remember seing this in MoMuSys reference software. Christoph Lampert -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Zi. 14, 53115 Bonn, Germany | nell erstellt und bedarf Tel. (0228) 73-2948 Fax. +49 228 73-7916 | keiner Unterschrift. AZ 27B-6 From khuber sorenson.com Wed Oct 2 11:40:23 2002 From: khuber sorenson.com (Kris Huber) Date: Wed Jul 30 14:08:56 2003 Subject: [M4IF Technotes] question about the value of fixed_vop_time_i ncrement Message-ID: <70A238C106788B49A1B7B46C050DEDFE9BD2FA@pandora.sorenson.com> Hi Leonid and Amit, This problem in the text is corrected in an upcoming MPEG-4 corrigenda item under ballot. Please inform the list if this text does not match the behavior of both reference software versions dated 12/2000 and beyond, which is the intent in this particular case (in earlier reference software, Microsoft used 0 bits and MoMuSys used 1 bit for vop_time_increment in the case of vop_time_increment_resolution=1). vop_time_increment: This value represents the absolute vop_time_increment from the synchronization point marked by the modulo_time_base measured in the number of clock ticks. It can take a value in the range of [0,vop_time_increment_resolution). The number of bits representing the value is calculated as the minimum number of unsigned integer bits required to represent the above range, unless vop_time_increment_resolution is equal to '1', in which case one bit with a value of '0' is used. The local time base in the units of seconds is recovered by dividing this value by the vop_time_increment_resolution. modulo_time_base: This value represents the local time base in one second resolution units (1000 milliseconds). It consists of a number of consecutive '1' followed by a '0'. Each '1' represents a duration of one second that have elapsed. For I-, S(GMC)-, and P-VOPs of a non scalable bitstream and the base layer of a scalable bitstream, the number of '1's indicate the number of seconds elapsed since the synchronization point marked by time_code of the previous GOV header or by non-zero modulo_time_base of the previously decoded I-, S(GMC)-, or P-VOP, in decoding order. For B-VOP of non scalable bitstream and base layer of scalable bitstream, the number of '1's indicate the number of seconds elapsed since the synchronization point marked in the previous GOV header, I-VOP, S(GMC)-VOP, or P-VOP, in display order. For I-, P-, or B-VOPs of enhancement layer of scalable bitstream, the number of '1's indicate the number of seconds elapsed since the synchronization point marked in the previous GOV header, I-VOP, P-VOP, or B-VOP, in display order. I think this matches what Chris Lampert indicated as well. Regards, Kris Huber -----Original Message----- Date: Tue, 1 Oct 2023 19:36:49 +0300 From: "Amit Klir" To: Cc: "Leonid Kolotigin" Subject: [M4IF Technotes] question about the value of fixed_vop_time_increment Hi, We have a question regarding the number of bits for the representation of VOP_time_increment_resolution where its value is 1. In the case of vop_time_increment_resolution == 1 and fixed_vop_rate == 1 the sending of the bit for the fixed_vop_time_increment is redundant. the only possible values of the vop_time_increment and fixed_vop_time_increment are in the range of [0,VOP_time_increment_resolution).In this case the value is 0. If we have the value of the VOP_time_increment_resolution == n , we need n-1 bits for the vop_time_increment representation. In this special case the number of bits for the fixed_vop_time_increment is 0. According the standard, what is the actual number of bits that need to specify the case of VOP_time_increment_resolution=1 0 - since 0 is the minimum between [0 1) ? or 1- since the value 0 is forbidden ? Thanks in advance Leonid & Amit **************************************************************************** ********************** The contents of this email and any attachments are confidential. It is intended for the named recipient(s) only. If you have received this email in error please notify the system manager or the sender immediately and do not disclose the contents to any one or make copies. ** eSafe scanned this email for viruses, vandals and malicious content ** **************************************************************************** ********************** From Giuliano.Catrambone h3g.it Wed Oct 2 20:11:40 2002 From: Giuliano.Catrambone h3g.it (Catrambone Giuliano) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] 'config' parameter in the case of MP4V-ES payload name Message-ID: <5A830F2A9C8FD144B6031E3D5D7EA2A74440F6@MIEXC01.h3g.it> Hi, I have a question regarding the 'config' field in the 'fmtp' parameter (within the SDP) in the case of MP4V-ES payload name. 1. How to extract the 'profile-level-id' field from this 'config' field? 2. Which is the syntax of this 'config' parameter? rgds giuliano -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021002/97efe69a/attachment.html From guraaf yahoo.co.in Thu Oct 3 14:56:19 2002 From: guraaf yahoo.co.in (Gaurav Aggarwal) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] max of 4 objects in SP/ASP In-Reply-To: <3C124172E7FDD511B510000347426D5901324BDE@exchange.epr.com> Message-ID: <000c01c26ab6$8cb4da90$6440840a@blr.broadcom.com> Thanks a lot for replying. I should have realized on my own :( Anyhow, what to do with the 4 objects? We would need some kind of compositor, right? But this doesn't seem to part of the spec. So are we open to the method used to display 1 or more out of the 4 objects that may come in the stream. Gaurav > -----Original Message----- > From: Rob Koenen [mailto:rkoenen@intertrust.com] > Sent: Tuesday, October 01, 2023 8:40 PM > To: 'Gaurav Aggarwal'; 'technotes@lists.m4if.org' > Subject: RE: [M4IF Technotes] max of 4 objects in SP/ASP > > > The total size of the objects is always the same, whether > you have one large object or a max of 4 smaller ones. > The encoder/author can choose how to distribute resources > (Macroblocks) over the different objects. > > So decode bandwidth is not affected. > > Rob > > > > -----Original Message----- > > From: Gaurav Aggarwal [mailto:guraaf@yahoo.co.in] > > Sent: Tuesday, October 01, 2023 7:47 > > To: technotes@lists.m4if.org > > Subject: [M4IF Technotes] max of 4 objects in SP/ASP > > > > > > Hi all, > > > > Need to understand the max of 4 objects in SP/ASP. If I want > > to build a decoder, then should I have the capability to > > decode 4 objects (coming on different PIDs, I suppose) in > > real time and then compose them?? Isn't that a lot of decode > > bandwidth expected? Or is it enough to decode a single object? > > > > Thanks in advance, > > Gaurav ________________________________________________________________________ Missed your favourite TV serial last night? Try the new, Yahoo! TV. visit http://in.tv.yahoo.com From jeansonh admtek.com.tw Thu Oct 3 19:47:06 2002 From: jeansonh admtek.com.tw (jeansonh@admtek.com.tw) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] b-vop with quarter pel Message-ID: Hi, all I got one question about b-vops with quarter-pel, and wish somebody may give me a hint. :) In the reference code, if B-VOPs with Backward, Forward, or Bi-directional prediction, the quarter-pel interpolation will be done on 16x16 pixels, while with Direct mode prediction it will be on 4 8x8 pixel blocks. Is this correct? I think in standard section 7.6.9.5.3, it remarks that the result will be totally different on those two different interpolation processes. But when decoding B-VOPs with quarterpel bitstreams generated by DivX 5, 16x16 interpolation will result in blocky images, while 8x8 will 'look' fine... Is this the problem of DivX or it changed in later amendant?My standard version is 14496-2:1999/Amd.1:2000 Thanks in advance! -- Jeanson Hung BU2, ADMtek Inc. Phone: +886-3-5788879-521 From vmorais inescporto.pt Thu Oct 3 13:46:47 2002 From: vmorais inescporto.pt (Vagner Morais) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] MOMUSYS configuration Message-ID: <000c01c26b15$9c4db9e0$611875c2@gauss> Hi, Where can i get the manual of MOMUSYS, to do the configuration of the files .ctl and .cfg. Thanks by your attention. *************************************************************************************************** Vagner David Pinto Morais INESC - Porto | Instituto de Engenharia de Sistemas e Computadores do Porto Unidade de Telecomunica??es e Multim?dia Grupo de An?lise e S?ntese de Imagem Email : vmorais@inescporto.pt Tel. : +351 222094233 ****************************************************************************************************** _________ http://www.inescn.pt __| \ / | Campus da FEUP ___| /\ ___/\ | Rua Dr. Roberto Frias, n? 378 ___ |/_______\| 4200-465 PORTO ------------------------------------------------------------------------------ ******************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021003/e2637178/attachment.html From rkoenen intertrust.com Thu Oct 3 10:26:59 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] max of 4 objects in SP/ASP Message-ID: <3C124172E7FDD511B510000347426D5901324C42@exchange.epr.com> MPEG-4 Systems would give you the ability to composite the objects. MPEG-4 Visual and MPEG-4 Systems have been defined in a coherent fashion. This reflects the object-based nature of MPEG-4. As of today, this isn't used in many applications, although on a PC you could do cool things with the different objects (and you are also free to define their aspect ratios given the MB size restrictions). Without MPEG-4 Systems, a stream would normally contain one single object. Rob > -----Original Message----- > From: Gaurav Aggarwal [mailto:guraaf@yahoo.co.in] > Sent: Thursday, October 03, 2023 1:26 > To: technotes@lists.m4if.org > Subject: RE: [M4IF Technotes] max of 4 objects in SP/ASP > > > Thanks a lot for replying. I should have realized on my own :( > > Anyhow, what to do with the 4 objects? We would need some > kind of compositor, right? But this doesn't seem to part of > the spec. So are we open to the method used to display 1 or more > out of the 4 objects that may come in the stream. > > Gaurav > > > -----Original Message----- > > From: Rob Koenen [mailto:rkoenen@intertrust.com] > > Sent: Tuesday, October 01, 2023 8:40 PM > > To: 'Gaurav Aggarwal'; 'technotes@lists.m4if.org' > > Subject: RE: [M4IF Technotes] max of 4 objects in SP/ASP > > > > > > The total size of the objects is always the same, whether > > you have one large object or a max of 4 smaller ones. > > The encoder/author can choose how to distribute resources > > (Macroblocks) over the different objects. > > > > So decode bandwidth is not affected. > > > > Rob > > > > > > > -----Original Message----- > > > From: Gaurav Aggarwal [mailto:guraaf@yahoo.co.in] > > > Sent: Tuesday, October 01, 2023 7:47 > > > To: technotes@lists.m4if.org > > > Subject: [M4IF Technotes] max of 4 objects in SP/ASP > > > > > > > > > Hi all, > > > > > > Need to understand the max of 4 objects in SP/ASP. If I want to > > > build a decoder, then should I have the capability to decode 4 > > > objects (coming on different PIDs, I suppose) in real > time and then > > > compose them?? Isn't that a lot of decode bandwidth > expected? Or is > > > it enough to decode a single object? > > > > > > Thanks in advance, > > > Gaurav > > ______________________________________________________________ > __________ > Missed your favourite TV serial last night? Try the new, Yahoo! TV. > visit http://in.tv.yahoo.com > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From rkoenen intertrust.com Thu Oct 3 13:04:16 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] object based coding, profiles and bandwidth limitation Message-ID: <3C124172E7FDD511B510000347426D5901324C71@exchange.epr.com> Sheldon, shape coding is not supported in Simple because Simple is kept just that way - simple. It was not deemed necessary and Simple is the lowest complexity Profile in MPEG-4 Visual. Low bitrate apps usually require low complexity as well. The jury is still out on whether object-based coding helps compression efficiency, but for those who want to try and prove, there is Core to play with. A few companies actually implemented this, but not all are around anymore. It may come back, who knows. I have seen promising results. As for bandwidth limitations: of course there are. Otherwise you'd not be able to implement a decoder and know it would execute in real time. See "Information on MPEG-4 Object Types, Profiles and Levels" on this page: http://www.m4if.org/resources.php Rob -----Original Message----- From: X.D.Cai@sussex.ac.uk [mailto:X.D.Cai@sussex.ac.uk] Sent: Tuesday, September 24, 2023 4:35 To: technotes@lists.m4if.org Subject: [M4IF Technotes] object based coding, profiles and bandwidth limitation Dear all : It is well known that mpeg-4 is object-based coding which supports arbitrary shape coding in its core profile. Does anybody can tell me why it is not supported in simple profile? I believe in many cases, the arbitrary shape object based coding can provide a better compression rate than rectangle shape or frame coding so that it can serve low bandwidth applications (in simple profile). So is it possible to use arbitrary shape coding technique (by video segmentation) for simple profile ? Also , are there anyone can tell me if there are bandwidth limitation designed in the standard for different profiles ? if so , why ? Many thanks . sheldon University of Sussex UK -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021003/e4cb925b/attachment.html From yangwj lucent.com Fri Oct 4 10:34:43 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] MOMUSYS configuration Message-ID: <6999EF22C12CD4119A6800508B65021F0244E1BD@CI0027EXCH001U> hello, I think you can refer to README under root directory for the .cal and .cog files. Yang, Wei Jian (Mac) Bell labs, Shanghai Lucent Technologies (China) Co., Ltd. Tel: 8621-5450-4555 ext. 8610 Fax: 8621-5490-2106 E-mail: yangwj@lucent.com -----Original Message----- From: Vagner Morais [mailto:vmorais@inescporto.pt] Sent: 2002?10?4? 3:47 To: MPEG4 Duvidas Subject: [M4IF Technotes] MOMUSYS configuration Hi, Where can i get the manual of MOMUSYS, to do the configuration of the files .ctl and .cfg. Thanks by your attention. **************************************************************************** *********************** Vagner David Pinto Morais INESC - Porto | Instituto de Engenharia de Sistemas e Computadores do Porto Unidade de Telecomunica??es e Multim?dia Grupo de An?lise e S?ntese de Imagem Email : vmorais@inescporto.pt Tel. : +351 222094233 **************************************************************************** ************************** _________ http://www.inescn.pt __| \ / | Campus da FEUP ___| /\ ___/\ | Rua Dr. Roberto Frias, n? 378 ___ |/_______\| 4200-465 PORTO ---------------------------------------------------------------------------- -- **************************************************************************** **** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021004/ef60ff36/attachment.html From mylanchen yahoo.com Fri Oct 4 02:09:57 2002 From: mylanchen yahoo.com (xin chen) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] generation of mp4 files Message-ID: <20021004080957.56751.qmail@web12104.mail.yahoo.com> Hi all, I've some questions during generation a mp4 file: 1.Video Encoder: when player of mpeg4ip (gmp4player) is used to decode the mp4 file generated by QT Pro, it reports to use xvid plug-in to decode video, while mpeg4ip itself supplies two kinds of encoder: ISO encoder(mp4venc) and xvid encoder(xvidenc). So xvid is different from ISO encoder, but it's ISO compliant? Who can give me a detailed explanation of QT Pro Encoder and its compliance with ISO? 2. Audio Encoder: I tried different ways to encode a wav file to AAC format, under windows .aac is generated but it is not recognized by QT, and when .aac is packaged to .mp4 by mpeg4ip, cannot decoded by QT either. Under Linux, I use the FAAC from mpeg4ip, it fails and says: faac -pLOW -b128 true.wav Without libsndfile, only raw format is allowed What does it mean? I think the wav file is correct because it can be encoded into mp3 by lame of mpeg4ip. Thank you very much. __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com From idimkovic nero.com Fri Oct 4 15:24:45 2002 From: idimkovic nero.com (Ivan Dimkovic) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] generation of mp4 files In-Reply-To: <20021004080957.56751.qmail@web12104.mail.yahoo.com> Message-ID: Dear Xin, All 1. (Actually, guys from MPEG4IP project would probably give more precise answer): MPEG4IP has several decoders - Xvid decoder used for Simple profile, and ISO reference decoder (Microsoft) for all other profiles, because (to my best knowledge) Xvid decoder does not support other profiles than simple (and maybe advanced simple). Xvid decoder is used for simple profile because it is much faster than unoptimized ISO reference source code. Xvid encoder should generate ISO compatible elementary bitstreams (simple profile, or in SigmaDesign's derrivate even Advanced Simple). To my best knowledge, QT Pro MPEG-4 video codec is ISO compliant. 2. QT6 won't recognize .aac (elementary stream audio) file, it can only read MP4 file format - so you must multiplex MPEG-4 Low Complexity AAC file into .mp4. As far as I know, FAAC generates ISO/MPEG-4 compliant files - make sure that you are using "MPEG-4 AAC" and not "MPEG-2 AAC" in FAAC, because some decoders won't play MP4 files with 'MPEG-2 AAC' object type inside. Best regards / Mit freundlichen Gru?en Ivan Dimkovic ---------------------------------------------------------------- Ivan Dimkovic MPEG-4 Software R&D Ahead Software AG phone: +49 (0)7248 911 822(direct line) Im Stoeckmaedle 18 fax: +49 (0)7248 911 888 76307 Karlsbad email: idimkovic@nero.com Germany web: www.nero.com ---------------------------------------------------------------- -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of xin chen Sent: Friday, October 04, 2023 10:10 AM To: technotes@lists.m4if.org Subject: [M4IF Technotes] generation of mp4 files Hi all, I've some questions during generation a mp4 file: 1.Video Encoder: when player of mpeg4ip (gmp4player) is used to decode the mp4 file generated by QT Pro, it reports to use xvid plug-in to decode video, while mpeg4ip itself supplies two kinds of encoder: ISO encoder(mp4venc) and xvid encoder(xvidenc). So xvid is different from ISO encoder, but it's ISO compliant? Who can give me a detailed explanation of QT Pro Encoder and its compliance with ISO? 2. Audio Encoder: I tried different ways to encode a wav file to AAC format, under windows .aac is generated but it is not recognized by QT, and when .aac is packaged to .mp4 by mpeg4ip, cannot decoded by QT either. Under Linux, I use the FAAC from mpeg4ip, it fails and says: faac -pLOW -b128 true.wav Without libsndfile, only raw format is allowed What does it mean? I think the wav file is correct because it can be encoded into mp3 by lame of mpeg4ip. Thank you very much. __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From neel2265 yahoo.com Fri Oct 4 07:31:19 2002 From: neel2265 yahoo.com (Neelakanth) Date: Wed Jul 30 14:08:57 2003 Subject: [M4IF Technotes] 'config' parameter in the case of MP4V-ES payload name In-Reply-To: <5A830F2A9C8FD144B6031E3D5D7EA2A74440F6@MIEXC01.h3g.it> Message-ID: <20021004133119.60742.qmail@web40202.mail.yahoo.com> > Hi, > I have a question regarding the 'config' field > in the 'fmtp' parameter (within the SDP) > in the case of MP4V-ES payload name. > > 1. How to extract the 'profile-level-id' field > from this 'config' field? > 2. Which is the syntax of this 'config' > parameter? syntax for config is DecoderSpecificInfo as mentioned in 8.6.7 of 14496-1 and 'profile-level-id' field can be parsed from config field, since it is mentioned Profile-level-id: For visual streams, this parameter is the decimal value from Table G-1 (FLC table for profile and level indication of ISO/IEC 14496- 2,indicating which MPEG-4 Visual tool subsets are applied to encode the visual stream. profile-level-id is same as profile_and_level_Indication found at VisualObjectSequence()in 6.2.2 of 14496-2. This is what I understand. If I am wrong please correct me. One question to u all For audio streams, 'profile-level-id' parameter is the decimal value from Table 5 audioProfileLevelIndicationValues) in ISO/IEC 14496-1, indicating which MPEG-4 Audio tool subsets are applied to encode the audio stream. For Visual streams, why it can't be visualProfileLevelIndication Values of Table 6 in ISO/IEC 14496-1 ? Is it correct to parse the visual stream? Thanks and Regards Neelakanth __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com From singer apple.com Fri Oct 4 10:16:58 2002 From: singer apple.com (Dave Singer) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] generation of mp4 files In-Reply-To: <20021004080957.56751.qmail@web12104.mail.yahoo.com> References: <20021004080957.56751.qmail@web12104.mail.yahoo.com> Message-ID: At 01:09 -0700 10/4/02, xin chen wrote: >Hi all, >I've some questions during generation a mp4 file: > >1.Video Encoder: when player of mpeg4ip (gmp4player) >is used to decode the mp4 file generated by QT Pro, it >reports to use xvid plug-in to decode video, while >mpeg4ip itself supplies two kinds of encoder: ISO >encoder(mp4venc) and xvid encoder(xvidenc). So xvid is >different from ISO encoder, but it's ISO compliant? >Who can give me a detailed explanation of QT Pro >Encoder and its compliance with ISO? I'd be happy to attempt to answer any questions you have. > >2. Audio Encoder: I tried different ways to encode a >wav file to AAC format, under windows .aac is >generated but it is not recognized by QT, and when >.aac is packaged to .mp4 by mpeg4ip, cannot decoded by >QT either. Under Linux, I use the FAAC from mpeg4ip, >it fails and says: >faac -pLOW -b128 true.wav >Without libsndfile, only raw format is allowed >What does it mean? I think the wav file is correct >because it can be encoded into mp3 by lame of mpeg4ip. > QuickTime handles only MP4 files with MPEG-4 AAC, as Ivan noted. Does his answer solve your question? I'd be happy to look at the possibility of our WAV importer handling files with MPEG-4 AAC embedded in them, also, if needed. I need to know where that is specified, and have an example file. -- David Singer Apple Computer/QuickTime From wmay cisco.com Fri Oct 4 10:42:17 2002 From: wmay cisco.com (Bill May) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] generation of mp4 files References: Message-ID: <3D9DC4E9.3060305@cisco.com> Ivan Dimkovic wrote: > 1. (Actually, guys from MPEG4IP project would probably give more precise > answer): MPEG4IP has several decoders - Xvid decoder used for Simple > profile, and ISO reference decoder (Microsoft) for all other profiles, > because (to my best knowledge) Xvid decoder does not support other profiles > than simple (and maybe advanced simple). Xvid decoder is used for simple > profile because it is much faster than unoptimized ISO reference source > code. > > Xvid encoder should generate ISO compatible elementary bitstreams (simple > profile, or in SigmaDesign's derrivate even Advanced Simple). That being me, the version of we use xvid supports a subset of simple profile - it is missing some tools such as video packets. We use the ISO reference code (1999 version, I believe) for all others. Bill May From rkoenen intertrust.com Sun Oct 6 10:24:45 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] FW: Quarter-pel AMV problem - what have deployed Advanced Simple code cs done? Message-ID: <3C124172E7FDD511B510000347426D5901324CB7@exchange.epr.com> Forward on behalf of Kris Huber who had trouble posting - hope this does work. Rob ---------------------------------------------------------------------------- - Hello all, I recently discovered what I consider an error in probably all versions of the MPEG-4 visual reference software to date that implement quarter-pel motion compensation. The problem is that the AMV (average motion vector) used for further motion vector coding after a GMC macroblock, is clipped to a range that is half the quarter-pel MV range. I noticed today that the same problem, plus some other related issues, were reported in the most recent "List of MPEG-4 Visual Problem Reports" MPEG document. That item was reporting a problem in the OptSimple code, but my searching found the corresponding location in the Microsoft 12/13/2000 code where incorrect GMC AVM clipping is performed. Note that the title of the item that reports this mentioned interlace, but it also reports several issues. None of those issues are confirmed or resolved in discussions that I've been involved with to date. The dilema is how this should be resolved. To promote highly-efficient interoperable information coding (not to mention causing fewer headaches to future consumers, engineers, and anyone else expecting interoperability to mean that the same bitstream will play on all conformant decoders), I think the wise course is to give the most weight to how the issue was resolved in the implementations currently deployed. This is a technical issue that cannot be avoided in Advanced Simple Profile implementations. Some things to consider are: - To the best of my knowledge the behavior is the same in all current versions of reference software, but does not match the text of the visual spec. - AMV problem will not impact coding efficiency of S(GMC)-VOPs more than a small (negligible) increase in proportion of motion vector prediction bits. - AMV problem impacts B-VOP coding efficiency due to modified temporal prediction in direct mode. Again, since most of probability for motion vectors is typically around a small range, coding efficiency is not compromised by the current reference software. - the current behavior of reference software was used during verification testing. Please respond if you have information about the details in these respects about implementations that are deployed or coming close to that stage. If you prefer to remain anonymous, reply to me directly and I will summarize results to the list without your company name included (or I'm sure Rob Koenen would do you the favor). Gathering such information can be helpful so this ambiguity of specification can be resolved in an way accomodating to all concerned. Without further information, I favor resolving the issue by changing the text through corrigendum to match the reference software. Best regards and thanks, Kris Huber From guraaf yahoo.co.in Mon Oct 7 11:15:55 2002 From: guraaf yahoo.co.in (Gaurav Aggarwal) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] generation of mp4 files In-Reply-To: <3D9DC4E9.3060305@cisco.com> Message-ID: <001701c26dbc$6c6f06d0$6440840a@blr.broadcom.com> Pardon me for jumping in, but what do you mean that support for video packets is missing? Will the entire VOP be a single video packet, i.e., all MBs would come soon after the VOP header and there will be no resync marker and video_packet header? Thanks, Gaurav > That being me, the version of we use xvid supports a subset of > simple profile - > it is missing some tools such as video packets. We use the ISO reference > code (1999 version, I believe) for all others. > > Bill May ________________________________________________________________________ Missed your favourite TV serial last night? Try the new, Yahoo! TV. visit http://in.tv.yahoo.com From d98rolb stud.hh.se Mon Oct 7 14:41:28 2002 From: d98rolb stud.hh.se (Roland Bengtsson) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] Some newbies questions Message-ID: <5.1.0.14.0.20021007114244.02e0ee48@studpop.hh.se> I have some very simple questions about MPEG-4. 1. I have found some source for MPEG-4 on http://megaera.ee.nctu.edu.tw/mpeg Is this the latest free software about MPEG-4? 2. What is the differens betwwen Microsoft-fdaml and MoMuSys-fdaml? Microsoft seems to be towards Windows and MoMuSys against Unix. 3. I downloaded the reference source from MS and compiled it in VC++. It compiled ok and the input from test-sequences worked well. But I imagine that I could have a MPEG-4 movie as input to the decoder. But the decoder has a cmp-file as input and a yuv-file as output. So my question is what is the difference between a cmp-file and a MPEG-4 file? 4. I have an idea to feed the decoder with a MPEG-4 stream containing a base layer and a enhancement layer. Then just handle the base layer for output because low bandwidth. Is there any software that show how this can be done? I'm look forward to the answers! Best regards Roland From mylanchen yahoo.com Sun Oct 6 23:53:20 2002 From: mylanchen yahoo.com (xin chen) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] generation of mp4 files In-Reply-To: Message-ID: <20021007055320.77985.qmail@web12104.mail.yahoo.com> Thank you all for your help. Here is my ideas: about the video of mpeg-4, the encoder of mpeg4ip, both xvid and ISO compliant one, and the encoder of QT Pro, are all ISO compliant, just different algorithms, different profiles. right? about the audio encoder, it's my mistake to not using the -r option in faac command. I tried faac -m4 -pLC -r -b96 *.wav *.aac and it works:) And David, I'm not clear about your idea of WAV importer handling files with MPEG-4 AAC embedded in them, what kind of example do you need? how to embed an AAC file in WAV? --- Dave Singer wrote: > At 01:09 -0700 10/4/02, xin chen wrote: > >Hi all, > >I've some questions during generation a mp4 file: > > > >1.Video Encoder: when player of mpeg4ip > (gmp4player) > >is used to decode the mp4 file generated by QT Pro, > it > >reports to use xvid plug-in to decode video, while > >mpeg4ip itself supplies two kinds of encoder: ISO > >encoder(mp4venc) and xvid encoder(xvidenc). So xvid > is > >different from ISO encoder, but it's ISO compliant? > >Who can give me a detailed explanation of QT Pro > >Encoder and its compliance with ISO? > > I'd be happy to attempt to answer any questions you > have. > > > > >2. Audio Encoder: I tried different ways to encode > a > >wav file to AAC format, under windows .aac is > >generated but it is not recognized by QT, and when > >.aac is packaged to .mp4 by mpeg4ip, cannot decoded > by > >QT either. Under Linux, I use the FAAC from > mpeg4ip, > >it fails and says: > >faac -pLOW -b128 true.wav > >Without libsndfile, only raw format is allowed > >What does it mean? I think the wav file is correct > >because it can be encoded into mp3 by lame of > mpeg4ip. > > > > QuickTime handles only MP4 files with MPEG-4 AAC, as > Ivan noted. > Does his answer solve your question? > > I'd be happy to look at the possibility of our WAV > importer handling > files with MPEG-4 AAC embedded in them, also, if > needed. I need to > know where that is specified, and have an example > file. > -- > David Singer > Apple Computer/QuickTime > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes __________________________________________________ Do you Yahoo!? Faith Hill - Exclusive Performances, Videos & More http://faith.yahoo.com From wmay cisco.com Mon Oct 7 10:30:28 2002 From: wmay cisco.com (Bill May) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] generation of mp4 files References: <001701c26dbc$6c6f06d0$6440840a@blr.broadcom.com> Message-ID: <3DA1B6A4.3040706@cisco.com> In the version that we have, yes. If you have more interest in the xvid code, I would suggest you visit their site; I do not wish to put words in their mouths. Bill Gaurav Aggarwal wrote: > Pardon me for jumping in, but what do you mean that support for > video packets is missing? Will the entire VOP be a single video > packet, i.e., all MBs would come soon after the VOP header and > there will be no resync marker and video_packet header? > > Thanks, > Gaurav > > >>That being me, the version of we use xvid supports a subset of >>simple profile - >>it is missing some tools such as video packets. We use the ISO reference >>code (1999 version, I believe) for all others. >> >>Bill May > > > ________________________________________________________________________ > Missed your favourite TV serial last night? Try the new, Yahoo! TV. > visit http://in.tv.yahoo.com > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From yangwj lucent.com Tue Oct 8 16:54:55 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] video codec DirectX filter Message-ID: <6999EF22C12CD4119A6800508B65021F0244E1C7@CI0027EXCH001U> hello, Does any one know how to write a video codec filter for MS DirectX framework? I can not find sample source code under directshow sub-dir in DirectX8.1 SDK. Thanks Mac From chenchunxi hotmail.com Tue Oct 8 18:36:22 2002 From: chenchunxi hotmail.com (chen chun) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] about RVLC Message-ID: It seems an error in the standard. Am I right? >-----Original Message----- >From: Wu, Gang [mailto:gang.wu@intel.com] >Sent: Monday, September 23, 2023 10:19 AM >To: Technotes@lists.m4if.org >Subject: [M4IF Technotes] about RVLC > > >hi, >would you please help me understand the standard clearly about RVLC, >in the standard, I am confused with the strategies for determining which MB >to been discarded. > in standard(detail see page 422 in standard) > L1 : Number of bits which can be decoded in a forward decoding. > N1 : Number of MBs which can be completely decoded in a forward >decoding. > f_mb(S): Number of decoded MBs which S bits can be decoded in a >forward direction. > (Equal to or more than one bit can be decoded in a MB, >f_mb(S) counter if up) > > what is the relationship between N1 and f_mb(L1) ? > (1)N1 == f_mb(L1). > (2)N1 == f_mb(L1) ro N1 == f_mb(L1) -1. > (3)others. >It's very important to implement the strategy. especially in strategy 3. >maybe N-b_mb(L2)>=N1? am I right? >thanks very much for you help. >Gang > > > >_______________________________________________ >Technotes mailing list >Technotes@lists.m4if.org >http://lists.m4if.org/mailman/listinfo/technotes _________________________________________________________________ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.microsoft.com/cn From rkoenen intertrust.com Tue Oct 8 08:51:48 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] video codec DirectX filter Message-ID: <3C124172E7FDD511B510000347426D5901324CEB@exchange.epr.com> Let's keep this list for MPEG-4 related questions. Rob > -----Original Message----- > From: Yang, Wei Jian (Mac) [mailto:yangwj@lucent.com] > Sent: Tuesday, October 08, 2023 0:55 > To: Technotes (E-mail) > Subject: [M4IF Technotes] video codec DirectX filter > > > hello, > > Does any one know how to write a video codec filter for MS > DirectX framework? I can not find sample source code under > directshow sub-dir in DirectX8.1 SDK. > > Thanks > Mac > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From brail mcst.ru Tue Oct 8 20:13:40 2002 From: brail mcst.ru (Ilya V. Brailovsky) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] MoMuSys with ASP bitstreams ( vcon_asp*.bits ) Message-ID: <3DA2F624.48351162@mcst.ru> Hi all, I've ran MoMuSys-000720 with the new AS profile bitstreams ( vcon_asp*.bits ). Decoder refuses to process any of this set, it says: ERROR(ecalloc): Wrong size: <= 0. Does anyone have the same problem or smth wrong with my MoMuSys decoder ? Thanks, -Ilya From vmorais inescporto.pt Tue Oct 8 18:39:34 2002 From: vmorais inescporto.pt (Vagner Morais) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] MOMUSYS Message-ID: <002d01c26f2c$56b03d70$611875c2@gauss> Good afternoon: I have 4 questions: 1 - When i use the encoder MOMUSYS to encode a sequence .yuv, and i configure .cfg and .ctl to use a mask, is supposed when i decode the file to obtain a video segmented (using binary or grey-level)? Or the segmented video is only in format .bits (mpeg-4)? 2 - What is the "alpha threshold" in .cfg files of MOMUSYS? 3 - Is there any tool (ideally open-source) to make masks of videos? 4 - I tried to encode the same file using different "type of alpha channel" (rectangular and bynary) and the file encoded with alpha channel rectangular had higher size than binary? Shouldn?t be the contrary? Thanks by your attention. *************************************************************************************************** Vagner David Pinto Morais INESC - Porto | Instituto de Engenharia de Sistemas e Computadores do Porto Unidade de Telecomunica??es e Multim?dia Grupo de An?lise e S?ntese de Imagem Email : vmorais@inescporto.pt Tel. : +351 222094233 ****************************************************************************************************** _________ http://www.inescn.pt __| \ / | Campus da FEUP ___| /\ ___/\ | Rua Dr. Roberto Frias, n? 378 ___ |/_______\| 4200-465 PORTO ------------------------------------------------------------------------------ ******************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021008/7e8e0449/attachment.html From krangam lsil.com Tue Oct 8 15:35:03 2002 From: krangam lsil.com (kasturi rangam) Date: Wed Jul 30 14:08:58 2003 Subject: [M4IF Technotes] CABAC - AVC (h264) question References: <0170DDAD0BADFA4CBEC3B55A0748DCCC05B67009@red-msg-02.redmond.corp.microsoft.com> Message-ID: <3DA34F87.3675B2A2@lsil.com> Hi, I have a question on CABAC decode for macroblock type in I slice. I have the following revision of the spec: JVT-D157, generated:2002-08-10 Table 9-20 on page 93 has the binarization for macroblock types for I slices. From the table it looks like by decoding the macroblock type, we get the Imode, nc (chroma cbp) and AC flag. There are 2 issues. 1. In table 9-23, the binarization type specified for ctx_mb_type_I is Table 9-14, which i think should be table 9-20 2. There is additional ctx_cbp_chroma (context from chroma cbp). From what I understand, this applies only for P and B slices, since the mb_type for I slices already has chroma cbp. Please let me know if my assumptions are right. Thanks, Kasturi From seal7100 media.gwu.ac.kr Wed Oct 9 10:43:21 2002 From: seal7100 media.gwu.ac.kr (=?ks_c_5601-1987?B?seix2b+1?=) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Final reference software number for ASP and FGS Message-ID: <003801c26f2c$de150980$bd418680@seal> Hello, I have a question. What is the number of final version of reference software for encoding Advanced simple profile and FGS. please tell me about them. Thanks, Keun-Young Kim -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021009/a1f19729/attachment.html From gaodd soft.thtf.com.cn Wed Oct 9 16:26:42 2002 From: gaodd soft.thtf.com.cn (gaodd) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] where can i find detailed description of [4-MV, Unrestricted MV]? Message-ID: <018201c26f65$3a0325a0$1fd06fa6@gaodd> Hi In Mpeg-4, [4-MV, Unrestricted MV] is a tool used in Simple Object of ASP Profile, i can't find its detailed description in N3056, now i wonder where i could find it? thanks very much! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021009/6f74a148/attachment.html From guraaf yahoo.co.in Wed Oct 9 18:31:59 2002 From: guraaf yahoo.co.in (Gaurav Aggarwal) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Test-suite for MPEG-4 SP/ASP conformance testing Message-ID: <002201c26f8b$abf76650$6440840a@blr.broadcom.com> Hi all, Is anyone aware of a comprehensive conformance test-suite for MPEG-4 SP/ASP? We were using Sarnoff for MPEG-2 and would appreciate if you could provide any pointers. Thanks, Gaurav ________________________________________________________________________ Missed your favourite TV serial last night? Try the new, Yahoo! TV. visit http://in.tv.yahoo.com From moses kampsax.dtu.dk Wed Oct 9 16:41:56 2002 From: moses kampsax.dtu.dk (moses woldeselassie) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] help with momusys in linux (make -e error) Message-ID: <3DA43224.30F63361@kampsax.dtu.dk> dear all I'm using MoMuSys-FPDAM1-1.0-000720 the problem is with these files (when I use gmake and make -e SYSTEM=LINUX) gcc: lib/lib_linux/libvm_enc.a : No such fies or directory gcc: lib/lib_linux/libvm_common.a what are these files *.a? how are connected the MoMusys (MPEG4)? additionally, i'm looking for the generate_doc.pl file. I'm looking forward to get your help. Best regards moses From yangwj lucent.com Thu Oct 10 15:37:47 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Coder efficiency Message-ID: <6999EF22C12CD4119A6800508B65021F03FE2E78@CI0027EXCH001U> Dear MPEG4 experts: When I use Momusys as reference MPEG4 codec, I find that something interesting. This DCT-like coder has much better efficiency than MPEG2 Simulating environment is: 1. MPEG2 coder: 1996, MPEG Software Simulation Group, MP@ML 2. MPEG4 coder: MoMuSys-FDIS-V1.0-990812 simple profile 3. Akiyo.yuv 4. 96000kbps, MPEG2 TM5 Rate Control Compare result is: (bits/PSNR-Y(db)) I P B ...... MPEG2 37740/32 7234/32.6 9979/33.1 MPEG4 47982/37.8 678/37.5 80/37.7 I find that MPEG2 coder produce much more MVs in P and B than those in MPEG4. I guess in MPEG4, it uses a more efficient MB type decision policy in Motion Estimation. I know MPEG4 adopts quarter-pixel prediction, but does it has such high impact on ME efficiency? Any comments? Thanks Mac Yang, Wei Jian (Mac) Bell labs, Shanghai Lucent Technologies (China) Co., Ltd. Tel: 8621-5450-4555 ext. 8610 Fax: 8621-5490-2106 E-mail: yangwj@lucent.com From guraaf yahoo.co.in Thu Oct 10 18:26:03 2002 From: guraaf yahoo.co.in (Gaurav Aggarwal) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] help with momusys in linux (make -e error) In-Reply-To: <3DA43224.30F63361@kampsax.dtu.dk> Message-ID: <002f01c27054$0273c320$6440840a@blr.broadcom.com> Moses, I am no authority on either MPEG-4 or on s/w installation, but I will try to help. I am using MoMuSys-FDIS-V1.0-990812 and not MoMuSys-FPDAM1-1.0-000720 but they should be similar. I have the files in my directory as lib/lib_sol2/*.a since I use solaris. I believe they are the library files for the encoder, decoder and the common ones that the eventual executable binary will use. Mine were created automatically when I compiled the software. Also, I compiled the encoder as well as the decoder. Gaurav > dear all > > I'm using MoMuSys-FPDAM1-1.0-000720 > the problem is with these files (when I use gmake and make -e > SYSTEM=LINUX) > gcc: lib/lib_linux/libvm_enc.a : No such fies or directory > gcc: lib/lib_linux/libvm_common.a > > what are these files *.a? > how are connected the MoMusys (MPEG4)? > > additionally, i'm looking for the generate_doc.pl file. > > > I'm looking forward to get your help. > > Best regards > moses ________________________________________________________________________ Missed your favourite TV serial last night? Try the new, Yahoo! TV. visit http://in.tv.yahoo.com From yangwj lucent.com Fri Oct 11 17:45:28 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Coder efficiency Message-ID: <6999EF22C12CD4119A6800508B65021F03FE2E7A@CI0027EXCH001U> Dear MPEG4 expert: I think I have find the reason why Momusys generate less bits in P and B frames. There is a SAD calculation in mot_est_mb.c: If calculate sad with MB in same position of previous picture sad=SAD_Macroblock(...) - (MB_Nb[currentMB]/2+1); else sad=SAD_Macroblock(...); MB_Nb[...] is number of pixels of MB in shape. If I use rectangle coding, MB_Nb[...] should be 255. So, MB at MV(0,0) has much more previlige than other candidate MBs. My question is, sad should be like the following to make more sense: sad=SAD_Macroblock(...) -(255-MB_Nb[currentMB]); Who is correct? Thanks Mac -----Original Message----- From: Yang, Wei Jian (Mac) [mailto:yangwj@lucent.com] Sent: 2002年10月10日 14:38 To: Technotes (E-mail) Subject: [M4IF Technotes] Coder efficiency Dear MPEG4 experts: When I use Momusys as reference MPEG4 codec, I find that something interesting. This DCT-like coder has much better efficiency than MPEG2 Simulating environment is: 1. MPEG2 coder: 1996, MPEG Software Simulation Group, MP@ML 2. MPEG4 coder: MoMuSys-FDIS-V1.0-990812 simple profile 3. Akiyo.yuv 4. 96000kbps, MPEG2 TM5 Rate Control Compare result is: (bits/PSNR-Y(db)) I P B ...... MPEG2 37740/32 7234/32.6 9979/33.1 MPEG4 47982/37.8 678/37.5 80/37.7 I find that MPEG2 coder produce much more MVs in P and B than those in MPEG4. I guess in MPEG4, it uses a more efficient MB type decision policy in Motion Estimation. I know MPEG4 adopts quarter-pixel prediction, but does it has such high impact on ME efficiency? Any comments? Thanks Mac Yang, Wei Jian (Mac) Bell labs, Shanghai Lucent Technologies (China) Co., Ltd. Tel: 8621-5450-4555 ext. 8610 Fax: 8621-5490-2106 E-mail: yangwj@lucent.com _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From k.ramasamy mmu.edu.my Fri Oct 11 10:37:01 2002 From: k.ramasamy mmu.edu.my (K. Ramasamy) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] video compression & transmission over wireless Message-ID: <002d01c270c6$b2407740$3b1a640a@cyber.mmu.edu.my> Dear sir, Greetings....... I'm planning to do my ph.D in video compression and transmission over wireless.I'm trying with the mpeg4 reference software for video compression. I think it is not handy. Is it possible to do both video compression(coder and decoder) and transmission over wireless channels using some hardware equipments. If so, Could you please inform me details of the manufacturers/dealers of hardware equipments for video(mpeg4) compression and transmission. Thanks in advance With best Regards, K.RAMASAMY LECTURER/FOE MULTIMEDIA UNIVERSITY JALAN MULTIMEDIA,63100 CYBERJAYA SELANGOR D.H MALAYSIA TEL:603-8312 5391 FAX:603 8318 3029 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021011/fc25a26d/attachment.html From rkoenen intertrust.com Fri Oct 11 09:48:11 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Some newbies questions Message-ID: <3C124172E7FDD511B510000347426D5901324D74@exchange.epr.com> > 1. I have found some source for MPEG-4 on > http://megaera.ee.nctu.edu.tw/mpeg > Is this the latest > free software about MPEG-4? The official free software is always at ISO's site. Check out http://www.m4if.org/resources.php for the correct links > 2. What is the differens betwwen Microsoft-fdaml and MoMuSys-fdaml? > Microsoft seems to be towards Windows and MoMuSys against Unix. They are two independent implementations of the same algorithm. Having two implementations to exchange bitstreams has helped MPEG to ensure that the standard is sound and does what it is supposed to do. Rob From myinamdar tataelxsi.co.in Sat Oct 12 20:28:23 2002 From: myinamdar tataelxsi.co.in (Mahesh Y Inamdar) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] (no subject) Message-ID: <001801c271f7$6e066990$b214010a@mahesh> Where shall i get mpeg4 aac multichannel streams other then mpeg site ? i need streams which are having one minute of multichannel audio. Thanks in Advance, Mahesh Inamdar -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021012/d81f18ca/attachment.html From pescatorimarco hotmail.com Mon Oct 14 12:42:27 2002 From: pescatorimarco hotmail.com (pesca) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] mpeg-4 scalability Message-ID: I'm a student of the University of Pavia.I'm doing a work about "Transmission of MPEG-4 video over mobile networks".I'm using the Microsoft Visual Codec,but I can't use the options for temporal/spatial scalability when the rate control(TM5 or MP4) is enabled. Is there anybody who can help me ? Please write me to my e-mail address peska78@tin.it Thak you, Hello. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021014/047d9e9f/attachment.html From dsearles mmc.atmel.com Mon Oct 14 12:21:29 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Latest reference software? Message-ID: <3DAAE0F9.501@mmc.atmel.com> I started at the link indicated below: > ... > The official free software is always at ISO's site. > Check out http://www.m4if.org/resources.php for the > correct links > > > > 2. What is the differens betwwen Microsoft-fdaml and MoMuSys-fdaml? > > Microsoft seems to be towards Windows and MoMuSys against Unix. > > They are two independent implementations of the same algorithm. > Having two implementations to exchange bitstreams has helped MPEG to > ensure that the standard is sound and does what it is supposed to do. > > Rob Specifically, Publicly available MPEG-4 reference software from ISO (ISO/IEC 14496-5) This got me to a list of directories at: www.iso.ch/iso/en/ittf/PubliclyAvailableStandards/14496-5_Compressed_directories/ "Visual", in that list took me to another list of files, Natural.zip" and "Snhc.zip" Natural.zip contains Microsoft and Momusys reference software. The Microsoft software appears to be in a directory called microsoft-vfdis-v10-990812. So where is the "fdaml" version? Secondly, going out on a limb, I noticed that there is another dirctory at ISO: www.iso.ch/iso/en/ittf/PubliclyAvailableStandards/ISO_IEC_14496-5_2001_Software_Reference/ which also contains "reference" software. Peeking into "video.zip" shows "microsoft-2.3-001213", and the code looks updated. Where does this code fit into the scheme of things? Thanks Dan Searles -- Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-0300 From salo rucus.ru.ac.za Mon Oct 14 23:51:33 2002 From: salo rucus.ru.ac.za (Nik) Date: Wed Jul 30 14:08:59 2003 Subject: [M4IF Technotes] Bifs-text scene description - streaming movietexture Message-ID: <001b01c273c3$7b4a0260$9679e792@beast> Hi streaming experts, I'd really appreciate help with this. I need a movietexture that uses live video as its source, such as "rtsp://localhost:554/vid.mp4". How do you specify the OD's and the ESD's for this? Can anyone give me some pointers on this, or even the url to a similar mp4 on the net somewhere? I'm also a little confused as to how a player will know which track to use from a url like "rtsp://localhost:554/vid.mp4" for the movietexture....? Any help very appreciated. Kind regards, Nicholas. From vmorais inescporto.pt Tue Oct 15 13:13:24 2002 From: vmorais inescporto.pt (Vagner Morais) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Masks Video Message-ID: <004101c2747e$eee547a0$611875c2@gauss> Hi, Anyone know if is there any tool (ideally open-source) to make masks of videos? Thanks by your attention *************************************************************************************************** Vagner David Pinto Morais INESC - Porto | Instituto de Engenharia de Sistemas e Computadores do Porto Unidade de Telecomunica??es e Multim?dia Grupo de An?lise e S?ntese de Imagem Email : vmorais@inescporto.pt Tel. : +351 222094233 ****************************************************************************************************** _________ http://www.inescn.pt __| \ / | Campus da FEUP ___| /\ ___/\ | Rua Dr. Roberto Frias, n? 378 ___ |/_______\| 4200-465 PORTO ------------------------------------------------------------------------------ ******************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021015/a94f5ddb/attachment.html From salo rucus.ru.ac.za Tue Oct 15 15:33:16 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Bifs-text scene description - streaming movietexture In-Reply-To: <45A1F95BB9D7D84FAB0A1EB4D67EEF964FB1E3@IS01EX01.ittiam.com> Message-ID: <000001c27447$0aee3ac0$9679e792@beast> > Hi, > > Movie textures are not supported in any streaming standard > (RFC 3016 or ISMA). Neither do I know about any proprietary > way of doing it. You may have to device your own scheme for this. > > Regards, > Parimal. > Hmmm, so then how are applications like (implemented versions of) the following from enst supposed to work? http://161.58.176.70/downloads/content/MIRRORSITE/interoperability/ENST/ ecran2.mp4 Inlineing the sources is fine, until you want to manipulate them in some way, such as full screening them.... Thanks, Nik. > -----Original Message----- > From: Nik [mailto:salo@rucus.ru.ac.za] > Sent: Tuesday, October 15, 2023 2:22 AM > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Bifs-text scene description - > streaming movietexture > > > Hi streaming experts, > > I'd really appreciate help with this. > > I need a movietexture that uses live video as its source, > such as "rtsp://localhost:554/vid.mp4". > > How do you specify the OD's and the ESD's for this? > > Can anyone give me some pointers on this, or even the url to > a similar mp4 on the net somewhere? > > I'm also a little confused as to how a player will know which > track to use from a url like "rtsp://localhost:554/vid.mp4" > for the movietexture....? > > > Any help very appreciated. > > Kind regards, > Nicholas. > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From BEJO de.ibm.com Tue Oct 15 17:29:46 2002 From: BEJO de.ibm.com (Jochen Bergdolt1) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] MPEG-4 with timecode Message-ID: Dear list, i am trying to find out a solution that enables me to put real timecode to iso mpeg-4. Is there any specification within the standard to implement such a feature? I know there is a solution based on quicktime. (You could put an additional timecode track to an existing video track. But that?s not what i want to do, because nevertheless when preparing for streaming, the timecode track can not be hinted, right?) cheers, jochen From dsearles mmc.atmel.com Tue Oct 15 14:34:43 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Macroblock stuffing and short headers Message-ID: <3DAC51B3.20104@mmc.atmel.com> Conformance test stream hit037 is a short headers stream, with macroblock stuffing (derived_mb_type == "stuffing"). The macroblock() syntax allows derived_mb_type == "stuffing" regardless of short header, but macroblock() is effectively iterated over during short header parsing. The Momusys decoder doesn't handle this stream, but I believe MS does. The H263 spec does say that the bits should be discarded... maybe that operation was missed during the writing of macroblock() in MPEG-4? -- Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-0300 From zhenzhongc hotmail.com Wed Oct 16 11:12:20 2002 From: zhenzhongc hotmail.com (Chen Zhenzhong) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Masks Video References: <004101c2747e$eee547a0$611875c2@gauss> Message-ID: Morais: I am afraid that there is no segmentation tools can make masks of video sequences perfectly. :-( But I am still interested in other's anwser on Morais' question. Regards, Zhenzhong ----- Original Message ----- From: Vagner Morais To: MPEG4 Duvidas Sent: Wednesday, October 16, 2023 3:13 AM Subject: [M4IF Technotes] Masks Video Hi, Anyone know if is there any tool (ideally open-source) to make masks of videos? Thanks by your attention *************************************************************************************************** Vagner David Pinto Morais INESC - Porto | Instituto de Engenharia de Sistemas e Computadores do Porto Unidade de Telecomunica??es e Multim?dia Grupo de An?lise e S?ntese de Imagem Email : vmorais@inescporto.pt Tel. : +351 222094233 ****************************************************************************************************** _________ http://www.inescn.pt __| \ / | Campus da FEUP ___| /\ ___/\ | Rua Dr. Roberto Frias, n? 378 ___ |/_______\| 4200-465 PORTO ------------------------------------------------------------------------------ ******************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021016/b9438a58/attachment.html From dsearles mmc.atmel.com Wed Oct 16 10:22:35 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Unaligned gob_resync_marker can be unaligned? Message-ID: <3DAD681B.8060607@mmc.atmel.com> The spec shows that gob_resync_markers may be aligned, or unaligned. The Momusys reference software checks for both, but MS (1.0 ans 2.3) does not. Several shortheader streams do not parse under MS for this reason. I've added a patch to my code, but would like to know which way it should be? Thanks -- Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-0300 From salo rucus.ru.ac.za Wed Oct 16 23:45:45 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:00 2003 Subject: [M4IF Technotes] Bifs-text scene description - streaming movietexture In-Reply-To: Message-ID: <000301c27555$028696c0$9679e792@beast> > Hi Nicholas, > > I'm very curious if somebody answered your questions already. > I really > would like to know the answers as well. Hi Mathijn - thanks for the info you sent in that url. Yes, a lot of people have helped trying to answer this - big thanks to Cyril, Guido, JC and Mikael. The short answer, as far as I can work out, is that yes you are correct: It is theoretically possible, but hasn't been defined - not by ISMA or MPEG / ISO. Largely due to lack of interest, apparently.(!) What would work, is to write your own (or extend another) server so that it can parse something like rtsp://example.com/stream.mp4#1 to return the first track in stream.mp4, or similar. This is given that your player understands rtsp. If your OD looked like this: ObjectDescriptor { objectDescriptorID 20 esdescr [ ES_Descriptor { es_id 3 URLString "rtsp://localhost:554/..." decConfigDescr DecoderConfigDescriptor { objectTypeIndication 32 streamType 4 decSpecificInfo DecoderSpecificInfoString { } } slConfigDescr SLConfigDescriptor { ... } } ] } Cyril noted that "Here the URL shall point to a SL-packetized stream following the SLConfigDescriptor. You could point to a particular track in a remote mp4. In that case, the SLConfigDescriptor is empty. But again, I'm not sure that MPEG has standardized the way to access a track within an MP4 file." HTH. Kind regards, Nik. > > > That url business is quite tricky and there are no really > good sources > about it. > The delivery faq, section 24 specifies which url's will be > supported, > however: > http://mpeg.telecomitalialab.com/faq/mp4-dmi/mp4-dmi.htm#24 > > I also know that Envivio supports live broadcasts of video > and audio, > so it should be possible. I don't know which url syntax they use for > it... and if it is the standarized way to do it. > I also know that they allow streaming of content from within > the same > file, but that it is not possible to do it cross-file. The latter is > only possible with Inline nodes. But as you said, that is not really > flexible enough. > > What I get from the standard, is that it should all be possible in > theory. I think we've to wait for the implementations... > > Please let me know of any answers you receive... > > Best regards, > > Mathijn > > On Tuesday, Oct 15, 2002, at 14:33 Europe/Amsterdam, Nicholas Smit > wrote: > > >> Hi, > >> > >> Movie textures are not supported in any streaming standard > >> (RFC 3016 or ISMA). Neither do I know about any proprietary > >> way of doing it. You may have to device your own scheme for this. > >> > >> Regards, > >> Parimal. > >> > > > > > > Hmmm, so then how are applications like (implemented > versions of) the > > following from enst supposed to work? > > > > http://161.58.176.70/downloads/content/MIRRORSITE/interoperability/ > > ENST/ > > ecran2.mp4 > > > > Inlineing the sources is fine, until you want to manipulate > them in > > some > > way, such as full screening them.... > > > > > > Thanks, > > Nik. > > > > > > > >> -----Original Message----- > >> From: Nik [mailto:salo@rucus.ru.ac.za] > >> Sent: Tuesday, October 15, 2023 2:22 AM > >> To: technotes@lists.m4if.org > >> Subject: [M4IF Technotes] Bifs-text scene description - > >> streaming movietexture > >> > >> > >> Hi streaming experts, > >> > >> I'd really appreciate help with this. > >> > >> I need a movietexture that uses live video as its source, > >> such as "rtsp://localhost:554/vid.mp4". > >> > >> How do you specify the OD's and the ESD's for this? > >> > >> Can anyone give me some pointers on this, or even the url to > >> a similar mp4 on the net somewhere? > >> > >> I'm also a little confused as to how a player will know which > >> track to use from a url like "rtsp://localhost:554/vid.mp4" > >> for the movietexture....? > >> > >> > >> Any help very appreciated. > >> > >> Kind regards, > >> Nicholas. > >> > >> _______________________________________________ > >> Technotes mailing list > >> Technotes@lists.m4if.org > >> http://lists.m4if.org/mailman/listinfo/technot> es > >> > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technotes > > > > > < Mathijn Elhorst | Lead Developer > > Lost Boys B.V. | www.lostboys.nl | mathijn.elhorst@lostboys.nl > Joop Geesinkweg 209 | 1096 AV Amsterdam | The Netherlands > Ph: +31 20 5356161 | Mob: +31 6 52516810 | Fax: +31 20 4604501 > > > > From rob.koenen m4if.org Wed Oct 16 14:55:36 2002 From: rob.koenen m4if.org (Rob Koenen) Date: Wed Jul 30 14:09:01 2003 Subject: [M4IF Technotes] Bifs-text scene description - streaming movietexture In-Reply-To: <000301c27555$028696c0$9679e792@beast> Message-ID: <3C124172E7FDD511B510000347426D59011B3CEB@exchange.epr.com> > Yes, a lot of people have helped trying to answer this - big thanks to > Cyril, Guido, JC and Mikael. That's great - also my thanks to the people that tried to help out. It would be great if answers could be sent to the list as well, so that we can all benefit (and I am not stuck with the impression that questions go unanswered) Rob From singer apple.com Wed Oct 16 15:25:39 2002 From: singer apple.com (Dave Singer) Date: Wed Jul 30 14:09:01 2003 Subject: [M4IF Technotes] Bifs-text scene description - streaming movietexture In-Reply-To: <000301c27555$028696c0$9679e792@beast> References: <000301c27555$028696c0$9679e792@beast> Message-ID: At 22:45 +0200 10/16/02, Nicholas Smit wrote: > > Hi Nicholas, >> >> I'm very curious if somebody answered your questions already. >> I really >> would like to know the answers as well. > > >Hi Mathijn - thanks for the info you sent in that url. > >Yes, a lot of people have helped trying to answer this - big thanks to >Cyril, Guido, JC and Mikael. > >The short answer, as far as I can work out, is that yes you are correct: >It is theoretically possible, but hasn't been defined - not by ISMA or >MPEG / ISO. Largely due to lack of interest, apparently.(!) well, usually we refer to whole MP4 files, so the question of an anchor format doesn't come up. I'm curious; who defines the URL format for referencing anchors in files? Is this something that MPEG defines as it owns the file format, or W3C as they define what a URL looks like? > >What would work, is to write your own (or extend another) server so that >it can parse something like rtsp://example.com/stream.mp4#1 to return >the first track in stream.mp4, or similar. This is given that your >player understands rtsp. > > >If your OD looked like this: > > ObjectDescriptor { > objectDescriptorID 20 > esdescr [ > ES_Descriptor { > es_id 3 > URLString >"rtsp://localhost:554/..." > decConfigDescr >DecoderConfigDescriptor { > objectTypeIndication 32 > streamType 4 > decSpecificInfo >DecoderSpecificInfoString { > } > } > slConfigDescr SLConfigDescriptor >{ ... } > } > ] > } > >Cyril noted that "Here the URL shall point to a SL-packetized stream >following the SLConfigDescriptor. You could point to a particular track >in a remote mp4. In that case, the SLConfigDescriptor is empty. But >again, I'm not sure that MPEG has standardized the way to access a track >within an MP4 file." > >HTH. > >Kind regards, > >Nik. > > > > > > >> > >> >> That url business is quite tricky and there are no really >> good sources >> about it. >> The delivery faq, section 24 specifies which url's will be >> supported, >> however: >> http://mpeg.telecomitalialab.com/faq/mp4-dmi/mp4-dmi.htm#24 >> >> I also know that Envivio supports live broadcasts of video >> and audio, >> so it should be possible. I don't know which url syntax they use for >> it... and if it is the standarized way to do it. >> I also know that they allow streaming of content from within >> the same >> file, but that it is not possible to do it cross-file. The latter is >> only possible with Inline nodes. But as you said, that is not really >> flexible enough. >> >> What I get from the standard, is that it should all be possible in >> theory. I think we've to wait for the implementations... >> >> Please let me know of any answers you receive... >> >> Best regards, >> >> Mathijn >> >> On Tuesday, Oct 15, 2002, at 14:33 Europe/Amsterdam, Nicholas Smit >> wrote: >> >> >> Hi, >> >> >> >> Movie textures are not supported in any streaming standard >> >> (RFC 3016 or ISMA). Neither do I know about any proprietary >> >> way of doing it. You may have to device your own scheme for this. >> >> >> >> Regards, >> >> Parimal. >> >> >> > >> > >> > Hmmm, so then how are applications like (implemented >> versions of) the >> > following from enst supposed to work? >> > >> > http://161.58.176.70/downloads/content/MIRRORSITE/interoperability/ >> > ENST/ >> > ecran2.mp4 >> > >> > Inlineing the sources is fine, until you want to manipulate >> them in >> > some >> > way, such as full screening them.... >> > >> > >> > Thanks, >> > Nik. >> > >> > >> > >> >> -----Original Message----- >> >> From: Nik [mailto:salo@rucus.ru.ac.za] >> >> Sent: Tuesday, October 15, 2023 2:22 AM >> >> To: technotes@lists.m4if.org >> >> Subject: [M4IF Technotes] Bifs-text scene description - >> >> streaming movietexture >> >> >> >> >> >> Hi streaming experts, >> >> >> >> I'd really appreciate help with this. >> >> >> >> I need a movietexture that uses live video as its source, >> >> such as "rtsp://localhost:554/vid.mp4". >> >> >> >> How do you specify the OD's and the ESD's for this? > > >> >> >> Can anyone give me some pointers on this, or even the url to >> >> a similar mp4 on the net somewhere? >> >> >> >> I'm also a little confused as to how a player will know which >> >> track to use from a url like "rtsp://localhost:554/vid.mp4" >> >> for the movietexture....? >> >> >> >> >> >> Any help very appreciated. >> >> >> >> Kind regards, >> >> Nicholas. >> >> >> >> _______________________________________________ >> >> Technotes mailing list >> >> Technotes@lists.m4if.org >> >> http://lists.m4if.org/mailman/listinfo/technot> es >> >> >> > >> > _______________________________________________ >> > Technotes mailing list >> > Technotes@lists.m4if.org >> > http://lists.m4if.org/mailman/listinfo/technotes >> > >> > >> < Mathijn Elhorst | Lead Developer >> >> Lost Boys B.V. | www.lostboys.nl | mathijn.elhorst@lostboys.nl >> Joop Geesinkweg 209 | 1096 AV Amsterdam | The Netherlands >> Ph: +31 20 5356161 | Mob: +31 6 52516810 | Fax: +31 20 4604501 >> >> >> >> > >_______________________________________________ >Technotes mailing list >Technotes@lists.m4if.org >http://lists.m4if.org/mailman/listinfo/technotes -- David Singer Apple Computer/QuickTime From CC_Ju mtk.com.tw Thu Oct 17 16:08:57 2002 From: CC_Ju mtk.com.tw (CC_Ju@mtk.com.tw) Date: Wed Jul 30 14:09:02 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? Message-ID: Dear all: I don't know if I misunderstand the OBMC implementation in Momusys software. It seems to use only up, left, right, and current macroblock's motion vector to do OBMC operation. The motion vector in bottom macroblock seems don't used in OBMC operation. This differs from 14496-2 specification. Anybody can help? Thanks in advance. --- Chi-Cheng From lmartel atsana.com Thu Oct 17 13:16:09 2002 From: lmartel atsana.com (Luc Martel) Date: Wed Jul 30 14:09:02 2003 Subject: [M4IF Technotes] Intra coded blocks in P-Frames Message-ID: <002c01c275f8$818a9f40$e805000a@lumictech.com> Hi, This question may have already been asked. Sorry if it's the case but I couldn't find it in the archives. I would like to know what's happenning to the predictions in the case in which we have I blocks inserted into a P Frame. I'm going back and forth from the standard to the VM and it seems unclear. 1- What happens to the Motion Vector Prediction? Are the INTRA blocks considers as being outside the VOP? 2- What happens to the DC/AC predictions of the INTRA blocks? Should we consider the INTER coded blocks as being outside the VOP? Tkx, in advance. ---------------------------------------------------------------------------- --- Luc Martel, PhD Senior Engineer Atsana Semiconductors Ottawa, Canada. lmartel@atsana.com From rob.koenen m4if.org Thu Oct 17 11:50:27 2002 From: rob.koenen m4if.org (Rob Koenen) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] video compression & transmission over wireless In-Reply-To: <002d01c270c6$b2407740$3b1a640a@cyber.mmu.edu.my> Message-ID: <3C124172E7FDD511B510000347426D59011B3D1C@exchange.epr.com> Look at the news archive and the home page of the MPEG-4 Industry Forum's website. www.m4if.org and http://www.m4if.org/news/mpeg4.php Also look at the products page (a bit outdated though) and perhaps do a google search for 'MPEg-4' and 'chip' (http://www.google.com/search?sourceid=navclient &q=mpeg%2D4+chip) There is too much for me to list here. Good luck, Rob -----Original Message----- From: K. Ramasamy [mailto:k.ramasamy@mmu.edu.my] Sent: Thursday, October 10, 2023 18:37 To: technotes@lists.m4if.org Subject: [M4IF Technotes] video compression & transmission over wireless Dear sir, Greetings....... I'm planning to do my ph.D in video compression and transmission over wireless.I'm trying with the mpeg4 reference software for video compression. I think it is not handy. Is it possible to do both video compression(coder and decoder) and transmission over wireless channels using some hardware equipments. If so, Could you please inform me details of the manufacturers/dealers of hardware equipments for video(mpeg4) compression and transmission. Thanks in advance With best Regards, K.RAMASAMY LECTURER/FOE MULTIMEDIA UNIVERSITY JALAN MULTIMEDIA,63100 CYBERJAYA SELANGOR D.H MALAYSIA TEL:603-8312 5391 FAX:603 8318 3029 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021017/c5b0b30d/attachment.html From salo rucus.ru.ac.za Thu Oct 17 23:04:02 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] Layout and Form Message-ID: <000001c27618$585abef0$9679e792@beast> Hi experts, Regarding the layout and form nodes... A) Could anyone mention any free (or evaluation) players that implement this profile@level? (ie. Simple2D Graphics@Level1) B) If there arent, I cant test them, but what I would like to know is - if you have your layout set as you like (horizontally even spaced for example), and you add a new node to that group - will the nodes in the group be rearranged so that they are still evenly spaced? Thanks very much, Nicholas From Jean-Claude.Dufourd enst.fr Fri Oct 18 13:25:00 2002 From: Jean-Claude.Dufourd enst.fr (Jean-Claude Dufourd) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] Layout and Form References: <000001c27618$585abef0$9679e792@beast> Message-ID: <3DAFE17C.4010709@enst.fr> Nicholas Smit wrote: > Regarding the layout and form nodes... > > A) Could anyone mention any free (or evaluation) players that implement > this profile@level? (ie. Simple2D Graphics@Level1) Osmose will: this will be released during the next week on our web site. However, the only profile including Layout and Form is Complete2D. > B) If there arent, I cant test them, but what I would like to know is - > if you have your layout set as you like (horizontally even spaced for > example), and you add a new node to that group - will the nodes in the > group be rearranged so that they are still evenly spaced? Yes, layout is dynamic and should change if you insert new nodes or change the shape of existing nodes. Best regards JC -- Jean-Claude Dufourd @======================================@ ENST, Dept COMELEC The wing, over the big rock... 46, rue Barrault @======================================@ 75013 Paris Tel: +33145817807 Fax: +33145804036 From piccarre elet.polimi.it Fri Oct 18 14:17:12 2002 From: piccarre elet.polimi.it (Luca Piccarreta) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? References: Message-ID: <00bc01c27697$e83657d0$4b7baf83@piccarreta> I'm also very interested in this point. The spec for OBMC look very encoder-oriented. Encoders usually have all the MVs available at encoding time. Decoders, instead, work in a pipeline. In MS software also bottom vector is not used. And having the right vector available really complicates decoding. In fact, Motion compensation for a MB is carried out only after the subsequent MB header and MV (and shape) have been decoded. This may be (Not sure, though) the reason why OBMC is not present in any profile/level combination. So one may wonder why there is a tool that no encoder can use being sure that any decoder will ever be able to decode! Strictly following the specs for OBMC would heavily impact on performances (DCT coefs or output) for a whole row should be cached. Partially following the specs makes the decoder code cumbersome to seay the least (see decode_PVOP... code in MS Ref). Thanks to anyone that will shed light on this issue. Luca Piccarreta. ----- Original Message ----- From: To: Sent: Thursday, October 17, 2023 9:08 AM Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? > Dear all: > I don't know if I misunderstand the OBMC implementation in Momusys > software. It seems to use only up, left, right, and current macroblock's > motion vector to do OBMC operation. The motion vector in bottom macroblock > seems don't used in OBMC operation. This differs from 14496-2 > specification. Anybody can help? > > Thanks in advance. > > --- > Chi-Cheng > > > > > > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes From dsearles mmc.atmel.com Fri Oct 18 10:12:17 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? References: Message-ID: <3DB008B1.5040307@mmc.atmel.com> CC_Ju@mtk.com.tw wrote: >Dear all: > I don't know if I misunderstand the OBMC implementation in Momusys >software. It seems to use only up, left, right, and current macroblock's >motion vector to do OBMC operation. The motion vector in bottom macroblock >seems don't used in OBMC operation. This differs from 14496-2 >specification. Anybody can help? > >Thanks in advance. > >--- >Chi-Cheng >_______________________________________________ >Technotes mailing list >Technotes@lists.m4if.org >http://lists.m4if.org/mailman/listinfo/technotes > > I didn't look at the Momusys code, so this may not be the answer to your question, but the spec has a line in it that implies that you will never actually have to have a motion vector from the macroblock under the one being processed: "if the current block is at the bottom of the macroblock, the remote motion vector corresponding with an 8*8 luminance block in the macroblock below the current macroblock is replaced by the motion vector for the current block." Dan Searles -- Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-0300 From dsearles mmc.atmel.com Fri Oct 18 11:31:44 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] MS reference sw bug References: <00bc01c27697$e83657d0$4b7baf83@piccarreta> Message-ID: <3DB01B50.3010202@mmc.atmel.com> The following applies to versions 1.0, and 2.3. Version 2.3 does have some modifications in the functions affected, but did not completely remove the bug. Test streams mit013 and mit014 demonstrate the bug. I noticed green specks in the decoded sequences, that were not present in the Momusys decoded sequence. I've been running under VC++ 5.0. The functions decodeIVOP_DataPartitioning(), decodeMBTextureModeOfPVOP_DataPartitioning() and possibly decodeIVOP_WithShape_DataPartitioning() and decodePVOP_WithShape_DataPartitioning(). Part of the fix that version 2.3 has is that pointer m_piMCBPC is initialized to point to an array with one more than the number of macroblocks in the frame. This is needed because some things are stored off through the pointer after it has been incremented past the last macroblock of the frame (due to mb stuffing after the frame). The real problem, however, is stuff stored off through pmbmd. This pointer has also been incremented out of the frame, and out of the memory array allocated. Even the new decodeMBTextureModeOfPVOP_DataPartitioning() in version 2.3 stores through pmbmd before the check for the stuffing code. The areas that are still questionable to me are in decodeIVOP_WithShape_DataPartitioning() and decodePVOP_WithShape_DataPartitioning(). Is mb stuffing after the "frame" valid? The uses of pmbmd are more complex in these routines, so the check for mb stuffing needs to be done earlier, if possible. Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-0300 From garysull microsoft.com Fri Oct 18 09:24:39 2002 From: garysull microsoft.com (Gary Sullivan) Date: Wed Jul 30 14:09:03 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? Message-ID: <3B8749EA44FC2748956538D4A3096E42E08A5A@RED-MSG-01.redmond.corp.microsoft.com> The bottom remote vectors should be used for the upper two 8x8 blocks within the macroblock, but not for the lower two 8x8 blocks (for which the remote motion vector would be from the next row of macrobloks). I think this is stated in the spec if you look closely enough. OBMC as in MPEG-4 was copied directly from H.263 with the same design, but was not ever put into a profile of MPEG-4. OBMC is actually used a lot in H.263 in the videoconferencing industry. It gives a significant improvement in visual quality, and it's not terribly difficult to decode - it just requires a one-macroblock pipelining delay in the decoding process. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Luca Piccarreta [mailto:piccarre@elet.polimi.it] +> Sent: Friday, October 18, 2023 4:17 AM +> To: technotes@lists.m4if.org +> Subject: Re: [M4IF Technotes] Is OBMC implementation +> different from Momusys and 14496-2?? +> +> +> I'm also very interested in this point. +> The spec for OBMC look very encoder-oriented. +> Encoders usually have all the MVs available at +> encoding time. +> Decoders, instead, work in a pipeline. +> In MS software also bottom vector is not used. +> And having the right vector available really +> complicates decoding. +> In fact, Motion compensation for a MB is carried +> out only after the subsequent MB header and MV +> (and shape) have been decoded. +> This may be (Not sure, though) the reason why +> OBMC is not present in any profile/level +> combination. +> So one may wonder why there is a tool that no encoder +> can use being sure that any decoder will ever be able +> to decode! +> Strictly following the specs for OBMC would heavily +> impact on performances (DCT coefs or output) for a whole +> row should be cached. +> Partially following the specs makes the decoder code +> cumbersome to seay the least (see decode_PVOP... +> code in MS Ref). +> Thanks to anyone that will shed light on this issue. +> Luca Piccarreta. +> +> ----- Original Message ----- +> From: +> To: +> Sent: Thursday, October 17, 2023 9:08 AM +> Subject: [M4IF Technotes] Is OBMC implementation different +> from Momusys and +> 14496-2?? +> +> +> > Dear all: +> > I don't know if I misunderstand the OBMC +> implementation in Momusys +> > software. It seems to use only up, left, right, and +> current macroblock's +> > motion vector to do OBMC operation. The motion vector in +> bottom macroblock +> > seems don't used in OBMC operation. This differs from 14496-2 +> > specification. Anybody can help? +> > +> > Thanks in advance. +> > +> > --- +> > Chi-Cheng +> > +> > +> > +> > +> > +> > +> > +> > +> > _______________________________________________ +> > Technotes mailing list +> > Technotes@lists.m4if.org +> > http://lists.m4if.org/mailman/listinfo/technotes +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> From piccarre elet.polimi.it Fri Oct 18 18:47:01 2002 From: piccarre elet.polimi.it (Luca Piccarreta) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? References: <3B8749EA44FC2748956538D4A3096E42E08A5A@RED-MSG-01.redmond.corp.microsoft.com> Message-ID: <002a01c276bd$9966e8b0$4b7baf83@piccarreta> Thanks for the answer. But I'm still a little bit perplexed. May be these questions have been asked lots of times, but how should a decoder behave at the end of a video packet (OBMC)? I suppose that the MV vector of the macroblock to the right should be used. (If I remember right, MV vectors of other video packets are valid for OBMC prediction). Does this imply that a Video Packet cannot be decoded correctly on its own? Regards, Luca. ----- Original Message ----- From: "Gary Sullivan" To: "Luca Piccarreta" ; Sent: Friday, October 18, 2023 5:24 PM Subject: RE: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? The bottom remote vectors should be used for the upper two 8x8 blocks within the macroblock, but not for the lower two 8x8 blocks (for which the remote motion vector would be from the next row of macrobloks). I think this is stated in the spec if you look closely enough. OBMC as in MPEG-4 was copied directly from H.263 with the same design, but was not ever put into a profile of MPEG-4. OBMC is actually used a lot in H.263 in the videoconferencing industry. It gives a significant improvement in visual quality, and it's not terribly difficult to decode - it just requires a one-macroblock pipelining delay in the decoding process. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Luca Piccarreta [mailto:piccarre@elet.polimi.it] +> Sent: Friday, October 18, 2023 4:17 AM +> To: technotes@lists.m4if.org +> Subject: Re: [M4IF Technotes] Is OBMC implementation +> different from Momusys and 14496-2?? +> +> +> I'm also very interested in this point. +> The spec for OBMC look very encoder-oriented. +> Encoders usually have all the MVs available at +> encoding time. +> Decoders, instead, work in a pipeline. +> In MS software also bottom vector is not used. +> And having the right vector available really +> complicates decoding. +> In fact, Motion compensation for a MB is carried +> out only after the subsequent MB header and MV +> (and shape) have been decoded. +> This may be (Not sure, though) the reason why +> OBMC is not present in any profile/level +> combination. +> So one may wonder why there is a tool that no encoder +> can use being sure that any decoder will ever be able +> to decode! +> Strictly following the specs for OBMC would heavily +> impact on performances (DCT coefs or output) for a whole +> row should be cached. +> Partially following the specs makes the decoder code +> cumbersome to seay the least (see decode_PVOP... +> code in MS Ref). +> Thanks to anyone that will shed light on this issue. +> Luca Piccarreta. +> +> ----- Original Message ----- +> From: +> To: +> Sent: Thursday, October 17, 2023 9:08 AM +> Subject: [M4IF Technotes] Is OBMC implementation different +> from Momusys and +> 14496-2?? +> +> +> > Dear all: +> > I don't know if I misunderstand the OBMC +> implementation in Momusys +> > software. It seems to use only up, left, right, and +> current macroblock's +> > motion vector to do OBMC operation. The motion vector in +> bottom macroblock +> > seems don't used in OBMC operation. This differs from 14496-2 +> > specification. Anybody can help? +> > +> > Thanks in advance. +> > +> > --- +> > Chi-Cheng +> > +> > +> > +> > +> > +> > +> > +> > +> > _______________________________________________ +> > Technotes mailing list +> > Technotes@lists.m4if.org +> > http://lists.m4if.org/mailman/listinfo/technotes +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> From garysull microsoft.com Sat Oct 19 04:02:32 2002 From: garysull microsoft.com (Gary Sullivan) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Is OBMC implementation different from Momusys and 14496-2?? Message-ID: <3B8749EA44FC2748956538D4A3096E42E08A5F@RED-MSG-01.redmond.corp.microsoft.com> Luca, The answer is Yes (in MPEG-4 OBMC). But the answer doesn't matter. In H.263 slice-mode operation, the answer is No. MPEG-4 OBMC was developed from the original version of H.263, which did not have slices/"video packets" as they are known in MPEG-1 or MPEG-4. When slice-structured coding was added to H.263, the decoding process for each slice was made independent of other slices within the same picture (at least prior to deblocking filtering). In fact the slices could even be sent and decoded out of order. MPEG-4 made a different choice. Some people indicated that perceptual quality was better if the OBMC would use MVs from other slices. Others argued that independent decoding of slices was a better idea - partly because of the out-of-order decoding capability this offered. But anyway this is all strictly an academic question as OBMC is not in any MPEG-4 profiles! And if MPEG ever decided to create a profile that included OBMC, MPEG would be entirely free to make OBMC operate differently in that new profile! So the question is purely academic. Best Regards, -Gary S. +> -----Original Message----- +> From: Luca Piccarreta [mailto:piccarre@elet.polimi.it] +> Sent: Friday, October 18, 2023 8:47 AM +> To: Gary Sullivan; technotes@lists.m4if.org +> Subject: Re: [M4IF Technotes] Is OBMC implementation +> different from Momusys and 14496-2?? +> +> +> Thanks for the answer. +> But I'm still a little bit perplexed. +> May be these questions have been asked lots of times, +> but how should a decoder behave at the end +> of a video packet (OBMC)? I suppose that the MV vector of +> the macroblock +> to the right should be used. (If I remember right, MV +> vectors of other +> video packets are valid for OBMC prediction). +> Does this imply that a Video Packet cannot be decoded correctly +> on its own? +> Regards, +> Luca. +> +> ----- Original Message ----- +> From: "Gary Sullivan" +> To: "Luca Piccarreta" ; +> +> Sent: Friday, October 18, 2023 5:24 PM +> Subject: RE: [M4IF Technotes] Is OBMC implementation +> different from Momusys +> and 14496-2?? +> +> +> +> The bottom remote vectors should be used for the upper two 8x8 +> blocks within the macroblock, +> but not for the lower two 8x8 blocks (for which the remote +> motion vector would be from the next row of macrobloks). I think +> this is stated in the spec if you look closely enough. +> +> OBMC as in MPEG-4 was copied directly from H.263 with the same +> design, but was not ever put into a profile of MPEG-4. +> +> OBMC is actually used a lot in H.263 in the +> videoconferencing industry. +> It gives a significant improvement in visual quality, and it's +> not terribly difficult to decode - it just requires a one-macroblock +> pipelining delay in the decoding process. +> +> Best Regards, +> +> Gary Sullivan +> +> +> -----Original Message----- +> +> From: Luca Piccarreta [mailto:piccarre@elet.polimi.it] +> +> Sent: Friday, October 18, 2023 4:17 AM +> +> To: technotes@lists.m4if.org +> +> Subject: Re: [M4IF Technotes] Is OBMC implementation +> +> different from Momusys and 14496-2?? +> +> +> +> +> +> I'm also very interested in this point. +> +> The spec for OBMC look very encoder-oriented. +> +> Encoders usually have all the MVs available at +> +> encoding time. +> +> Decoders, instead, work in a pipeline. +> +> In MS software also bottom vector is not used. +> +> And having the right vector available really +> +> complicates decoding. +> +> In fact, Motion compensation for a MB is carried +> +> out only after the subsequent MB header and MV +> +> (and shape) have been decoded. +> +> This may be (Not sure, though) the reason why +> +> OBMC is not present in any profile/level +> +> combination. +> +> So one may wonder why there is a tool that no encoder +> +> can use being sure that any decoder will ever be able +> +> to decode! +> +> Strictly following the specs for OBMC would heavily +> +> impact on performances (DCT coefs or output) for a whole +> +> row should be cached. +> +> Partially following the specs makes the decoder code +> +> cumbersome to seay the least (see decode_PVOP... +> +> code in MS Ref). +> +> Thanks to anyone that will shed light on this issue. +> +> Luca Piccarreta. +> +> +> +> ----- Original Message ----- +> +> From: +> +> To: +> +> Sent: Thursday, October 17, 2023 9:08 AM +> +> Subject: [M4IF Technotes] Is OBMC implementation different +> +> from Momusys and +> +> 14496-2?? +> +> +> +> +> +> > Dear all: +> +> > I don't know if I misunderstand the OBMC +> +> implementation in Momusys +> +> > software. It seems to use only up, left, right, and +> +> current macroblock's +> +> > motion vector to do OBMC operation. The motion vector in +> +> bottom macroblock +> +> > seems don't used in OBMC operation. This differs from 14496-2 +> +> > specification. Anybody can help? +> +> > +> +> > Thanks in advance. +> +> > +> +> > --- +> +> > Chi-Cheng +> +> > +> +> > +> +> > +> +> > +> +> > +> +> > +> +> > +> +> > +> +> > _______________________________________________ +> +> > Technotes mailing list +> +> > Technotes@lists.m4if.org +> +> > http://lists.m4if.org/mailman/listinfo/technotes +> +> +> +> _______________________________________________ +> +> Technotes mailing list +> +> Technotes@lists.m4if.org +> +> http://lists.m4if.org/mailman/listinfo/technotes +> +> +> +> From Manu.Batura patni.com Sat Oct 19 17:09:39 2002 From: Manu.Batura patni.com (Manu V Batura) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Escape coding Message-ID: <001101c2775b$d4b69520$5402a8c0@pcp33246> Hi, The MPEG-4 video standard indicates that the values 0 and 128 are reserved while performing escape coding. They should not be transmitted by the encoder. The decoder should also convert a received 255 level to 128. Can anyone please explain the reason for doing so? Which levels should be transmitted if the levels are 0, 128 or 255 after prediction/quantization? Thanks & Regards, Manu From garysull microsoft.com Sat Oct 19 05:44:08 2002 From: garysull microsoft.com (Gary Sullivan) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Escape coding Message-ID: <3B8749EA44FC2748956538D4A3096E42E08A62@RED-MSG-01.redmond.corp.microsoft.com> The purpose of avoiding use of fixed-length coded values of 0 and 128 is to prevent accidental start-code emulation within the video data bitstream. Both of those values end in seven zeros, which is too many. (And one of them is all zero, which is even worse.) I believe there is no need for an escape-coded value of zero. If the thing is zero, you should just skip the coefficient value (at least if it's AC). If you want to represent something as close as possible to 255, send 254. If you want to represent something as close as possible to 0, either send 1 or represent the coefficient by skipping it. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Manu V Batura [mailto:Manu.Batura@patni.com] +> Sent: Saturday, October 19, 2023 3:40 AM +> To: technotes@lists.m4if.org +> Subject: [M4IF Technotes] Escape coding +> +> +> Hi, +> +> The MPEG-4 video standard indicates that the values 0 and +> 128 are reserved +> while performing escape coding. +> They should not be transmitted by the encoder. The decoder +> should also +> convert a received 255 level +> to 128. Can anyone please explain the reason for doing so? +> Which levels +> should be transmitted if the levels are +> 0, 128 or 255 after prediction/quantization? +> +> Thanks & Regards, +> Manu +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> From hwang jdl.ac.cn Sun Oct 20 12:03:51 2002 From: hwang jdl.ac.cn (WangHui) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] help for yuv player Message-ID: <001f01c277e5$51313270$242ae29f@WangHui> Hi, dear all Now I am engaged in building up a test platform for MP4, I use MoMusys encoder and decoder. Could you tell me where I can get the player for .yuv file under the Redhat Linux 7.2. Thank you very much best regards WangHui -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021020/e4dc0806/attachment.html From yangwj lucent.com Sun Oct 20 22:18:36 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Free codec support Error Resilience in DirectShouw filter Message-ID: <6999EF22C12CD4119A6800508B65021F03FE2E90@CI0027EXCH001U> hello, Can anyone tell me which free mpeg4 codec that support Error Resilience is in DirectShouw filter format? Thanks a lot Mac -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021020/5ffaa99d/attachment.html From michaelni gmx.at Sun Oct 20 19:29:09 2002 From: michaelni gmx.at (Michael Niedermayer) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Free codec support Error Resilience in DirectShouw filter In-Reply-To: <6999EF22C12CD4119A6800508B65021F03FE2E90@CI0027EXCH001U> References: <6999EF22C12CD4119A6800508B65021F03FE2E90@CI0027EXCH001U> Message-ID: <200210201829.10268.michaelni@gmx.at> Hi On Sunday 20 October 2023 15:18, Yang, Wei Jian (Mac) wrote: > hello, > > Can anyone tell me which free mpeg4 codec that support Error Resilience is > in DirectShouw filter format? ffdshow, it supports data partitioning & video packets (RVLC is not supported yet) see: http://cutka.szm.sk/ffdshow [...] Michael From moses kampsax.dtu.dk Mon Oct 21 00:35:00 2002 From: moses kampsax.dtu.dk (moses woldeselassie) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Alpha file Message-ID: hi all > I'm using MoMuSys-FDIS-V1.0-990812 in SOLARIS platform. But > I've a problems with the alpha file: > > 1. when type of alpha channel is 1-Binary > - ERROR: can not open Alpha file > GetBoundingRect: alpha plane without any object!! > Empty VOP at 3266.67 > > 2. Type of alpha channel is 0-Rectangular > > P[ 3] : time=100.000 video_obj_id=0 layer_id=0 quant=4 > P[ 3] flds : Y=0 C=0 MVs=0 cbpy=0 mcbpc=0 ilace=0 > P[ 3] mblk : inter=396 inter4v=0 intra=0 field=0 skip=0 > fldDCT=0 P[ 3] bits : shape=0 motion=0 texture=5 vop=55 > total=456 P[ 3] PSNR : Y=48.130802 U=Inf V=Inf > > - ERROR: can not open Alpha file > Performing motion estimation > - OBMC mode ON > - 4 MV per MB mode ON > - unrestricted mode ON > Coding INTER texture > > 3. Additionally, encoder write to the output > akiyo_cif_enc.yuv, but it does not write the alpha file. > > ORIG2/akiyo_cif_enc.yuv {Name of source VOP Y file on disk} > ORIG2/akiyo_cif_enc.yuv {Name of source VOP U file on disk} > ORIG2/akiyo_cif_enc.yuv {Name of source VOP V file on disk} > ORIG2/akiyo_cif.seg {Name of source VOP Alpha file on disk} > > how do I solve this problems > > I'm looking forward to get your help. Best regards Moses (Mussie) <<<>>><<<>>><<<>>><<<>>><<<>>><<<>>><<<>>><<>>><<<>>> YOUR BRAIN IS HUNGRY Wireless Networks Technology, News and Opinion. <<<>>><<<>>><<<>>><<<>>><<<>>><<<>>><<<>>><<<>>><<<>>> From Manu.Batura patni.com Mon Oct 21 10:44:21 2002 From: Manu.Batura patni.com (Manu V Batura) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Escape coding In-Reply-To: <3B8749EA44FC2748956538D4A3096E42E08A62@RED-MSG-01.redmond.corp.microsoft.com> Message-ID: <000601c278b8$55b0f050$5402a8c0@pcp33246> Hello Gary, Does it mean that we also need to perform inv. prediction at the encoder end rather than just inv Quantizaton and IDCT to avoid propagation of errors introduced due to such changes (like 255 to 254) since a coefficient might get a value of 255 after prediction. Regards, Manu -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Gary Sullivan Sent: Saturday, October 19, 2023 5:14 PM To: Manu.Batura@patni.com; technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Escape coding The purpose of avoiding use of fixed-length coded values of 0 and 128 is to prevent accidental start-code emulation within the video data bitstream. Both of those values end in seven zeros, which is too many. (And one of them is all zero, which is even worse.) I believe there is no need for an escape-coded value of zero. If the thing is zero, you should just skip the coefficient value (at least if it's AC). If you want to represent something as close as possible to 255, send 254. If you want to represent something as close as possible to 0, either send 1 or represent the coefficient by skipping it. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Manu V Batura [mailto:Manu.Batura@patni.com] +> Sent: Saturday, October 19, 2023 3:40 AM +> To: technotes@lists.m4if.org +> Subject: [M4IF Technotes] Escape coding +> +> +> Hi, +> +> The MPEG-4 video standard indicates that the values 0 and +> 128 are reserved +> while performing escape coding. +> They should not be transmitted by the encoder. The decoder +> should also +> convert a received 255 level +> to 128. Can anyone please explain the reason for doing so? +> Which levels +> should be transmitted if the levels are +> 0, 128 or 255 after prediction/quantization? +> +> Thanks & Regards, +> Manu +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From egoam libero.it Mon Oct 21 11:02:50 2002 From: egoam libero.it (egoam@libero.it) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] =?iso-8859-1?Q?CBR_encoding?= Message-ID: Hi all! Can anyone tell me how to do constant bitrate encoding with Momusys? Best regards, Andrea Ambrosioni From garysull microsoft.com Mon Oct 21 03:56:14 2002 From: garysull microsoft.com (Gary Sullivan) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] Escape coding Message-ID: <3B8749EA44FC2748956538D4A3096E42E08A6C@RED-MSG-01.redmond.corp.microsoft.com> Strictly speaking, the encoder can do what it wants. Practically speaking, the encoder needs to track what the decoder will do in order to use the decoded pictures for subsequent motion-compensated prediction. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Manu V Batura [mailto:Manu.Batura@patni.com] +> Sent: Sunday, October 20, 2023 9:14 PM +> To: Gary Sullivan; technotes@lists.m4if.org +> Subject: RE: [M4IF Technotes] Escape coding +> +> +> Hello Gary, +> +> Does it mean that we also need to perform inv. prediction at +> the encoder +> end rather than just inv Quantizaton and IDCT to avoid propagation of +> errors introduced due to such changes (like 255 to 254) +> since a coefficient +> might get a value of 255 after prediction. +> +> Regards, +> Manu +> +> -----Original Message----- +> From: technotes-admin@lists.m4if.org +> [mailto:technotes-admin@lists.m4if.org]On Behalf Of Gary Sullivan +> Sent: Saturday, October 19, 2023 5:14 PM +> To: Manu.Batura@patni.com; technotes@lists.m4if.org +> Subject: RE: [M4IF Technotes] Escape coding +> +> +> +> The purpose of avoiding use of fixed-length coded values of 0 and 128 +> is to prevent accidental start-code emulation within the video data +> bitstream. +> Both of those values end in seven zeros, which is too many. +> (And one of +> them +> is all zero, which is even worse.) +> +> I believe there is no need for an escape-coded value of zero. If the +> thing +> is zero, you should just skip the coefficient value (at least if it's +> AC). +> +> If you want to represent something as close as possible to 255, send +> 254. +> If you want to represent something as close as possible to 0, either +> send +> 1 or represent the coefficient by skipping it. +> +> Best Regards, +> +> Gary Sullivan +> +> +> -----Original Message----- +> +> From: Manu V Batura [mailto:Manu.Batura@patni.com] +> +> Sent: Saturday, October 19, 2023 3:40 AM +> +> To: technotes@lists.m4if.org +> +> Subject: [M4IF Technotes] Escape coding +> +> +> +> +> +> Hi, +> +> +> +> The MPEG-4 video standard indicates that the values 0 and +> +> 128 are reserved +> +> while performing escape coding. +> +> They should not be transmitted by the encoder. The decoder +> +> should also +> +> convert a received 255 level +> +> to 128. Can anyone please explain the reason for doing so? +> +> Which levels +> +> should be transmitted if the levels are +> +> 0, 128 or 255 after prediction/quantization? +> +> +> +> Thanks & Regards, +> +> Manu +> +> +> +> _______________________________________________ +> +> Technotes mailing list +> +> Technotes@lists.m4if.org +> +> http://lists.m4if.org/mailman/listinfo/technotes +> +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> From pescatorimarco hotmail.com Mon Oct 21 16:10:57 2002 From: pescatorimarco hotmail.com (Marco Pescatori) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] MPEG-4 bitstream syntax Message-ID: I need to extract the MB's motion vector information from the coded bitsteam .What can I do? If somebody can help me,please contact me. Best Regard. Marco. _________________________________________________________________ MSN Search: la risposta alle tue ricerche online http://search.msn.it/ From wmay cisco.com Mon Oct 21 10:34:07 2002 From: wmay cisco.com (Bill May) Date: Wed Jul 30 14:09:04 2003 Subject: [M4IF Technotes] help for yuv player References: <001f01c277e5$51313270$242ae29f@WangHui> Message-ID: <3DB42C7F.9080707@cisco.com> You can try our mpeg4ip open source package - it has a yuvdump utility included. Bill May WangHui wrote: > Hi, dear all > > Now I am engaged in building up a test platform > for MP4, I use MoMusys encoder and decoder. > > Could you tell me where I can get the player for > .yuv file under the Redhat Linux 7.2. > > Thank you very much > > best regards > > WangHui From zhenzhongc hotmail.com Tue Oct 22 10:51:55 2002 From: zhenzhongc hotmail.com (Zhenzhong Chen) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] CBR encoding References: Message-ID: Dear AA, in the example control file (.ctl file) n <-- {Type of rate control: 0-NONE, 1-VM4, 2-VM5+, 3-TM5} n|n|n|n|n|n <-- {RC String: Alg|rate_mod|dist_mod|Target|buffer|mode} set these parameters properly. For CBR, you should choose a rate control type (VM4, VM5+ or TM5). You'd better read the readme file which is provided with the source code. Regards, Zhenzhong ----- Original Message ----- From: To: "technotes" Sent: Monday, October 21, 2023 4:02 PM Subject: [M4IF Technotes] CBR encoding > Hi all! > > Can anyone tell me how to do constant bitrate encoding with Momusys? > > Best regards, > Andrea Ambrosioni > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From yangwj lucent.com Tue Oct 22 13:08:57 2002 From: yangwj lucent.com (Yang, Wei Jian (Mac)) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] display YUV in windows Message-ID: <6999EF22C12CD4119A6800508B65021F03FE2E93@CI0027EXCH001U> hello, Does any one know the API that can display YUV(4:2:0) images in windows? Thanks Yang, Wei Jian (Mac) Bell labs, Shanghai Lucent Technologies (China) Co., Ltd. Tel: 8621-5450-4555 ext. 8610 Fax: 8621-5490-2106 E-mail: yangwj@lucent.com From ben interframemedia.com Tue Oct 22 00:18:55 2002 From: ben interframemedia.com (Ben Waggoner) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] CELP test files or encoder? In-Reply-To: Message-ID: Folks, All the MPEG-4 encoders I have my hands on only support AAC-LC audio, not CELP. Does anyone have a MPEG-4 file with CELP audio I can use for some tests, or know of an encoder that can make CELP audio? Ben Waggoner Compressed Video Consulting, Training, and Encoding My Book: http://www.benwaggoner.com/books.htm Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html Compression Books: http://www.benwaggoner.com/bookshelf.htm From gertzen irt.de Tue Oct 22 10:24:08 2002 From: gertzen irt.de (Gertzen) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] display YUV in windows In-Reply-To: <6999EF22C12CD4119A6800508B65021F03FE2E93@CI0027EXCH001U> Message-ID: Here is a perfect yuv player. It helped me so much. http://kbs.cs.tu-berlin.de/~stewe/vceg/tools.htm Am 22.10.2023 06:08:57, schrieb "Yang, Wei Jian (Mac)" : >hello, > >Does any one know the API that can display YUV(4:2:0) images in windows? > >Thanks > >Yang, Wei Jian (Mac) >Bell labs, Shanghai >Lucent Technologies (China) Co., Ltd. >Tel: 8621-5450-4555 ext. 8610 >Fax: 8621-5490-2106 >E-mail: yangwj@lucent.com > >_______________________________________________ >Technotes mailing list >Technotes@lists.m4if.org >http://lists.m4if.org/mailman/listinfo/technotes > > From egoam libero.it Tue Oct 22 11:56:24 2002 From: egoam libero.it (egoam@libero.it) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] =?iso-8859-1?Q?Re:_[M4IF_Technotes]_CBR_encoding?= Message-ID: I read the readme file and I used rate control, but the bitrate I obtained seems not to be so constant as I expected: the frame sizes are distributed in a large range of values. How can I obtain a flatter coding? Best regards, Andrea > Dear AA, > > in the example control file (.ctl file) > > n <-- {Type of rate control: 0-NONE, 1-VM4, 2-VM5+, 3-TM5} > n|n|n|n|n|n <-- {RC String: Alg|rate_mod|dist_mod|Target|buffer|mode} > > set these parameters properly. > For CBR, you should choose a rate control type (VM4, VM5+ or TM5). > > You'd better read the readme file which is provided with the source code. > > Regards, > > Zhenzhong > ----- Original Message ----- > From: > To: "technotes" > Sent: Monday, October 21, 2023 4:02 PM > Subject: [M4IF Technotes] CBR encoding > > > > Hi all! > > > > Can anyone tell me how to do constant bitrate encoding with Momusys? > > > > Best regards, > > Andrea Ambrosioni > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technotes > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From a.ilangovan gdatech.co.in Tue Oct 22 16:17:58 2002 From: a.ilangovan gdatech.co.in (A.Ilangovan) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] Re: [M4IF Technotes] CBR encoding References: Message-ID: <000601c279b0$1b611f30$de00a8c0@ilangovan> Dear Andrea Ambrosioni Don't expect your compressed frame sizes to be exactly the same as you set/expect. I have my experiences with only TM5, which will generate some statistics from the previous frame!! Hence, be happy if the size is +/- 10% of the target size and gets adjusted(in TM5) in the following frames!!! With Love A.Ilangovan ========================================== ----- Original Message ----- From: To: "zhenzhongc" Cc: "technotes" Sent: Tuesday, October 22, 2023 2:26 PM Subject: [M4IF Technotes] Re: [M4IF Technotes] CBR encoding > I read the readme file and I used rate control, but the bitrate I > obtained seems not to be so constant as I expected: the frame sizes are > distributed in a large range of values. > How can I obtain a flatter coding? > > Best regards, > Andrea > > > Dear AA, > > > > in the example control file (.ctl file) > > > > n <-- {Type of rate control: 0-NONE, 1-VM4, 2-VM5+, > 3-TM5} > > n|n|n|n|n|n <-- {RC String: > Alg|rate_mod|dist_mod|Target|buffer|mode} > > > > set these parameters properly. > > For CBR, you should choose a rate control type (VM4, VM5+ or TM5). > > > > You'd better read the readme file which is provided with the source > code. > > > > Regards, > > > > Zhenzhong > > ----- Original Message ----- > > From: > > To: "technotes" > > Sent: Monday, October 21, 2023 4:02 PM > > Subject: [M4IF Technotes] CBR encoding > > > > > > > Hi all! > > > > > > Can anyone tell me how to do constant bitrate encoding with Momusys? > > > > > > Best regards, > > > Andrea Ambrosioni > > > > > > _______________________________________________ > > > Technotes mailing list > > > Technotes@lists.m4if.org > > > http://lists.m4if.org/mailman/listinfo/technotes > > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technotes > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.401 / Virus Database: 226 - Release Date: 10/9/2023 From wmay cisco.com Tue Oct 22 10:42:06 2002 From: wmay cisco.com (Bill May) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] display YUV in windows References: <6999EF22C12CD4119A6800508B65021F03FE2E93@CI0027EXCH001U> Message-ID: <3DB57FDE.7020006@cisco.com> Try yuvdump, which is part of mpeg4ip (www.mpeg4ip.net) Bill May Yang, Wei Jian (Mac) wrote: > hello, > > Does any one know the API that can display YUV(4:2:0) images in windows? > > Thanks > > Yang, Wei Jian (Mac) > Bell labs, Shanghai > Lucent Technologies (China) Co., Ltd. > Tel: 8621-5450-4555 ext. 8610 > Fax: 8621-5490-2106 > E-mail: yangwj@lucent.com > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From grl iis.fhg.de Wed Oct 23 13:05:10 2002 From: grl iis.fhg.de (Bernhard Grill) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] CELP test files or encoder? In-Reply-To: References: Message-ID: <3DB67456.5030201@iis.fhg.de> Ben Waggoner wrote: > Folks, > > All the MPEG-4 encoders I have my hands on only support AAC-LC audio, > not CELP. Does anyone have a MPEG-4 file with CELP audio I can use for some > tests, or know of an encoder that can make CELP audio? The MPEG-4 reference software includes a CELP encoder which produces reasonable quality. Best regards, Bernhard Grill > > > > Ben Waggoner > Compressed Video Consulting, Training, and Encoding > > My Book: http://www.benwaggoner.com/books.htm > Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html > Compression Books: http://www.benwaggoner.com/bookshelf.htm > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes -- Dr. Bernhard Grill email: grl@iis.fhg.de Head of Audio Department, FhG-IIS A phone: +49 9131 776-351 Am Weichselgarten 3, D-91058 Erlangen, Germany FAX: +49 9131 776-398 From rkoenen intertrust.com Wed Oct 23 10:21:59 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] CELP test files or encoder? Message-ID: <3C124172E7FDD511B510000347426D5901324F2F@exchange.epr.com> > The MPEG-4 reference software includes a CELP encoder which > produces reasonable quality. I think Ben is looking for products, not source code, but he can speak for himself. Envivio has a CELP coder in their product, and there must be others - please speak up. Also, I have forwarded the request to a few parties that may have or know about CELP technology. Rob > -----Original Message----- > From: Bernhard Grill [mailto:grl@iis.fhg.de] > Sent: Wednesday, October 23, 2023 3:05 > To: Ben Waggoner > Cc: technotes@lists.m4if.org > Subject: Re: [M4IF Technotes] CELP test files or encoder? > > > Ben Waggoner wrote: > > > Folks, > > > > All the MPEG-4 encoders I have my hands on only support AAC-LC > > audio, not CELP. Does anyone have a MPEG-4 file with CELP > audio I can > > use for some tests, or know of an encoder that can make CELP audio? > > > The MPEG-4 reference software includes a CELP encoder which > produces reasonable > quality. > > > Best regards, > Bernhard Grill > > > > > > > > > Ben Waggoner > > Compressed Video Consulting, Training, and Encoding > > > > My Book: http://www.benwaggoner.com/books.htm > > Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html > > Compression Books: http://www.benwaggoner.com/bookshelf.htm > > > > > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technotes > > > > -- > Dr. Bernhard Grill email: > grl@iis.fhg.de > Head of Audio Department, FhG-IIS A phone: > +49 9131 776-351 > Am Weichselgarten 3, D-91058 Erlangen, Germany FAX: > +49 9131 776-398 > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From ben interframemedia.com Wed Oct 23 10:49:40 2002 From: ben interframemedia.com (Ben Waggoner) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] CELP test files or encoder? In-Reply-To: <3C124172E7FDD511B510000347426D5901324F2F@exchange.epr.com> Message-ID: on 10/23/02 9:21, Rob Koenen at rkoenen@intertrust.com wrote: >> The MPEG-4 reference software includes a CELP encoder which >> produces reasonable quality. > > I think Ben is looking for products, not source code, but he can > speak for himself. I'd really just be happy with a sample file at this point. As for source code, I don't mind as long as it'll compile under MacOS X 10.2. I don't have a complier on my Windows machines, and won't have my Linux system set up for a few more weeks. Given the popularity of MacOS for media folks, it'd be lovely if the reference software was tested for compiling under X. > Also, I have forwarded the request to a few parties that > may have or know about CELP technology. Thanks! I've already received some useful leads. Ben Waggoner Compressed Video Consulting, Training, and Encoding My Book: http://www.benwaggoner.com/books.htm Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html Compression Books: http://www.benwaggoner.com/bookshelf.htm From rkoenen intertrust.com Wed Oct 23 14:08:21 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:09:05 2003 Subject: [M4IF Technotes] scalability questions Message-ID: <3C124172E7FDD511B510000347426D5901324F52@exchange.epr.com> Technoters, There were some questions about scalability in the reference SW lately. I asked an expert, who told me: About the scalability tools in the reference software, my understanding is that spatial scalability and temporal scalability in the Simple Scalable Profile are implemented in the reference software, but may not be working properly with the rate control part in the reference software, namely, rate control may not work for the enhancement layer. The FGS and FGST tools work fine in the reference software and do not need rate control for the enhancement layer at the encoding time. A simple bitstream truncation program is included in the reference software package to cut an FGS bitstream to any bit rate that a user specifies. Since FGST is also a temporal scalability tool (using FGS bitplane coding for the DCT coefficients), only spatial scalability tool may not be working properly with rate control in the enhancement layer (fixed QP should be ok). Hope this is of some help. I am still trying to understand the status of the reference SW - what has been released by ISO and what hasn't. Will keep the list posted. Rob -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021023/08aa417a/attachment.html From Lefan mediaq.com Wed Oct 23 15:10:31 2002 From: Lefan mediaq.com (Lefan Zhong) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Reference software Version 2 cannot decode comformance visual testing stream Message-ID: Hello folks, Has any one try to use Reference software version 2 (especially microsoft-2.3-001213) to decode all conformance visual test stream (14496-4)? As I know the software cannot decode hit035.m4v. It even cannot finish the first I-VOP. Thanks, Lefan -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021023/9d0e234a/attachment.html From rkoenen intertrust.com Wed Oct 23 15:12:57 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Reference software Version 2 cannot decode c omformance visual testing stream Message-ID: <3C124172E7FDD511B510000347426D5901324F5E@exchange.epr.com> Lefan, thanks for the note. Some problems in the ref soft are known, and still being corrected. Some of the MPEG folks are listening on this list. For commercial products, fortunately, there is the rigorous interoperability testing in M4IF with now over 30 participants in the 4th round, and cross-forum testing with ISMA and others. Best, Rob -----Original Message----- From: Lefan Zhong [mailto:Lefan@mediaq.com] Sent: Wednesday, October 23, 2023 14:11 To: technotes@lists.m4if.org Subject: [M4IF Technotes] Reference software Version 2 cannot decode comformance visual testing stream Hello folks, Has any one try to use Reference software version 2 (especially microsoft-2.3-001213) to decode all conformance visual test stream (14496-4)? As I know the software cannot decode hit035.m4v. It even cannot finish the first I-VOP. Thanks, Lefan -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021023/036d1bc3/attachment.html From dsearles mmc.atmel.com Wed Oct 23 18:38:21 2002 From: dsearles mmc.atmel.com (dsearles) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Reference software Version 2 cannot decode comformance visual testing stream References: Message-ID: <3DB716CD.4020209@mmc.atmel.com> I worked on getting about 150 of the simple profile streams to work. There are various bugs that need to be fixed. Some as simple as uninitialized variables but others are much more complex. Regarding the "rigorous interoperability testing in M4IF with now over 30 participants in the 4th round, and cross-forum testing with ISMA and others." mentioned by Rob Koenen in a near by post, will that result in public code? Dan Searles Atmel Multimedia & Communications 3800 Gateway Centre, Suite 311 Morrisville, NC 27560 phone: (919)462-6553 fax: (919)462-03 Lefan Zhong wrote: > Hello folks, > > Has any one try to use Reference software version 2 (especially > microsoft-2.3-001213) to decode all conformance visual test stream > (14496-4)? As I know the software cannot decode hit035.m4v. It even > cannot finish the first I-VOP. > > Thanks, > > Lefan From rkoenen intertrust.com Wed Oct 23 16:00:19 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Reference software Version 2 cannot decode c omformance visual testing stream Message-ID: <3C124172E7FDD511B510000347426D5901324F69@exchange.epr.com> Dan, > Regarding the "rigorous interoperability testing in M4IF with > now over 30 participants in the 4th round, and cross-forum > testing with ISMA and others." > mentioned by Rob Koenen in a near by post, will that result > in public code? The main goal of the tests is getting interoperable *and* conformant products (we are not satisfied with products that are all interoperable but not conformant). Interop testing does *not* test the reference source code, nor does it produce any source code. Interop testing also makes only limited use of the conformance bitstreams, because the approach is to have n parties involved, where approximately n parties encode content and approximately n parties decode that, so that we get O(n x (n-1)) results. (Not all parties do both encoding *and* decoding) Whatever bugs are found in the standard are fed back into MPEG, and this should in return result in improvements of the public code as well as the bitstreams. Those bitstreams will be the basis for the certification program that M4IF has currently under development. Rob ps: Forgive the plug - but those listeners who are interested in interop testing: please consider joining M4IF. We are a low-threshold forum, and joining is straightforward. Info is here: http://www.m4if.org/m4if/ From Lefan mediaq.com Wed Oct 23 17:25:13 2002 From: Lefan mediaq.com (Lefan Zhong) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] AC prediction, follow document or follow reference software? Message-ID: Hello folks, In MPEG-4 document 14496-2, subclause 7.4.3.3, adaptive ac coefficient prediction, it says: If the prediction block (block 'A' or block 'C') is outside of the boundary of VOP or video packet, then all the prediction coefficients of that block are assumed to be zero. Beware that statement doesn't tell us what to do when the prediction block is 'inter'. Should we use the prediction block coefficients to do prediction or set to zeros when the prediction block is 'inter' type? In the comformance test bitstream Hit007.m4v, 1st P-VOP, the 2nd macroblock (coordinate (16,0)), the macroblock is an INTRA block, and AC prediction flag is on, and the 1st Y-block needs to do AC prediction from the left block. But left block is an INTER block, the AC coefficients are: -1 0 0 -1 0 0 0 0 1 -1 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Should we use 1024 1 0 0 0 0 0 0 to do prediction or use 1024 0 0 0 0 0 0 0 to do prediction? In the reference softwares (no matter momusys or microsoft, no matter version 1 or version 2), they all use zeros to do prediction in such cases. For scence of prediction, it is better to use zeros in such cases. But that is conflict to document. Thank you for your consideration. Lefan From mithun.nayak patni.com Thu Oct 24 15:32:09 2002 From: mithun.nayak patni.com (Mithun N) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Rate Control constraints In-Reply-To: <200210221016.g9MAGHJG001028@mx3.magma.ca> Message-ID: <000301c27b3c$09a297a0$fe02a8c0@patni.com> Hi all, Could anyone please suggest what steps do encoders normally take in order to meet the bitrate and fps constraints specified by the user. It seems that the microsoft encoder produces very high localized disortion to meet the requirements. Can anyone suggest a graceful degradation scheme. Thanks n regds Mithun N From garysull microsoft.com Thu Oct 24 03:16:57 2002 From: garysull microsoft.com (Gary Sullivan) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] AC prediction, follow document or follow reference software? Message-ID: <3B8749EA44FC2748956538D4A3096E42E08A98@RED-MSG-01.redmond.corp.microsoft.com> Lefan, That appears to be a bug in the MPEG-4 visual spec. I just looked too, and I couldn't find a clear statement that non-intra macroblocks are not used for AC coefficient prediction of Intra MBs, although I think it would be silly to use them. I don't recall whether that problem has been previously reported to MPEG or not. I will try to do so. Best Regards, Gary Sullivan +> -----Original Message----- +> From: Lefan Zhong [mailto:Lefan@mediaq.com] +> Sent: Wednesday, October 23, 2023 4:25 PM +> To: technotes@lists.m4if.org +> Subject: [M4IF Technotes] AC prediction, follow document or +> follow reference software? +> +> +> Hello folks, +> +> In MPEG-4 document 14496-2, subclause 7.4.3.3, adaptive ac +> coefficient prediction, it says: +> +> If the prediction block (block 'A' or block 'C') is outside +> of the boundary of VOP or video packet, then all the +> prediction coefficients of that block are assumed to be zero. +> +> +> Beware that statement doesn't tell us what to do when the +> prediction block is 'inter'. Should we use the prediction +> block coefficients to do prediction or set to zeros when the +> prediction block is 'inter' type? +> +> In the comformance test bitstream Hit007.m4v, 1st P-VOP, the +> 2nd macroblock (coordinate (16,0)), the macroblock is an +> INTRA block, and AC prediction flag is on, and the 1st +> Y-block needs to do AC prediction from the left block. But +> left block is an INTER block, the AC coefficients are: +> +> -1 0 0 -1 0 0 0 0 +> 1 -1 0 0 0 0 0 0 +> 0 -1 0 0 0 0 0 0 +> 0 0 0 0 0 0 0 0 +> 0 0 0 0 0 0 0 0 +> 0 0 0 0 0 0 0 0 +> 0 0 0 0 0 0 0 0 +> 0 0 0 0 0 0 0 0 +> +> Should we use +> 1024 +> 1 +> 0 +> 0 +> 0 +> 0 +> 0 +> 0 +> to do prediction or use +> 1024 +> 0 +> 0 +> 0 +> 0 +> 0 +> 0 +> 0 +> to do prediction? +> +> In the reference softwares (no matter momusys or microsoft, +> no matter version 1 or version 2), they all use zeros to do +> prediction in such cases. +> +> For scence of prediction, it is better to use zeros in such +> cases. But that is conflict to document. +> +> Thank you for your consideration. +> +> Lefan +> +> _______________________________________________ +> Technotes mailing list +> Technotes@lists.m4if.org +> http://lists.m4if.org/mailman/listinfo/technotes +> From salo rucus.ru.ac.za Fri Oct 25 05:44:49 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Streaming MPEG-4 live Message-ID: <000301c27bd0$7fd64cc0$9679e792@beast> Hi, I'm using the mpeg4ip mp4live server, and darwin streaming server on linux. Using windows, I can open my live stream using quicktime 6 with envivio, qt6 without envivio, but (strangely enough) not with realone w/ envivio, or win media player w/ envivio. But that's fine, at least I know that it is streaming properly. The problem comes when I want to inline that stream. Inline { URL [ "rtsp://myserver/sample_300kbit.mp4"] } works fine, but... Inline { URL [ "rtsp://myserver/mysdp"] } Doesn't work, despite that URL working in the player. Ideas? Thanks very much, Nicholas From salo rucus.ru.ac.za Fri Oct 25 07:10:24 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Streaming MPEG-4 live In-Reply-To: Message-ID: <000401c27bdc$73be7b40$9679e792@beast> Thanks Boris, But it didn't work - the players dont even attempt to make a network connection with inlined e-rtsp. Using rtsp they (qt and real) do make a connection for the duration of the clip, but don't display anything. In fact, after receiving about 1k of data, they (presumably) don't ask for any more, and don't get more. Using e-rtsp (as a normal "open url") did create a nice garbled green image in realone though... :) which is better than nothing. Who's job is it to parse these URL's - the player, the plug-in, or MPEG-4 itself? Does the player look at something like "rtsp" and decide it knows how to do that, get the data, and then pass it to the (MPEG-4) codec to decode? According to the MPEG-4 component, my sdp url is well formed right? Thanks, Nik > -----Original Message----- > From: Boris Felts [mailto:bfelts@sfo.envivio.com] > Sent: 25 October 2023 05:38 > To: Nicholas Smit > Subject: RE: [M4IF Technotes] Streaming MPEG-4 live > > > Try: > > Inline { URL [ "e-rtsp://myserver/mysdp"] } > > This may work... > You can also post more specific questions on our forum on the > support page. > > Best, > > Boris > > > -----Original Message----- > > From: Nicholas Smit [mailto:salo@rucus.ru.ac.za] > > Sent: Thursday, October 24, 2023 7:45 PM > > To: technotes@lists.m4if.org > > Subject: [M4IF Technotes] Streaming MPEG-4 live > > > > Hi, > > > > I'm using the mpeg4ip mp4live server, and darwin streaming > server on > > linux. > > > > Using windows, I can open my live stream using quicktime 6 with > envivio, > > qt6 without envivio, but (strangely enough) not with realone w/ > envivio, > > or win media player w/ envivio. > > > > But that's fine, at least I know that it is streaming properly. > > > > The problem comes when I want to inline that stream. > > > > Inline { URL [ "rtsp://myserver/sample_300kbit.mp4"] } > > > > works fine, but... > > > > Inline { URL [ "rtsp://myserver/mysdp"] } > > > > Doesn't work, despite that URL working in the player. > > > > Ideas? > > > > Thanks very much, > > > > Nicholas > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technotes > From sps iis.fhg.de Fri Oct 25 11:59:26 2002 From: sps iis.fhg.de (Ralph Sperschneider) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] CELP test files or encoder? References: Message-ID: <3DB907EE.7060606@iis.fhg.de> Hi Ben, you will find mp4 files with MPEG-4 CELP content on the conformance MPEG-4 Audio ftp server: ftp://mpaudconf:adif2mp4@ftp.iis.fhg.de/mpeg4audio-conformance/compressedMp4/ All files starting with 'ce' contain CELP data in various configurations. Beside this, the current MPEG-4 reference software (ISO/IEC 14496-5:2001) should be able to produce CELP content. Best regards, Ralph Ben Waggoner wrote: > Folks, > > All the MPEG-4 encoders I have my hands on only support AAC-LC audio, > not CELP. Does anyone have a MPEG-4 file with CELP audio I can use for some > tests, or know of an encoder that can make CELP audio? > > > Ben Waggoner > Compressed Video Consulting, Training, and Encoding > > My Book: http://www.benwaggoner.com/books.htm > Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html > Compression Books: http://www.benwaggoner.com/bookshelf.htm > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes -- Dipl.-Ing. Ralph Sperschneider | Phone: +49 9131 776 344 FhG IIS-A | Fax: +49 9131 776 398 Am Wolfsmantel 33 | mailto:sps@iis.fhg.de D 91058 Erlangen | http://www.iis.fhg.de/amm/ From bfelts sfo.envivio.com Fri Oct 25 12:06:08 2002 From: bfelts sfo.envivio.com (Boris Felts) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Streaming MPEG-4 live Message-ID: See answer on: http://ubb.envivio.com/6/ubb.x?a=tpc&s=31860025&f=942607201&m=5386059151 &r=6306042251#6306042251 Best, Boris > -----Original Message----- > From: Nicholas Smit [mailto:salo@rucus.ru.ac.za] > Sent: Thursday, October 24, 2023 9:10 PM > To: Boris Felts; technotes@lists.m4if.org > Subject: RE: [M4IF Technotes] Streaming MPEG-4 live > > Thanks Boris, > > But it didn't work - the players dont even attempt to make a network > connection with inlined e-rtsp. > > Using rtsp they (qt and real) do make a connection for the duration of > the clip, but don't display anything. In fact, after receiving about 1k > of data, they (presumably) don't ask for any more, and don't get more. > > Using e-rtsp (as a normal "open url") did create a nice garbled green > image in realone though... :) which is better than nothing. > > Who's job is it to parse these URL's - the player, the plug-in, or > MPEG-4 itself? Does the player look at something like "rtsp" and decide > it knows how to do that, get the data, and then pass it to the (MPEG-4) > codec to decode? > > According to the MPEG-4 component, my sdp url is well formed right? > > Thanks, > > Nik > > > > > -----Original Message----- > > From: Boris Felts [mailto:bfelts@sfo.envivio.com] > > Sent: 25 October 2023 05:38 > > To: Nicholas Smit > > Subject: RE: [M4IF Technotes] Streaming MPEG-4 live > > > > > > Try: > > > > Inline { URL [ "e-rtsp://myserver/mysdp"] } > > > > This may work... > > You can also post more specific questions on our forum on the > > support page. > > > > Best, > > > > Boris > > > > > -----Original Message----- > > > From: Nicholas Smit [mailto:salo@rucus.ru.ac.za] > > > Sent: Thursday, October 24, 2023 7:45 PM > > > To: technotes@lists.m4if.org > > > Subject: [M4IF Technotes] Streaming MPEG-4 live > > > > > > Hi, > > > > > > I'm using the mpeg4ip mp4live server, and darwin streaming > > server on > > > linux. > > > > > > Using windows, I can open my live stream using quicktime 6 with > > envivio, > > > qt6 without envivio, but (strangely enough) not with realone w/ > > envivio, > > > or win media player w/ envivio. > > > > > > But that's fine, at least I know that it is streaming properly. > > > > > > The problem comes when I want to inline that stream. > > > > > > Inline { URL [ "rtsp://myserver/sample_300kbit.mp4"] } > > > > > > works fine, but... > > > > > > Inline { URL [ "rtsp://myserver/mysdp"] } > > > > > > Doesn't work, despite that URL working in the player. > > > > > > Ideas? > > > > > > Thanks very much, > > > > > > Nicholas > > > > > > _______________________________________________ > > > Technotes mailing list > > > Technotes@lists.m4if.org > > > http://lists.m4if.org/mailman/listinfo/technotes > > From hbyu sjtu.edu.cn Sat Oct 26 16:54:10 2002 From: hbyu sjtu.edu.cn (Hongbin Yu) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] How to do streaming research using MPEG-4? Message-ID: <001801c27cc4$deeaf6d0$5f2350d3@yhb> Hi, all I am Hong-Bin Yu from China. I am considering doing some research about streaming media, but I am not aware how to simulate the network environment. Do I have to use some network simulation software such as OPNET or NS-2? Or I can find some simple and easier approach? Does MPEG offer any simple tools to do this job? Hong-Bin Yu -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20021026/010c723a/attachment.html From salo rucus.ru.ac.za Sun Oct 27 01:33:08 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:06 2003 Subject: [M4IF Technotes] Directly Call Conditional In-Reply-To: <3DB62DB4.1000107@enst.fr> Message-ID: <001301c27d48$0a5dec40$9679e792@beast> Hi experts. Is there any way to directly call a conditional? What I would like to do is something like : AT 2100 { CALL Media1GoSmall.activate } But of course I know one can only REPLACE, INSERT AND DELETE. As a work around at the moment I have: DEF TS1 TouchSensor { enabled false } DEF Go ROUTE TS1.enabled_changed to TS1.Media1GoBig.activate #"call" first event.... AT 1000 { REPLACE TS1.enabled by true } #insert next event.... AT 2000 { REPLACE Route Go by TS1.enabled_changed to Media1GoSmall.Activate Replace TS1.enabled by False } #trigger the event.... AT 2100 { REPLACE TS1.enabled by True } But I've noticed there seems to be a latency introduced by the event queue, hence the 2100. Sometimes the triggering event can be introduced as little as 10 milliseonds after the new event has been introduced, but othertimes the player will crash, or ignore it. 100ms seems to work fine though... But that's not great. There must be a better means to directly call a conditional...? Thanks! Nik. From christian.weigel epost.de Sun Oct 27 11:11:10 2002 From: christian.weigel epost.de (Christian Weigel) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Directly Call Conditional In-Reply-To: <001301c27d48$0a5dec40$9679e792@beast> Message-ID: <000001c27da1$2c7cac90$0500a8c0@christianpc> Hi Nicholas, Did you try this: AT 2100 { REPLACE Media1GoSmall.activate BY TRUE } I tried it and it worked. Regards Christian Nicholas Smit wrote: > Hi experts. > > Is there any way to directly call a conditional? > > What I would like to do is something like : > > AT 2100 { CALL Media1GoSmall.activate } > > But of course I know one can only REPLACE, INSERT AND DELETE. > > > > As a work around at the moment I have: > > DEF TS1 TouchSensor { enabled false } > DEF Go ROUTE TS1.enabled_changed to TS1.Media1GoBig.activate > > > #"call" first event.... > AT 1000 { > REPLACE TS1.enabled by true > } > > #insert next event.... > AT 2000 { > REPLACE Route Go by TS1.enabled_changed to > Media1GoSmall.Activate > Replace TS1.enabled by False > } > > #trigger the event.... > AT 2100 { > REPLACE TS1.enabled by True > } > > > But I've noticed there seems to be a latency introduced by > the event queue, hence the 2100. Sometimes the triggering > event can be introduced as little as 10 milliseonds after the > new event has been introduced, but othertimes the player will > crash, or ignore it. 100ms seems to work fine though... But > that's not great. > > There must be a better means to directly call a conditional...? > > Thanks! > > Nik. > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technot> es > From salo rucus.ru.ac.za Sun Oct 27 15:53:06 2002 From: salo rucus.ru.ac.za (Nicholas Smit) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Directly Call Conditional In-Reply-To: <000001c27da1$2c7cac90$0500a8c0@christianpc> Message-ID: <001901c27dc0$2fdcfd80$9679e792@beast> Works like a charm, thanks. > -----Original Message----- > From: technotes-admin@lists.m4if.org > [mailto:technotes-admin@lists.m4if.org] On Behalf Of Christian Weigel > Sent: 27 October 2023 12:11 > To: technotes@lists.m4if.org > Subject: RE: [M4IF Technotes] Directly Call Conditional > > > Hi Nicholas, > > Did you try this: > > AT 2100 { > REPLACE Media1GoSmall.activate BY TRUE > } > > I tried it and it worked. > > Regards > Christian > > > Nicholas Smit wrote: > > Hi experts. > > > > Is there any way to directly call a conditional? > > > > What I would like to do is something like : > > > > AT 2100 { CALL Media1GoSmall.activate } > > > > But of course I know one can only REPLACE, INSERT AND DELETE. > > > > > > > > As a work around at the moment I have: > > > > DEF TS1 TouchSensor { enabled false } > > DEF Go ROUTE TS1.enabled_changed to TS1.Media1GoBig.activate > > > > > > #"call" first event.... > > AT 1000 { > > REPLACE TS1.enabled by true > > } > > > > #insert next event.... > > AT 2000 { > > REPLACE Route Go by TS1.enabled_changed to > > Media1GoSmall.Activate > > Replace TS1.enabled by False > > } > > > > #trigger the event.... > > AT 2100 { > > REPLACE TS1.enabled by True > > } > > > > > > But I've noticed there seems to be a latency introduced by > > the event queue, hence the 2100. Sometimes the triggering > > event can be introduced as little as 10 milliseonds after the > > new event has been introduced, but othertimes the player will > > crash, or ignore it. 100ms seems to work fine though... But > > that's not great. > > > > There must be a better means to directly call a conditional...? > > > > Thanks! > > > > Nik. > > > > _______________________________________________ > > Technotes mailing list > > Technotes@lists.m4if.org > > http://lists.m4if.org/mailman/listinfo/technot> es > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From cloverleafland mail.hongkong.com Mon Oct 28 18:31:51 2002 From: cloverleafland mail.hongkong.com (cloverleafland@mail.hongkong.com) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need advice on Motion Estimation, please advise.... Message-ID: Dear all, I am currently working on a project to build a MPEG-4 encoder. I got confusion on the process of motion estimation. would anyone help explaining why the motion estimation is done on 16x16 block based but not in 8x8 block based? Also, for a macroblock, it consists of 4 8x8 Y blocks and 1 8x8 U block and 1 8x8 V block, would anyone advise whetehr the motion estimation on Y, U and V are done seperately or not? Thanks so much.... Regards, Clover ---------------------------------------------- \305w\252\357\250\317\245\316 hongkon.com \271q\244l\266l\245\363\250t\262\316 Thank you for using hongkong.com Email system From cloverleafland mail.hongkong.com Mon Oct 28 18:33:00 2002 From: cloverleafland mail.hongkong.com (cloverleafland@mail.hongkong.com) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need Advice on Motion Estimation for MPEG-4 encoder ... pls advise.... Message-ID: Dear all, I am currently working on a project to build a MPEG-4 encoder. I got confusion on the process of motion estimation. would anyone help explaining why the motion estimation is done on 16x16 block based but not in 8x8 block based? Also, for a macroblock, it consists of 4 8x8 Y blocks and 1 8x8 U block and 1 8x8 V block, would anyone advise whetehr the motion estimation on Y, U and V are done seperately or not? Thanks so much.... Regards, Clover ---------------------------------------------- \305w\252\357\250\317\245\316 hongkon.com \271q\244l\266l\245\363\250t\262\316 Thank you for using hongkong.com Email system From mithun.nayak patni.com Mon Oct 28 16:04:58 2002 From: mithun.nayak patni.com (Mithun N) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need advice on Motion Estimation, please advise.... In-Reply-To: Message-ID: <000a01c27e6d$aa5000a0$fe02a8c0@patni.com> Hi ... Motion estimation need not necessarily be done on 16 * 16 blocks. Infact if the search is performed on 4blocks of 8*8 then it is called 4MV searching. But that would consume a lot of time..searching for 4 blocks independently. Also ...motion estimation isnt carried out for chroma blocks..but the decoder assumes that the value of motion vector shall be 1) half that of luma vector (if search is for 16*16 block) 2) sum of non transparent luma vectors % num of non transparent luma blocks in a mcroblock (for 4 MV) The chrominance Motion vectors have to be rounded appropriately as suggested in ISO 14496-2.. regards Mithun Nayak -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of cloverleafland@mail.hongkong.com Sent: Monday, October 28, 2023 4:02 PM To: technotes@lists.m4if.org Subject: [M4IF Technotes] Re: Need advice on Motion Estimation, please advise.... Dear all, I am currently working on a project to build a MPEG-4 encoder. I got confusion on the process of motion estimation. would anyone help explaining why the motion estimation is done on 16x16 block based but not in 8x8 block based? Also, for a macroblock, it consists of 4 8x8 Y blocks and 1 8x8 U block and 1 8x8 V block, would anyone advise whetehr the motion estimation on Y, U and V are done seperately or not? Thanks so much.... Regards, Clover ---------------------------------------------- \305w\252\357\250\317\245\316 hongkon.com \271q\244l\266l\245\363\250t\262\316 Thank you for using hongkong.com Email system _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From cloverleafland mail.hongkong.com Mon Oct 28 21:29:35 2002 From: cloverleafland mail.hongkong.com (cloverleafland@mail.hongkong.com) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need advice on Motion Estimation, please advise.... Message-ID: Dear Mithun Nayak, Thanks for that... but I am still confused about the need of motion vector of chroma blocks... If we only need to perform motion estimation on Y blocks, then why do we still need to store the motion vectors for U & V blocks? Regards, Clover > Hi ... > > Motion estimation need not necessarily be done on > 16 * 16 blocks. > Infact if the search is performed on 4blocks of > 8*8 then it is called 4MV searching. > But that would consume a lot of time..searching > for 4 blocks independently. > > Also ...motion estimation isnt carried out for > chroma blocks..but the decoder assumes that > the value of motion vector shall be > 1) half that of luma vector (if search is for > 16*16 block) > 2) sum of non transparent luma vectors % num of > non transparent luma blocks in a mcroblock (for 4 > MV) > > The chrominance Motion vectors have to be rounded > appropriately as suggested in ISO 14496-2.. > > regards > Mithun Nayak > > > -----Original Message----- > From: technotes-admin@lists.m4if.org > [mailto:technotes-admin@lists.m4if.org]On Behalf > Of > cloverleafland@mail.hongkong.com > Sent: Monday, October 28, 2023 4:02 PM > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Re: Need advice on > Motion Estimation, please > advise.... > > > Dear all, > > I am currently working on a project to build a > MPEG-4 encoder. I got confusion on the process of > motion estimation. would anyone help explaining > why the motion estimation is done on 16x16 block > based but not in 8x8 block based? > > Also, for a macroblock, it consists of 4 8x8 Y > blocks and 1 8x8 U block and 1 8x8 V block, would > anyone advise whetehr the motion estimation on Y, > U and V are done seperately or not? > > Thanks so much.... > > Regards, > Clover > > > ---------------------------------------------- > \305w\252\357\250\317\245\316 hongkon.com > \271q\244l\266l\245\363\250t\262\316 > Thank you for using hongkong.com Email system > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > --------------------------------------------- 舧ㄏノHongKong.com秎ン╰参 Thank you for using hongkong.com Email system From chl math.uni-bonn.de Mon Oct 28 15:25:39 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need advice on Motion Estimation, please advise.... In-Reply-To: Message-ID: On Mon, 28 Oct 2023 cloverleafland@mail.hongkong.com wrote: > Dear Mithun Nayak, > > Thanks for that... but I am still confused about the need of motion vector of chroma blocks... > If we only need to perform motion estimation on Y blocks, then why do we > still need to store the motion vectors for U & V blocks? We don't. Either one 1 motion vector is saved (INTER) which describes the motion of whole 16x16 lumi block and it is used for both 8x8 chroma blocks, too. Or 4 motion vectors are saved (INTER4V) which describe motion of the 4 8x8 block in lumi and chroma vectors are calculated from those 4. No vectors is of course possible, too, and B-frames are also different, but this is the "ordinary" case. Christoph Lampert -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Zi. 15, 53115 Bonn, Germany | nell erstellt und bedarf Tel. (0228) 73-4708 Fax. +49 228 73-7916 | keiner Unterschrift. AZ 27B-6 From christian.weigel epost.de Mon Oct 28 18:11:54 2002 From: christian.weigel epost.de (Christian Weigel) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] URLs in PROTOS In-Reply-To: Message-ID: <000601c27ea5$1df19370$0500a8c0@christianpc> Hi, I'm having problems using a url in a PROTO. Assuming that the objectDescriptor with ID 2 contains a Jpeg the scene below should display the image. But it doesn't. If i give the OD_id directly in the PROTOs url field it works. I encoded it with both BIFSENC and mp4tool from ENST with the same result. Any hints? Christian --- PROTO testURL [ exposedField MFString theURL [] ] { Shape { geometry Bitmap {} appearance Appearance { texture ImageTexture { url IS theURL } } } } OrderedGroup { children [ DEF testInstance testURL { url ["2"] # alternatively I tried: "url 2" and "url "2" } ] } AT 0 { UPDATE OD [ { objectDescriptorID 2 muxScript .... } ] } --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.386 / Virus Database: 218 - Release Date: 09.09.2023 From mithun.nayak patni.com Tue Oct 29 10:18:46 2002 From: mithun.nayak patni.com (Mithun N) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Doubt on Padding of VOP Message-ID: <000801c27f06$780c5280$fe02a8c0@patni.com> Hi.. I have a certain doubt on the process of padding followed for MPEG4. After the horizontal and repetitive padding has been completed, we continue with the extended padding...which pads all the macroblocks inside the bounding rectangle of the VOP.(as stated in ISO 14496-2) Assuming this is the case and the search range is very small(f code =1 ) and the object has moved considerably between 2 time instants,and the reference vop falls outside the search range, and unrestricted motion vector is disabled ,how do we form the error macroblock??? Is it taken as ... error macroblock = current macroblock - zero(reference is zero since there is no search range ) ???????? Or is it that the whole frame has to be padded? If so how do we proceed with the padding of the macroblocks outside the VOP (between the bounding rectangle and the frame boundary)???? Please explain in detail....the reference microsoft decoder seems to pad the macroblocks outside the VOP. But there is no unrestricted motion vector parameter setting in the Microsoft encoder parameter file.. Thanks n Regards, Mithun Nayak From mithun.nayak patni.com Tue Oct 29 10:23:11 2002 From: mithun.nayak patni.com (Mithun N) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Re: Need Advice on Motion Estimation for MPEG-4 encoder ... pls advise.... In-Reply-To: Message-ID: <000901c27f07$15954d40$fe02a8c0@patni.com> Hi Mr clover Sorry for the delay in reply. the encoder need not calculate any motion vectors for the chroma blocks... the answer i gave was with respect to the decoder. the encoder will just perform motion vector calculation for the luminance macroblock/4 individual blocks. As it is there is no chrominance motion vector in the bitstream syntax.. Thanks, Mithun Nayak -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of cloverleafland@mail.hongkong.com Sent: Monday, October 28, 2023 4:03 PM To: technotes@lists.m4if.org Subject: [M4IF Technotes] Re: Need Advice on Motion Estimation for MPEG-4 encoder ... pls advise.... Dear all, I am currently working on a project to build a MPEG-4 encoder. I got confusion on the process of motion estimation. would anyone help explaining why the motion estimation is done on 16x16 block based but not in 8x8 block based? Also, for a macroblock, it consists of 4 8x8 Y blocks and 1 8x8 U block and 1 8x8 V block, would anyone advise whetehr the motion estimation on Y, U and V are done seperately or not? Thanks so much.... Regards, Clover ---------------------------------------------- \305w\252\357\250\317\245\316 hongkon.com \271q\244l\266l\245\363\250t\262\316 Thank you for using hongkong.com Email system _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From Cyril.Concolato enst.fr Tue Oct 29 08:59:15 2002 From: Cyril.Concolato enst.fr (Cyril Concolato) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] URLs in PROTOS References: <000601c27ea5$1df19370$0500a8c0@christianpc> Message-ID: <3DBE3FD3.55A1201E@enst.fr> Hi Christian, Since the proto field is an MFString you need to use a particular syntax to signal to then encoder that the string will actually be an odid. Use the following syntax "od:2" instead of "2" and it will work, at least with mp4tool. > PROTO testURL [ > exposedField MFString theURL [] > ] > { > Shape { > geometry Bitmap {} > appearance Appearance { > texture ImageTexture { > url IS theURL > } > } > } > } > OrderedGroup { > children [ > DEF testInstance testURL { > url ["2"] # alternatively I tried: "url 2" and "url "2" And of course the proto field name is not 'url' but 'theURL'. > } > ] > } Regards, Cyril Concolato -- Ecole Nationale Sup?rieure des T?l?communications, Paris Dept. Comelec 46, rue Barrault 75013 Paris Tel: +33145817991 Fax: +33145804036 From santosh.chapaneri patni.com Tue Oct 29 13:53:11 2002 From: santosh.chapaneri patni.com (Santosh Chapaneri) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] Doubt on Padding of VOP In-Reply-To: <000801c27f06$780c5280$fe02a8c0@patni.com> Message-ID: <000001c27f24$840f2350$8c03a8c0@pcp33248> Hi I have a similar kind of doubt regarding the padding in case of B-vop. In P VOPs, the problem (of ref vop lying outside the search range) can be solved by using intra macroblocks. But there is no scope of using Intra Macroblocks in a B VOP.(isn't it????) Can someone help me out on this??? thanks Santosh -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Mithun N Sent: Tuesday, October 29, 2023 10:19 AM To: technotes@lists.m4if.org Subject: [M4IF Technotes] Doubt on Padding of VOP Hi.. I have a certain doubt on the process of padding followed for MPEG4. After the horizontal and repetitive padding has been completed, we continue with the extended padding...which pads all the macroblocks inside the bounding rectangle of the VOP.(as stated in ISO 14496-2) Assuming this is the case and the search range is very small(f code =1 ) and the object has moved considerably between 2 time instants,and the reference vop falls outside the search range, and unrestricted motion vector is disabled ,how do we form the error macroblock??? Is it taken as ... error macroblock = current macroblock - zero(reference is zero since there is no search range ) ???????? Or is it that the whole frame has to be padded? If so how do we proceed with the padding of the macroblocks outside the VOP (between the bounding rectangle and the frame boundary)???? Please explain in detail....the reference microsoft decoder seems to pad the macroblocks outside the VOP. But there is no unrestricted motion vector parameter setting in the Microsoft encoder parameter file.. Thanks n Regards, Mithun Nayak _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From christian.weigel epost.de Tue Oct 29 10:29:11 2002 From: christian.weigel epost.de (Christian Weigel) Date: Wed Jul 30 14:09:07 2003 Subject: [M4IF Technotes] URLs in PROTOS In-Reply-To: <3DBE3FD3.55A1201E@enst.fr> Message-ID: <000301c27f2d$a3f259e0$0500a8c0@christianpc> Thank you. Works with mp4tool as well as BIFSENC. Christian Cyril Concolato wrote: > Hi Christian, > > Since the proto field is an MFString you need to use a > particular syntax to signal to then encoder that the string > will actually be an odid. Use the following syntax "od:2" > instead of "2" and it will work, at least with mp4tool. > > > PROTO testURL [ > > exposedField MFString theURL [] > > ] > > { > > Shape { > > geometry Bitmap {} > > appearance Appearance { > > texture ImageTexture { > > url IS theURL > > } > > } > > } > > } > > OrderedGroup { > > children [ > > DEF testInstance testURL { > > url ["2"] # alternatively I tried: "url 2" and "url "2" > And of course the proto field name is not 'url' but 'theURL'. > > } > > ] > > } > > Regards, > > Cyril Concolato > -- > Ecole Nationale Sup?rieure des T?l?communications, Paris > Dept. Comelec > 46, rue Barrault 75013 Paris > Tel: +33145817991 Fax: +33145804036 > > --- > Incoming mail is certified Virus Free. > Checked by AVG anti-virus system (http://www.grisoft.com). > Version: 6.0.408 / Virus Database: 230 - Release Date: 24.10.2023 > > From baumor iis.fhg.de Tue Oct 29 17:04:51 2002 From: baumor iis.fhg.de (Oliver Baum) Date: Wed Jul 30 14:09:08 2003 Subject: [M4IF Technotes] Sound2D: listening plane Message-ID: <3DBEB1A3.5090606@iis.fhg.de> Hi, I have a question concerning the Sound2D node: Refering to the Systems spec. (section 9.4.2.96.2) this node is intended for the auralisation of 2D-objects in the screen plane; i.e. the virtual sound sources can be moved only in the xy-plane (in a right-handed coordinate system, where z is the axis coming out of the screen pointing to the viewer/listener). Does this really make sense? I am asking this because the playback hardware - at least the ones available today - allows virtual sound sources to be positioned only in the xz-plane, which is the plane the (5.1- or 4.0-) speakers are placed. To move sources in the xz-plan thus the Sound or DirectiveSound node has to be used, which requires at least the "audio", "3D audio" or "complete" scene graph profile according to section 19.3.3 of the Systems spec. -- even if the other parts of the scene require only e.g. the "advanced 2D" or "complete 2D" profile. So wouldn't it be more handy to implement the Sound2D node's location field as coordinates in the xz-plane (in contradiction to the Systems spec.) or are these ideas completely wrong? Comments and hints are welcome! Greetings, -- Dipl.-Ing. Oliver Baum FhG IIS-A, Am Weichselgarten 3, D-91058 Erlangen, Germany dept. Audio & Multimedia phone: +49 (0) 91 31 / 7 76 - 39 51 mailto: baumor@iis.fhg.de From yossi neomagic.co.il Wed Oct 30 15:43:31 2002 From: yossi neomagic.co.il (Yossi Shain) Date: Wed Jul 30 14:09:08 2003 Subject: [M4IF Technotes] Downsample and Upsample filters In-Reply-To: <200210291704.g9TH4D4r014299@mx3.magma.ca> Message-ID: <006101c2801a$57173940$753da8c0@neomagic.co.il> Hello, I am trying to find information on downsample and upsample filters for YUV. The mpeg4 verification model (WG11/N3908) Section 2.2.2 "Filtering Process" provides a table of filter taps for downsampling. The references given are the document MPEG95/0322 and an ftp site 'drop.chips.ibm.com:Tampere/Contrib/m0896.zip' How can I locate these two sources? Does anyone have pointers to more reference material on upsampling and downsampling filters? Kind regards, Yossi Shain From chl math.uni-bonn.de Wed Oct 30 15:50:58 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:09:08 2003 Subject: [M4IF Technotes] Rounding of motion vectors for chroma? Message-ID: Hi, I have a question about rounding motion vectors, especially for chroma if quarterpel is active. I'll use a simple example. Consider a macro block in inter4v mode with luminance motion vectors mv[0] to mv[3] in halfpel or quarterpel resolution. How do I calculate the motion vector for chroma from them? I know of course that it's mv[0]+mv[1]+mv[2]+mv[3] / (2*4) but how exactly is rounding supposed to be done? What I read from the standard about this is as unclear as possible. Thank you in advance, Christoph -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Zi. 15, 53115 Bonn, Germany | nell erstellt und bedarf Tel. (0228) 73-4708 Fax. +49 228 73-7916 | keiner Unterschrift. AZ 27B-6 From michael.ditze c-lab.de Thu Oct 31 12:06:35 2002 From: michael.ditze c-lab.de (Michael Ditze) Date: Wed Jul 30 14:09:08 2003 Subject: [M4IF Technotes] WG: MPEG-4 Profiles and Scalability Message-ID: <001501c280cd$939f0b00$a751ea83@clab.de> Hello there, I have a question concerning MPEG-4 profiles and corresponding Codecs. Can someone name me some codecs that implement either the Simple Scalable or the Core Scalable Profile of MPEG-4? I am especially interested in Linux Open Source Projects and the scalability features in MPEG-4 to address network media with extremely low bandwidth, e.g. the Control Area Network (CAN). Furthermore, it would be interesting to know to which degree scalability features are implemented in other MPEG-4 Profiles. Kind Regards, Michael Ditze Michael Ditze C-LAB, 33094 Paderborn, Germany F?rstenallee 11, Office FU 304 Voice: +49-5251-60-6124, Fax: +49-5251-60-6065 Email: mailto:michael.ditze@c-lab.de From ben interframemedia.com Thu Oct 31 08:06:15 2002 From: ben interframemedia.com (Ben Waggoner) Date: Wed Jul 30 14:09:08 2003 Subject: [M4IF Technotes] WG: MPEG-4 Profiles and Scalability In-Reply-To: <001501c280cd$939f0b00$a751ea83@clab.de> Message-ID: Michael, The only vendor I know of that uses Simple Scalable in shipping products is PacketVideo. Scalability is a core competitive advantage of Windows Media and RealMedia over MPEG-4 today for projects streamed over the public internet. Ben Waggoner Compressed Video Consulting, Training, and Encoding My Book: http://www.benwaggoner.com/books.htm Cleaner Tutorial: http://www.saferseas.com/navseries/adclean.html Compression Books: http://www.benwaggoner.com/bookshelf.htm on 10/31/02 3:06, Michael Ditze at michael.ditze@c-lab.de wrote: > I have a question concerning MPEG-4 profiles and corresponding Codecs. > Can someone name me some codecs that implement either the Simple > Scalable or the Core Scalable Profile of MPEG-4? I am especially > interested in Linux Open Source Projects and the scalability features in > MPEG-4 to address network media with extremely low bandwidth, e.g. the > Control Area Network (CAN). Furthermore, it would be interesting to know > to which degree scalability features are implemented in other MPEG-4 > Profiles.