From zhenzhongc hotmail.com Thu Aug 1 00:35:38 2002 From: zhenzhongc hotmail.com (Chen Zhenzhong) Date: Wed Jul 30 14:08:33 2003 Subject: [M4IF Technotes] Mask pictures (binary or shape images) problem References: <001401c2389b$21ea8860$66a8758c@GRAPHIC> Message-ID: Dear BianSH, To encode the arbitrarily shaped object, you are required to set proper parameters in the parameter files. In the MS reference software, please set the Alpha.Type=Bianry (or ShapeOnly for encoding the shape only) in the file example.par. The "seg" file has a format similar to YUV luminance file (no file header). It contains only two values, 0 or 255 where 0 indicate it does not belong to the object and 255 indicate it belongs to the object. You 'd better extract the mask picture frame by frame from the "seg" file before you transform it into bmp format. For example, you may save the extracted picture with the extend name ".raw", which can be opened with PhotoShop such that you may save it as a bmp file. Regards, Chen Zhenzhong ----- Original Message ----- From: BianSH To: technotes@lists.m4if.org Sent: Wednesday, July 31, 2023 10:04 PM Subject: [M4IF Technotes] Mask pictures (binary or shape images) problem I have a question about the Masks for MPEG-4 video sequences on M4IF's resource page. I download all .seg.zip files and unzip them, but i don't realize how to use the .seg files. How can I get the mask sequences (ex:bmp files) from the "seg" file? On the other hand, i wonder that where we can obtain the mask sequences on Internet to test our research? Grateful for your answers! -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020731/06d18044/attachment.html From alisson dsc.ufpb.br Thu Aug 1 10:29:16 2002 From: alisson dsc.ufpb.br (Alisson Brito) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Getting frame by frame Message-ID: <002001c23957$0d36ac20$bf4ba596@alissonpessoal> Hi, Is there any enconder that I can ask for it enconde frame by frame, instead of encode all the bitstream? Thanks, Alisson -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020801/320b021a/attachment.html From sg jhu.edu Thu Aug 1 14:24:47 2002 From: sg jhu.edu (Sachin S Gangaputra) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Quarter pel issues. Message-ID: Hi all, The MPEG4 documentation that I have describes "how" to implement quarter pel interpolation but fails to describe "why" we choose this 8-tap symmetric filter. Can someone point me towards references in this direction that suggest - using this filter is better than the alternative method of doing a simple linear interpolation. Further I'm not sure why one needs to perform the mirroring during motion compensation of a given 8x8 or 16x16 block. In compensating for the next block from the current block, we do have the whole decoded current frame - so then do we really need this mirroring - why not just use neighboring blocks in the interpolation. At the edges of the frame mirroring does make sense. Moreover during motion compensation at the encoder mirroring is never performed as one can interpolate the whole frame in the search of the right quarter pel motion vector. Shouldn't both sides be consistent in using or not using mirroring? Thanks, Sachin Gangaputra Graduate Student, Johns Hopkins Univ. Baltimore. From youngall ms8.hinet.net Sat Aug 3 13:58:57 2002 From: youngall ms8.hinet.net (Jacky Young) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] list delete Message-ID: <009801c23ab5$809168e0$579be33d@laptop> Please kindly delete my email address from your mailing address. My email address is youngall@ms8.hinet.net Best Regards, Jacky Young -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020803/e3c668a0/attachment.html From Christoph.Stadler dynapel.de Mon Aug 5 14:12:40 2002 From: Christoph.Stadler dynapel.de (Christoph Stadler) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Getting frame by frame References: <002001c23957$0d36ac20$bf4ba596@alissonpessoal> Message-ID: <3D4E5DA8.F9F108B@dynapel.de> > Alisson Brito wrote: > > Hi, > Is there any enconder that I can ask for it enconde frame by frame, > instead of encode all the bitstream? > Thanks, > Alisson Dear Alisson, our company (DynaPel Laboratories, Germany) has only recently announced its new MPEG-4 Video Encoder/Decoder Software Development Kits which offer MPEG-4 video in Simple and Advanced Simple Profile and the very feature you are looking for, i.e. encoding frame by frame. For a licensing quote please turn to our Sales Team (Sales@DynaPel.com). Best Regards, Christoph Stadler -------------------------------------------------------------- DynaPel Labs. GmbH christoph.stadler@dynapel.de Dr. Christoph Stadler Fraunhofer Strasse 9 Tel. +49 (0)89 962428-25 D-85737 Ismaning, GERMANY Fax. +49 (0)89 962428-90 From Michael.Ditze c-lab.de Tue Aug 6 10:40:09 2002 From: Michael.Ditze c-lab.de (Michael Ditze) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] MPEG-4 Questions on Scalability Message-ID: <3D4F7D59.6020509@c-lab.de> Hello! I have some questions with regard to scalability in MPEG-4. Can you anyone answer some of these questions? - Is there a difference between temporal and spatial scalability when it comes to interdependencies bewteen the base stream and the enhancement stream. I assume that the enhancement stream for temporal scalability depends on the base stream because of the applied Motion Compensation. Is that the same for spatial scalability? - In MPEG-2 you normally have fixed spacing in a Group of Pictures and between successive I-frames and successive reference frames (I- and P-frames) that may be represented by the M and N value. Would this be the same for MPEG-4? - Even if MPEG-4 is scalable, does that mean that all streams are sent in any case through the network where the decodes neglects the enhancement streams if it cannot decode them in time? Or is there a mechanism that allows to send Enhancement Streams only when the server knows that the client may decode the stream in time (e.g. though an Admission Control) - As the scene description (BIFS) and video streams are sent in different Elementry streams, is there a mechanism in MPEG-4 that ensures that the scene description is decoded before the respective streams. Hoes does that work? I hope you can answer some of my question as this really important stuff I need for my thesis. If you have sources where I can find further information with regard to my questions, I would be very grateful. Thanks for your efforts anf 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 stevenp pacificsatellite.com Tue Aug 6 20:09:30 2002 From: stevenp pacificsatellite.com (Steven Pang) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Microprocessor power References: <002001c23957$0d36ac20$bf4ba596@alissonpessoal> Message-ID: <002601c23d39$bd1d4f00$4c5645ca@PACIFIC> Hi, Anyone know the Microprocessor processing power requirement to do the MPEG4 decoding and support all the MPEG4 formats? For Digital Satellite / Cable Set Top Box platform? Best regards, Steven Pang. Product Engineer / Technology & MIS Department. Pacific Satellite International Ltd. Room 1101 China Merchants Tower, Shun Tak Centre, 168 Connaught Road Central, Hong Kong Tel (Direct) : 852 - 2965 3161, Fax: 852 - 2558 0406 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020806/bee30616/attachment.html From wjf NetworkXXIII.com Tue Aug 6 05:24:47 2002 From: wjf NetworkXXIII.com (William J. Fulco) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Microprocessor power In-Reply-To: <002601c23d39$bd1d4f00$4c5645ca@PACIFIC> Message-ID: I would be interested in seeing any answers to this question also. ++Bill William J. Fulco wjf@NetworkXXIII.com Logic: When you absolutely, positively have to refute every fallacy in the room. -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Steven Pang Sent: Tuesday, August 06, 2023 4:10 AM To: technotes@lists.m4if.org Subject: Re: [M4IF Technotes] Microprocessor power Hi, Anyone know the Microprocessor processing power requirement to do the MPEG4 decoding and support all the MPEG4 formats? For Digital Satellite / Cable Set Top Box platform? Best regards, Steven Pang. Product Engineer / Technology & MIS Department. Pacific Satellite International Ltd. Room 1101 China Merchants Tower, Shun Tak Centre, 168 Connaught Road Central, Hong Kong Tel (Direct) : 852 - 2965 3161, Fax: 852 - 2558 0406 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020806/09932639/attachment.html From Michael.Ditze c-lab.de Tue Aug 6 16:46:11 2002 From: Michael.Ditze c-lab.de (Michael Ditze) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] Transmission Order in MPEG-4? Message-ID: <3D4FD323.5060507@c-lab.de> Hello, I have another questions with reagard to the Transmission Order in MPEG-4. Is it similar to MPEG-2. i.e. if a VoP relies on a succeeding VoP for decoding, than the latter is sent before. E.g. If a B-VoP relies on a suceesing P VoP, the P-VoP is inserted in the Transmission Stream such as it is sent before the B-VoP. Thanks for your answers. Cheers, Michael -- 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 chl math.uni-bonn.de Wed Aug 7 11:49:35 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... Message-ID: Hi, I still have problem understanding the correct specifications for writing timecodes to the VOP header. First there is modulo_time_base, it consists of a number of 1s followed by a 0, one 1 for every second that passed since the last GOV header or last module_time_base of the previously decoded I-, S- or P-VOP. Am I right, that this does not mean elapsed "full seconds", but is some kind of "carry" flag, when the ticks-counter "overflows"? So if we talked minutes and hours, it would be 1 when the clock jumps from 0:59 to 1:00, right? In real life, the chances of this value ever getting 2 is very small, correct? And then, there is vop_time_increment: It has a value in [0,vop_time_increment_resolution) and the last comment on it is "The local time base in the units of seconds is recovered by dividing this value by the vop_time_increment_resolution." That's simply wrong, isn't it? You will always end up with 0 when dividing this value by vop_time_increment_resolution... but what _is_ really meant? Thank you for your help, Christoph -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 From chl math.uni-bonn.de Wed Aug 7 12:30:07 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... In-Reply-To: Message-ID: On Wed, 7 Aug 2002, Christoph Lampert wrote: > Hi, > > I still have problem understanding the correct specifications > for writing timecodes to the VOP header. > > First there is modulo_time_base, it consists of a number of > 1s followed by a 0, one 1 for every second that passed since the last > GOV header or last module_time_base of the previously decoded I-, S- or > P-VOP. sorry, a correction. It was supposed to be "previously displayed I,S,P", not "decoded". Christoph -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 From Michael.Ditze c-lab.de Wed Aug 7 14:21:05 2002 From: Michael.Ditze c-lab.de (Michael Ditze) Date: Wed Jul 30 14:08:34 2003 Subject: [M4IF Technotes] free VOP header fields for additional information? Message-ID: <3D5102A1.5000506@c-lab.de> Hello there, Once again a question concerning the header of a Video Object Plane. Are there any fields that are free for editing to put in additional information, e.g. Worst Case Execution time information for the decoder? Can you name these fields and how long they are? 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 arcin atsana.com Wed Aug 7 09:50:06 2002 From: arcin atsana.com (Arcin Bozkurt) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... In-Reply-To: Message-ID: <004701c23e10$f4e71920$d505000a@lumictech.com> I will be answering your questions in a different order and hopefully will be clear. I will write my interpretation and how I am doing it. You are right, modulo_time_base in normal circumstances will never be two. You can create a setup where it will be, but it would not be very realistic I believe. The only time modulo_time_base is non-zero, is when the real time completes a one full second from the previous time a modulo_time_base was non-zero. Assume the simple fixed rate case: 30 fps. And the vop_time_increment_resolution is 300. 300 means each tick is 1/300 seconds long and the vop_time_increment is given as number of ticks in this resolution. Your display times will be 0, 1/30, 2/30, 3/30, .... 29/30, 1, 31/30, etc. In this case the values for modulo_time_base will be 0, 0, (repeat 29 times), 1, 0 repeat 30 times), 1, 0 (repeat 30 tiems) and the Vop_time_increment will be 0 10, 20, ... 290, 0 (this corresponds to a "1" in modulo_time_base), 10, 20, ... 290, 0 , etc... The confusing part is "local time base" related parts I believe. The standard says: modulo_time_base: This value represents the local time base in one second resolution units (1000 milliseconds). vop_time_increment: The local time base in the units of seconds is recovered by dividing this value by the vop_time_increment_resolution. Therefore, local time base in resolution of seconds, is given by modulo_time_base (as an increment from the previous) and local time base in resolution of "1/vop_time_increment_resolution" is given by vop_time_increment. (as an increment from the last modulo_time_base) If you look at this "division" as a floating point operation you will get (increment/increment_resoultion) (in seconds) which is exactly what we want. I hope this clarifies a bit. Once again, this is my understanding. -arcin -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Christoph Lampert Sent: Wednesday, August 07, 2023 4:50 AM To: technotes@lists.m4if.org Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... Hi, I still have problem understanding the correct specifications for writing timecodes to the VOP header. First there is modulo_time_base, it consists of a number of 1s followed by a 0, one 1 for every second that passed since the last GOV header or last module_time_base of the previously decoded I-, S- or P-VOP. Am I right, that this does not mean elapsed "full seconds", but is some kind of "carry" flag, when the ticks-counter "overflows"? So if we talked minutes and hours, it would be 1 when the clock jumps from 0:59 to 1:00, right? In real life, the chances of this value ever getting 2 is very small, correct? And then, there is vop_time_increment: It has a value in [0,vop_time_increment_resolution) and the last comment on it is "The local time base in the units of seconds is recovered by dividing this value by the vop_time_increment_resolution." That's simply wrong, isn't it? You will always end up with 0 when dividing this value by vop_time_increment_resolution... but what _is_ really meant? Thank you for your help, Christoph -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From chl math.uni-bonn.de Wed Aug 7 17:52:41 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... In-Reply-To: <004701c23e10$f4e71920$d505000a@lumictech.com> Message-ID: Hello, thank you for your answer. I have one small additional clarification to ask for: On Wed, 7 Aug 2002, Arcin Bozkurt wrote: > I will be answering your questions in a different order and hopefully will > be clear. I will write my interpretation and how I am doing it. > > You are right, modulo_time_base in normal circumstances will never be two. > You can create a setup where it will be, but it would not be very realistic > I believe. > > The only time modulo_time_base is non-zero, is when the real time completes > a one full second from the previous time a modulo_time_base was non-zero. > > Assume the simple fixed rate case: 30 fps. And the > vop_time_increment_resolution is 300. 300 means each tick is 1/300 seconds > long and the vop_time_increment is given as number of ticks in this > resolution. > > Your display times will be 0, 1/30, 2/30, 3/30, .... 29/30, 1, 31/30, etc. > In this case the values for modulo_time_base will be > 0, 0, (repeat 29 times), 1, 0 repeat 30 times), 1, 0 (repeat 30 tiems) > and the Vop_time_increment will be > 0 10, 20, ... 290, 0 (this corresponds to a "1" in modulo_time_base), 10, > 20, ... 290, 0 , etc... That is what I expected, fine. However, this is just for I/P/S-VOPs, not for B-VOPs. Consider almost a sequence encoded with IBBPBBPBBP... and only 12fps to keep the lines shorter (vop_time_incr_res=120), I understand it would be (display order): I B B P B B P B B P B B P B B P 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 000 010 020 030 040 050 060 070 080 090 100 110 000 010 020 030 now with 11fps (vop_time_incr_res=110) things get complicated. Would it be: I B B P B B P B B P B B P B B P 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 000 010 020 030 040 050 060 070 080 090 100 000 010 020 030 040 because reference is always the last I/P/S-VOP, not a B-VOP? Then I would finally start getting the point... > The confusing part is "local time base" related parts I believe. The > standard says: > modulo_time_base: This value represents the local time base in one second > resolution units (1000 milliseconds). > vop_time_increment: The local time base in the units of seconds is recovered > by dividing this value by the vop_time_increment_resolution. > > Therefore, local time base in resolution of seconds, is given by > modulo_time_base (as an increment from the previous) and local time base in > resolution of "1/vop_time_increment_resolution" is given by > vop_time_increment. (as an increment from the last modulo_time_base) > > If you look at this "division" as a floating point operation you will get > (increment/increment_resoultion) (in seconds) which is exactly what we want. Of course, I didn't think about floating point calculation with integer variables. Okay, you are right, then. So it's rather a statement about time "units" than something deep about values... Yours, Christoph > Hi, > > I still have problem understanding the correct specifications > for writing timecodes to the VOP header. > > First there is modulo_time_base, it consists of a number of > 1s followed by a 0, one 1 for every second that passed since the last > GOV header or last module_time_base of the previously decoded I-, S- or > P-VOP. > > Am I right, that this does not mean elapsed "full seconds", but is some > kind of "carry" flag, when the ticks-counter "overflows"? > So if we talked minutes and hours, it would be 1 when the clock jumps from > 0:59 to 1:00, right? > In real life, the chances of this value ever getting 2 is very small, > correct? > > And then, there is vop_time_increment: > > It has a value in [0,vop_time_increment_resolution) and the last comment > on it is "The local time base in the units of seconds is recovered by > dividing this value by the vop_time_increment_resolution." > > That's simply wrong, isn't it? You will always end up with 0 when dividing > this value by vop_time_increment_resolution... but what _is_ really meant? > > Thank you for your help, > > Christoph > > -- > Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- > Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf > Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > > -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 From sachin smartyantra.com Thu Aug 8 15:53:47 2002 From: sachin smartyantra.com (Sachin Patil) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] Problem : Half Pixel ME Message-ID: <5.1.0.14.0.20020808143720.00a0edc0@192.168.1.3> Hello all, I have an integer pel search Motion Estimation algorithm implemented for my MPEG4 video encoder. Now I am trying to refine the search using half pixel resolution. I find that there is an increase in file size with half pel refined search. Can anyone tell me the advantages of using Half Pel search with respect to the following -- 1. I want to know if there can be an increase in the output file size. 2. Is half pel refinement used only for an improvement in quality ( quality measure used : PSNR ) ? 3. What is the increased computational complexity? Thankyou, Sachin. -- Sachin Patil. Smart Yantra Technologies Pvt. Ltd. Bangalore INDIA Phone : +91 80 363 4917, 363 4918 Email : sachin@smartyantra.com Website : www.smartyantra.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020808/8e897186/attachment.html From salo rucus.ru.ac.za Thu Aug 8 14:07:59 2002 From: salo rucus.ru.ac.za (Nik) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] Interactive within interactive Message-ID: <001b01c23ecb$dc29df10$9679e792@beast> Hi there, I understand the concept of abstracting out audio video objects, which can then be composed as you like, but could one of these objects in turn be *another* mpeg-4 file with its own interactivity etc.? So if we had a complex mpeg-4 file exhibiting interactivity (still not sure-is this likely to be advanced simple profile?), could we then interact with a similar file (same profile) embedded somewhere within it? Thanks for your help! Nik. From nico.oorts xmt.be Thu Aug 8 15:37:33 2002 From: nico.oorts xmt.be (Nico Oorts) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] multiple audio tracks into an mp4 file Message-ID: <3D52660D.2080409@xmt.be> Hi all, With the mp4tool (JC Dufourd), I try to make a .mp4 file which contains multiple audiotracks. I make a scene with an audiosource. This audiosource has an url to the OD which contains the elementary stream. After some time I replace the audiosource by another which has also an url to an OD and ES. The result is an .mp4 which contains the tracks (I can check this with some tools and size), but play only the first track. Some ideas why this does not work? Thx Nico Oorts Research team VRT From Peter.Haighton m4if.org Thu Aug 8 10:28:08 2002 From: Peter.Haighton m4if.org (Peter Haighton) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] Problem : Half Pixel ME In-Reply-To: <5.1.0.14.0.20020808143720.00a0edc0@192.168.1.3> Message-ID: Sachin, The answers to your questions are definitely dependant on implendation details. Depending on what the ME finds as it's best match will determine whether or not the file size will increase. This is the case also for full pixel ME since the best match that you obtain through for example SAD of macroblocks may not translate to the smallest VLC codes. Half pixel ME is designed to improve quality of the image but again that is dependant on many factors. For example, if you find that your ME technique consistently uses many more bits than full pixel ME, and you are compressing at a fix rate, you may find that the quality decreases. On the otherhand it is typical that higher quality is obtained both in PSNR and visually. Increased Complexity comes right down to a case by case basis on how the ME is built. There are a few ways of building in half pixel ME with very different complexities. Peter -- Peter Haighton VideoSpheres Inc. 84 Hines Road Kanata, Ontario Canada, K2K 3G3 Tel: (613) 270-9646 x3022 Fax: (613) 271-9442 email: peterh@VideoSpheres.com See http://www.m4if.org for the latest on MPEG-4 -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Sachin Patil Sent: Thursday, August 08, 2023 5:24 AM To: technotes@lists.m4if.org Subject: [M4IF Technotes] Problem : Half Pixel ME Hello all, I have an integer pel search Motion Estimation algorithm implemented for my MPEG4 video encoder. Now I am trying to refine the search using half pixel resolution. I find that there is an increase in file size with half pel refined search. Can anyone tell me the advantages of using Half Pel search with respect to the following -- 1. I want to know if there can be an increase in the output file size. 2. Is half pel refinement used only for an improvement in quality ( quality measure used : PSNR ) ? 3. What is the increased computational complexity? Thankyou, Sachin. -- Sachin Patil. Smart Yantra Technologies Pvt. Ltd. Bangalore INDIA Phone : +91 80 363 4917, 363 4918 Email : sachin@smartyantra.com Website : www.smartyantra.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020808/bb5b6cd8/attachment.html From chl math.uni-bonn.de Thu Aug 8 17:09:47 2002 From: chl math.uni-bonn.de (Christoph Lampert) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] Problem : Half Pixel ME In-Reply-To: <5.1.0.14.0.20020808143720.00a0edc0@192.168.1.3> Message-ID: Hi, I can answer you something, not as an expert but with some experience from the Simple Profile Open Source MPEG-4 Codec XviD (www.xvid.org): On Thu, 8 Aug 2002, Sachin Patil wrote: > Hello all, > I have an integer pel search Motion Estimation algorithm > implemented for my MPEG4 video encoder. Now I am trying to refine the > search using half pixel resolution. I find that there is an increase in > file size with half pel refined search. > > Can anyone tell me the advantages of using Half Pel search with respect to > the following -- > 1. I want to know if there can be an increase in the output file size. It is (of course) possible that the Half Pel motion vectors need more space in VLC than the Full Pel ones, and that the gain in prediction accuracy does not compensate for this. However, if you just "refine" motion vectors from a Full Pel search by a small amount, I consider this effect rather unlikely for "normal" input material. You might want to check if your implementation is correct. > 2. Is half pel refinement used only for an improvement in quality ( quality > measure used : PSNR ) ? For us it has both effects: Smaller bitrate at fixed quantization parameter, _and_ improvement in quality (more than 1dB gain in PSNR, often much more). > 3. What is the increased computational complexity? That depends only on your implementation. In our codec on i386 with SIMD optimization, speed with Half Pel usage is about two thirds of speed without (always remember it's Simple Profile). This is mostly because we interpolate the whole image, then perform ME without distriminating between Full Pel and Half Pel positions, and after that there is a simple +-1 refinement check in x and y of the so far best motion vector. Hope that helps, Christoph -- Christoph H. Lampert chl@math,uni-bonn,de | Diese Signature wurde maschi- Beringstr. 6, Raum 14 Tel. (0228) 73-2948 | nell erstellt und bedarf Sprechstunden: keine, aber meistens da | keiner Unterschrift. AZ 27B-6 From arcin atsana.com Thu Aug 8 12:12:20 2002 From: arcin atsana.com (Arcin Bozkurt) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] confusion about modulo_time_base, vop_time_increment, ... In-Reply-To: Message-ID: <003701c23eed$ff530f80$d505000a@lumictech.com> > -----Original Message----- > From: technotes-admin@lists.m4if.org > [mailto:technotes-admin@lists.m4if.org]On Behalf Of Christoph Lampert > Sent: Wednesday, August 07, 2023 10:53 AM > To: technotes@lists.m4if.org > Subject: RE: [M4IF Technotes] confusion about modulo_time_base, > vop_time_increment, ... > > > Hello, > > thank you for your answer. I have one small additional > clarification to > ask for: > > On Wed, 7 Aug 2002, Arcin Bozkurt wrote: > > I will be answering your questions in a different order and > hopefully will > > be clear. I will write my interpretation and how I am doing it. > > > > You are right, modulo_time_base in normal circumstances > will never be two. > > You can create a setup where it will be, but it would not > be very realistic > > I believe. > > > > The only time modulo_time_base is non-zero, is when the > real time completes > > a one full second from the previous time a modulo_time_base > was non-zero. > > > > Assume the simple fixed rate case: 30 fps. And the > > vop_time_increment_resolution is 300. 300 means each tick > is 1/300 seconds > > long and the vop_time_increment is given as number of ticks in this > > resolution. > > > > Your display times will be 0, 1/30, 2/30, 3/30, .... 29/30, > 1, 31/30, etc. > > In this case the values for modulo_time_base will be > > 0, 0, (repeat 29 times), 1, 0 repeat 30 times), 1, 0 > (repeat 30 tiems) > > and the Vop_time_increment will be > > 0 10, 20, ... 290, 0 (this corresponds to a "1" in > modulo_time_base), 10, > > 20, ... 290, 0 , etc... > > That is what I expected, fine. However, this is just for > I/P/S-VOPs, not > for B-VOPs. Consider almost a sequence encoded with IBBPBBPBBP... > and only 12fps to keep the lines shorter (vop_time_incr_res=120), > I understand it would be (display order): > > I B B P B B P B B P B B P B B P > 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 > 000 010 020 030 040 050 060 070 080 090 100 110 000 010 020 030 > > now with 11fps (vop_time_incr_res=110) things get complicated. > Would it be: > > I B B P B B P B B P B B P B B P > 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 > 000 010 020 030 040 050 060 070 080 090 100 000 010 020 030 040 ^^^ The Marked "B" is based on the P at 90. I would say what you wrote is correct. for the next one which is a P, it is also based on the same P at 90. From lcheng62 yahoo.com Sun Aug 11 20:24:13 2002 From: lcheng62 yahoo.com (Liang Cheng) Date: Wed Jul 30 14:08:35 2003 Subject: [M4IF Technotes] about FGS specification Message-ID: <20020812022413.6817.qmail@web21106.mail.yahoo.com> Hi, Where can I find the latest FGS specification? Is it in the latest 14496-2? Thanks, Liang __________________________________________________ Do You Yahoo!? HotJobs - Search Thousands of New Jobs http://www.hotjobs.com From rkoenen intertrust.com Sun Aug 11 21:42:34 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] about FGS specification Message-ID: <3C124172E7FDD511B510000347426D59011B3727@exchange.epr.com> FGS is in the following amendment: ISO/IEC 14496-2:2001/Amd 2:2002 (Streaming video profile) See http://www.iso.ch/iso/en/StandardsQueryFormHandler.StandardsQueryFormHandler ?languageCode=en&keyword=&lastSearch=false&isoNumber=14496&isoPartNumber=&IC S=&stageCode=&stageDate=&committee=ALL&subcommittee=&scope=CATALOGUE&sortOrd er=ISO make sure to remove any linebreaks. Or go to http://www.m4if.org/resources.php and look under "Obtaining the standard" Rob > -----Original Message----- > From: Liang Cheng [mailto:lcheng62@yahoo.com] > Sent: Sunday, August 11, 2023 19:24 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] about FGS specification > > > Hi, > > Where can I find the latest FGS specification? Is it > in the latest 14496-2? > > Thanks, > Liang > > __________________________________________________ > Do You Yahoo!? > HotJobs - Search Thousands of New Jobs > http://www.hotjobs.com > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From salo rucus.ru.ac.za Mon Aug 12 12:42:18 2002 From: salo rucus.ru.ac.za (Nik) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Live streaming BIFS commands Message-ID: <003801c241e4$8d4edd20$9679e792@beast> Hi there. I hope you can help me with this fairly high level question. In a live mpeg-4 streaming situation, what tool would one use to manipulate the scene description on the fly so as to remove/add more live streams etc.?Assume for example, one wished to add a live stream with a URL that was unknown at the start of transmission. Do these tools exist, or would they have to be integrated with the streaming server? Thanks for your time :) Nik. From piccarre elet.polimi.it Wed Aug 14 02:02:18 2002 From: piccarre elet.polimi.it (Luca Piccarreta) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Wavelet Test Sequences References: <32808FAD125F8A49A3145D4BB67ADC4024C35D@lanmhs50.rd.francetelecom.fr> Message-ID: <001901c2431d$79866a80$8d0ed8d4@bigclit> DMIF and Amendment 7Hi all. I tried my best but I'm really not able to decoder Envivio wavelet streams in the sequences on their site. Can anyone (may be them) tell me how to compose DecoderSpecificInfo and AU and feed them into the decoder? This is what I did: 1) Strip the VO_Start code, End codes, etc., just left the VTC (0xbe) start code 2) Write both into a file 3) Decode them with: -Decoder MS Version 1.0 -Decoder MS Version 2.0 w4711 (tried version 1 and version 2 syntax) -Decoder MS Version 2.0 w4866 (tried version 1 and version 2 syntax) Moreover. Anyone knows where the updated test sequences are? (I mean, the old ms-vfdis-v10-ba1... bs4... vt4, but with the updated syntax especially for shape and vtc). Thanks in advance. Luca Piccarreta Tilab -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020814/d4130751/attachment.html From manu.batura patni.com Wed Aug 14 10:23:01 2002 From: manu.batura patni.com (Manu V Batura) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder Message-ID: <000201c24346$162cc640$5402a8c0@patni.com> Hi, As per Part-4 of the standard, it seems that any video encoder that can generate a compliant bitstream would be compliant to the standard. However, is there aso a corresponding distortion measure that needs to be acheived? (something like RMSE etc). If yes, what are the thresholds that should be used for testing 'acceptable video quality'/'allowable video distortion'? Thanks & Regards, Manu V Batura From rkoenen intertrust.com Tue Aug 13 23:24:28 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder Message-ID: <3C124172E7FDD511B510000347426D59011B377F@exchange.epr.com> Indeed, any decoder that produces a valid bistream is compliant to the standard. There are deliberately no prescribed quality metrics, nor is there any prescribed algorithm to perform the encoding. This means there will be good encoders and bad ones... Best, Rob > -----Original Message----- > From: Manu V Batura [mailto:manu.batura@patni.com] > Sent: Tuesday, August 13, 2023 20:53 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Distortion Testing for Encoder > > > Hi, > > As per Part-4 of the standard, it seems that any video > encoder that can > generate a compliant bitstream would be > compliant to the standard. However, is there aso a > corresponding distortion > measure that needs to be acheived? > (something like RMSE etc). If yes, what are the thresholds > that should be > used for testing 'acceptable video quality'/'allowable video > distortion'? > > Thanks & Regards, > Manu V Batura > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From manu.batura patni.com Wed Aug 14 12:53:38 2002 From: manu.batura patni.com (Manu V Batura) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder In-Reply-To: <3C124172E7FDD511B510000347426D59011B377F@exchange.epr.com> Message-ID: <000901c2435b$21029ee0$5402a8c0@patni.com> Thanks Rob, But are there any objective tests for distortion that can be used for testing which encoders are better. If distortion is the competitive measure, there should be distortion thresholds available that may be used to call some encoders good and some bad. Thanks & Regards, Manu -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Rob Koenen Sent: Wednesday, August 14, 2023 10:54 AM To: 'manu.batura@patni.com'; technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Indeed, any decoder that produces a valid bistream is compliant to the standard. There are deliberately no prescribed quality metrics, nor is there any prescribed algorithm to perform the encoding. This means there will be good encoders and bad ones... Best, Rob > -----Original Message----- > From: Manu V Batura [mailto:manu.batura@patni.com] > Sent: Tuesday, August 13, 2023 20:53 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Distortion Testing for Encoder > > > Hi, > > As per Part-4 of the standard, it seems that any video > encoder that can > generate a compliant bitstream would be > compliant to the standard. However, is there aso a > corresponding distortion > measure that needs to be acheived? > (something like RMSE etc). If yes, what are the thresholds > that should be > used for testing 'acceptable video quality'/'allowable video > distortion'? > > Thanks & Regards, > Manu V Batura > > _______________________________________________ > 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 rkoenen intertrust.com Wed Aug 14 10:09:55 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder Message-ID: <3C124172E7FDD511B510000347426D59011B3788@exchange.epr.com> > But are there any objective tests for distortion that can be used > for testing which encoders are better. If distortion is the > competitive measure, there > should be distortion thresholds available that may be used to > call some encoders good and some bad. The point is that there are no encoding quality requirements in the MPEG-4 standard. Of course there are many things you can do that go beyond the standard. NB - Nothing beats human evaluation, and I have yet to see proof that anything non-human beats signal-to-noise ratios. (Interestingly, this point comes up every 2-3 months on this list.) Best Regards, Rob From jp skystream.com Wed Aug 14 15:35:37 2002 From: jp skystream.com (Jeyendran Balakrishnan) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder Message-ID: Try the JND Toolkit from David Sarnoff labs (http://www.sarnoff.com/internet_telecom/jndmetrix/index.asp) available for about US$ 9K, Windows SW, offline analysis. Or you can get the more pricey Tektronix PQA box (about $30K I think, but don't hold me to this), which uses the same Sarnoff code inside, but (I believe) works real-time and for streaming content, and probably adds other bells and whistles. This generally gives some objective measures of video quality, taking into account the nature of the human psycho-visual system. You can use this to compare quality of different encoders on the same input content. However, to decide if any one encoder is good enough on an absolute basis is subjective, and generally addressable only by having a (sufficiently large) bunch of people look at the reconstructed output and judge the quality. -Jeyendran -----Original Message----- From: Manu V Batura [mailto:manu.batura@patni.com] Sent: Tuesday, August 13, 2023 11:24 PM To: technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Thanks Rob, But are there any objective tests for distortion that can be used for testing which encoders are better. If distortion is the competitive measure, there should be distortion thresholds available that may be used to call some encoders good and some bad. Thanks & Regards, Manu -----Original Message----- From: technotes-admin@lists.m4if.org [mailto:technotes-admin@lists.m4if.org]On Behalf Of Rob Koenen Sent: Wednesday, August 14, 2023 10:54 AM To: 'manu.batura@patni.com'; technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Indeed, any decoder that produces a valid bistream is compliant to the standard. There are deliberately no prescribed quality metrics, nor is there any prescribed algorithm to perform the encoding. This means there will be good encoders and bad ones... Best, Rob > -----Original Message----- > From: Manu V Batura [mailto:manu.batura@patni.com] > Sent: Tuesday, August 13, 2023 20:53 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Distortion Testing for Encoder > > > Hi, > > As per Part-4 of the standard, it seems that any video > encoder that can > generate a compliant bitstream would be > compliant to the standard. However, is there aso a > corresponding distortion > measure that needs to be acheived? > (something like RMSE etc). If yes, what are the thresholds > that should be > used for testing 'acceptable video quality'/'allowable video > distortion'? > > Thanks & Regards, > Manu V Batura > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020814/52e61bd6/attachment.html From rkoenen intertrust.com Wed Aug 14 21:36:23 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] Distortion Testing for Encoder Message-ID: <3C124172E7FDD511B510000347426D59011B37A9@exchange.epr.com> Please allow me write a Word of Caution. While 'objective' measures may be starting to get better than SNR, and are highly useful for monitoring purposes and optimizing encoder quality (including making coding decisions), they do not allow one to choose between one coder and the other. Their correlation with human subjective quality judgement is still too low for such purposes. Kind Regards, Rob -----Original Message----- From: Jeyendran Balakrishnan [mailto:jp@skystream.com] Sent: Wednesday, August 14, 2023 14:36 To: 'manu.batura@patni.com'; technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Try the JND Toolkit from David Sarnoff labs ( http://www.sarnoff.com/internet_telecom/jndmetrix/index.asp ) available for about US$ 9K, Windows SW, offline analysis. Or you can get the more pricey Tektronix PQA box (about $30K I think, but don't hold me to this), which uses the same Sarnoff code inside, but (I believe) works real-time and for streaming content, and probably adds other bells and whistles. This generally gives some objective measures of video quality, taking into account the nature of the human psycho-visual system. You can use this to compare quality of different encoders on the same input content. However, to decide if any one encoder is good enough on an absolute basis is subjective, and generally addressable only by having a (sufficiently large) bunch of people look at the reconstructed output and judge the quality. -Jeyendran -----Original Message----- From: Manu V Batura [ mailto:manu.batura@patni.com ] Sent: Tuesday, August 13, 2023 11:24 PM To: technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Thanks Rob, But are there any objective tests for distortion that can be used for testing which encoders are better. If distortion is the competitive measure, there should be distortion thresholds available that may be used to call some encoders good and some bad. Thanks & Regards, Manu -----Original Message----- From: technotes-admin@lists.m4if.org [ mailto:technotes-admin@lists.m4if.org ]On Behalf Of Rob Koenen Sent: Wednesday, August 14, 2023 10:54 AM To: 'manu.batura@patni.com'; technotes@lists.m4if.org Subject: RE: [M4IF Technotes] Distortion Testing for Encoder Indeed, any decoder that produces a valid bistream is compliant to the standard. There are deliberately no prescribed quality metrics, nor is there any prescribed algorithm to perform the encoding. This means there will be good encoders and bad ones... Best, Rob > -----Original Message----- > From: Manu V Batura [ mailto:manu.batura@patni.com ] > Sent: Tuesday, August 13, 2023 20:53 > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Distortion Testing for Encoder > > > Hi, > > As per Part-4 of the standard, it seems that any video > encoder that can > generate a compliant bitstream would be > compliant to the standard. However, is there aso a > corresponding distortion > measure that needs to be acheived? > (something like RMSE etc). If yes, what are the thresholds > that should be > used for testing 'acceptable video quality'/'allowable video > distortion'? > > Thanks & Regards, > Manu V Batura > > _______________________________________________ > 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020814/8275cdd0/attachment.html From yi.li skyworksinc.com Mon Aug 19 20:18:24 2002 From: yi.li skyworksinc.com (Yi Li) Date: Wed Jul 30 14:08:36 2003 Subject: [M4IF Technotes] De-blocking/De-ringing Message-ID: <200208192319.g7JNJYWv017193@mx3.magma.ca> Hello, I have some question regarding the de-blocking and de-ringing filters. The MPEG-4 visual discussed briefly a method to reduce the computational cost of de-blocking/de-ringing. This is achieved by defining ringing and blocking semaphores. However, it is not quite clear to me how to use those semaphores since the descriptions are too short. Does any one have more detailed info on that? The spec also mentions a reference simply called M2723 (maybe this is a MPEG-4 spec ref number). Do anyone have a electronic copy of that document? Thanks Yi Li From rm technopark.lt Wed Aug 21 12:28:04 2002 From: rm technopark.lt (Remigijus) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? Message-ID: <001201c248f5$0d693020$3301a8c0@technopark.dom> Hi All I would like to ask: Is standard conform to put mp3 sound in a MPEG-4 stream: I've created mp4 file with mp4creator (MPEG4IP) with mp3 sound track inside. When I'm playing this file with QuickTime 6 It generates Error message: "File Contains invalid audio data: An audio Decoder Configuration descriptor corrupted. Expected ObjectType = Audio(0x40); found object type ^0 instead." Does it means that QuickTime doesn't supports mp3 sound inside mp4 or something else. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020821/031983b9/attachment.html From rchandpur tataelxsi.co.in Wed Aug 21 17:47:14 2002 From: rchandpur tataelxsi.co.in (RUTURAJ CHANDPUR ) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? References: <001201c248f5$0d693020$3301a8c0@technopark.dom> Message-ID: <014e01c24904$4df4f070$0314010a@tataelxsi.co.in> Hi, The objectTypeIndication for mp3 audio is 0x6B as per Table 8 of ISO/IEC 14496-1:2001(E). Hope this helps u. rgds, Ruturaj Anand Chandpur ----- Original Message ----- From: Remigijus To: technotes@lists.m4if.org Sent: Wednesday, August 21, 2023 2:58 PM Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? Hi All I would like to ask: Is standard conform to put mp3 sound in a MPEG-4 stream: I've created mp4 file with mp4creator (MPEG4IP) with mp3 sound track inside. When I'm playing this file with QuickTime 6 It generates Error message: "File Contains invalid audio data: An audio Decoder Configuration descriptor corrupted. Expected ObjectType = Audio(0x40); found object type ^0 instead." Does it means that QuickTime doesn't supports mp3 sound inside mp4 or something else. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020821/04bcd1e0/attachment.html From rkoenen intertrust.com Wed Aug 21 10:19:44 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MP EG-4 stream ? Message-ID: <3C124172E7FDD511B510000347426D59011B3873@exchange.epr.com> That's right. In addition, I'd like to note that while mp3 is recognized as a native object type at the systems level, there is currently no MPEG-4 Profile that includes mp3 object types, so it may not be supported in some or even many systems. Popular demand will probably decide where this will go. Best Regards, Rob -----Original Message----- From: RUTURAJ CHANDPUR [mailto:rchandpur@tataelxsi.co.in] Sent: Wednesday, August 21, 2023 4:17 To: Remigijus Cc: technotes@lists.m4if.org Subject: Re: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? Hi, The objectTypeIndication for mp3 audio is 0x6B as per Table 8 of ISO/IEC 14496-1:2001(E). Hope this helps u. rgds, Ruturaj Anand Chandpur ----- Original Message ----- From: Remigijus To: technotes@lists.m4if.org Sent: Wednesday, August 21, 2023 2:58 PM Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? Hi All I would like to ask: Is standard conform to put mp3 sound in a MPEG-4 stream: I've created mp4 file with mp4creator (MPEG4IP) with mp3 sound track inside. When I'm playing this file with QuickTime 6 It generates Error message: "File Contains invalid audio data: An audio Decoder Configuration descriptor corrupted. Expected ObjectType = Audio(0x40); found object type ^0 instead." Does it means that QuickTime doesn't supports mp3 sound inside mp4 or something else. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020821/b342d940/attachment.html From singer apple.com Wed Aug 21 10:46:39 2002 From: singer apple.com (Dave Singer) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] Is standard conform to put mp3 sound in a MPEG-4 stream ? In-Reply-To: <001201c248f5$0d693020$3301a8c0@technopark.dom> References: <001201c248f5$0d693020$3301a8c0@technopark.dom> Message-ID: At 11:28 +0200 8/21/02, Remigijus wrote: >Hi All > >I would like to ask: Is standard conform to put mp3 sound in a MPEG-4 stream: > >I've created mp4 file with mp4creator (MPEG4IP) with mp3 sound track >inside. When I'm playing this file with QuickTime 6 It generates >Error message: > > > >"File Contains invalid audio data: > >An audio Decoder Configuration descriptor corrupted. > >Expected ObjectType = Audio(0x40); found object type ^0 instead." > > > >Does it means that QuickTime doesn't supports mp3 sound inside >mp4 or something else. > > Well yes, it does mean that QT6 doesn't do layer 3 in MP4. QT6 supports a few profiles in MP4, and we currently don't cover the 'old' formats (MPEG1 and MPEG2) in an MP4 context. I'd be curious to know, however, if we have the same understanding of the structures which would say "this is layer3 audio". can you briefly summarize your understanding of the profile, level, streamtype, objecttype, decoderspecificinfo, and access unit format? -- David Singer Apple Computer/QuickTime -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020821/8edda840/attachment.html From peter mltc.com.tw Thu Aug 22 10:47:53 2002 From: peter mltc.com.tw (Peter Hsu) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] About short video header..... Message-ID: <005701c2497d$ee2a89a0$5001a8c0@mltc.com.tw> Hi All, The ISO reference code can't encode short video header bitstream. Where may I get short video header encoder and bitstream? Thanks. Peter Hsu From rchandpur tataelxsi.co.in Thu Aug 22 16:03:35 2002 From: rchandpur tataelxsi.co.in (RUTURAJ CHANDPUR ) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] MP4 hint track Message-ID: <007401c249be$fd174610$0314010a@tataelxsi.co.in> Hi, Can I create MP4 file containing only hint tracks(rtp hint), OD and BIFS track? If yes, what are the hint track settings? How can I test this MP4 file? Thanks in advance. Ruturaj Anand Chandpur Sr. Engineer, Tata Elxsi Ltd., www.tataelxsi.com From manu.batura patni.com Fri Aug 23 16:46:28 2002 From: manu.batura patni.com (Manu V Batura) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] MPEG-4 bitstreams for conformance tests Message-ID: <000201c24a8e$259c31c0$5402a8c0@patni.com> Hi, From gajananl yahoo.com Sat Aug 24 04:10:29 2002 From: gajananl yahoo.com (gajanan lonkar) Date: Wed Jul 30 14:08:37 2003 Subject: [M4IF Technotes] Doubts!! Message-ID: <20020824101029.6902.qmail@web13108.mail.yahoo.com> Respected Sir(s), My name is Gajanan Lonkar & i m doing my Final Year Engg in E&TC from C.O.E.P(INDIA) which is one of the most reputed colleges in the country. I am doing a Final Year project at CIRRUS LOGIC titled "MPEG4 SYSTEM PARSER".Before jumping to MPEG4 we were asked to go through the MPEG1 & MPEG2 ISO Standards.We have been asked to find the answer to 2 questions but after much deliberation (and searching on the net) we were not able to come up with a substantial solution . I would be grateful if u could throw some light on the matter. The questions are as follows 1) Why DTS is Requires Please note : in our discussion with our guide we gave a few reasons as to DTS for decoding instant and PTS for presentation instant but their argument was that we know the (I,P,B) picture order so we can decode them just by reordering algorithm. Is it related to BUFFER MANAGMENT? 2)What is VBV(Video Buffer Verifier) Where is it located, What parmeters govern its operation. Specifically the lower and higher threshold. PS-We think its operation is somehow related to the DTS concept. I hope that u will spare some of ur valuable time and enlighten us regarding the above queries. THANKING U IN ANTICIPATION Your's Sincerely Gajanan Lonkar PS- PLease note that the quesions are generic in manner and can be related to any MPEG version. __________________________________________________ Do You Yahoo!? Yahoo! Finance - Get real-time stock quotes http://finance.yahoo.com From jaganan hotmail.com Sat Aug 24 16:25:50 2002 From: jaganan hotmail.com (Gajanan) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Doubts!! Message-ID: Respected Sir(s), My name is Gajanan Lonkar & i m doing my Final Year Engg in E&TC from C.O.E.P(INDIA) which is one of the most reputed colleges in the country. I am doing a Final Year project at CIRRUS LOGIC titled "MPEG4 SYSTEM PARSER".Before jumping to MPEG4 we were asked to go through the MPEG1 & MPEG2 ISO Standards.We have been asked to find the answer to 2 questions but after much deliberation (and searching on the net) we were not able to come up with a substantial solution . I would be highly obliged if u could throw some light on the matter. The questions are as follows: 1) Why DTS is Requires Please note : in our discussion with our guide we gave a few reasons as to DTS for decoding instant and PTS for presentation instant but their argument was that we know the (I,P,B)picture order so we can decode them just by reordering. Is it related to BUFFER MANAGMENT? 2)What is VBV(Video Buffer Verifier) Where is it located, What parmeters govern its operation. Specifically the lower and higher threshold. PS-We think its operation is somehow related to the DTS concept. I hope that u will spare some of ur valuable time and enlighten us regarding the above queries. THANKING U IN ANTICIPATION Your's Sincerely Gajanan Lonkar PS- PLease note that the quesions are generic in manner and can be related to any MPEG version. -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020824/a52ec9f3/attachment.html From hanxy magima.com.cn Mon Aug 26 11:50:54 2002 From: hanxy magima.com.cn (hanxy) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? Message-ID: <000e01c24cab$65a18d00$b800000a@sh.magima.com> Hi, All, the description of "bval" in psychoacoustic model of MPEG-4 AAC is "the median bark value of the specific partion", I don't know the physical meaning of it. So, Can you do me a favor to explain it in detail? Thanks in advance. your secerely Speer Han -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020826/508dddb0/attachment.html From st018487 ait.ac.th Mon Aug 26 12:10:04 2002 From: st018487 ait.ac.th (Suthinee Thanapirom) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] (no subject) Message-ID: <006d01c24cb6$74d65e80$47129fcb@tc04> In VTC mode, what is the quantization step size or quantization matrix? How does arithmetic coding do?Where do it get the probability table from? Can anyone do me a favor by answering these questions. Thank you in advance for your help. Regards, Suthinee Email: st018487@ait.ac.th or suthinee_t@hotmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020826/51d46a0c/attachment.html From idimkovic nero.com Mon Aug 26 10:59:39 2002 From: idimkovic nero.com (Ivan Dimkovic) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? In-Reply-To: <000e01c24cab$65a18d00$b800000a@sh.magima.com> Message-ID: Hi, "BVAL" value represents mean value in bark scale of the specified partition - i.e. this is the frequency in bark units of the middle of the partition. For bark units, you need to consult some psychoacoustic books - I suggest to take a look at Zwicker's publications. Bark scale is non-linear frequency scale (and Hz is linear frequency scale) and Bark units correlate to bandpass frequencies of human hearing - there are several approximation formulas for conversion of Hertz (Hz) to Bark. Take a look at google for these. Best regards / Mit freundlichen Gr��?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 hanxy Sent: Monday, August 26, 2023 4:51 AM To: technotes@lists.m4if.org Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? Hi, All, the description of "bval" in psychoacoustic model of MPEG-4 AAC is "the median bark value of the specific partion", I don't know the physical meaning of it. So, Can you do me a favor to explain it in detail? Thanks in advance. your secerely Speer Han From hanxy magima.com.cn Mon Aug 26 18:46:35 2002 From: hanxy magima.com.cn (hanxy) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? References: Message-ID: <001e01c24ce5$778b8810$b800000a@sh.magima.com> ----- Original Message ----- From: "Ivan Dimkovic" To: "hanxy" ; Sent: Monday, August 26, 2023 3:59 PM Subject: RE: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? > Hi, > > "BVAL" value represents mean value in bark scale of the specified > partition - i.e. this is the frequency in bark units of the middle of the > partition. For bark units, you need to consult some psychoacoustic books - I > suggest to take a look at Zwicker's publications. > > Bark scale is non-linear frequency scale (and Hz is linear frequency scale) > and Bark units correlate to bandpass frequencies of human hearing - there > are several approximation formulas for conversion of Hertz (Hz) to Bark. > Take a look at google for these. > > > Best regards / Mit freundlichen Gr��?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 hanxy > Sent: Monday, August 26, 2023 4:51 AM > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model > in detail? > > > Hi, All, the description of "bval" in psychoacoustic model of MPEG-4 AAC is > "the median bark value of the specific partion", I don't know the physical > meaning of it. So, Can you do me a favor to explain it in detail? > > Thanks in advance. > > your secerely > Speer Han > > From hanxy magima.com.cn Mon Aug 26 18:47:33 2002 From: hanxy magima.com.cn (hanxy) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? References: Message-ID: <002301c24ce5$9a44ff30$b800000a@sh.magima.com> Thank a lot, I have calculated and get the median bark value of each partition! ----- Original Message ----- From: "Ivan Dimkovic" To: "hanxy" ; Sent: Monday, August 26, 2023 3:59 PM Subject: RE: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model in detail? > Hi, > > "BVAL" value represents mean value in bark scale of the specified > partition - i.e. this is the frequency in bark units of the middle of the > partition. For bark units, you need to consult some psychoacoustic books - I > suggest to take a look at Zwicker's publications. > > Bark scale is non-linear frequency scale (and Hz is linear frequency scale) > and Bark units correlate to bandpass frequencies of human hearing - there > are several approximation formulas for conversion of Hertz (Hz) to Bark. > Take a look at google for these. > > > Best regards / Mit freundlichen Gr��?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 hanxy > Sent: Monday, August 26, 2023 4:51 AM > To: technotes@lists.m4if.org > Subject: [M4IF Technotes] Can anyone explain "bval" in psychoacoustic model > in detail? > > > Hi, All, the description of "bval" in psychoacoustic model of MPEG-4 AAC is > "the median bark value of the specific partion", I don't know the physical > meaning of it. So, Can you do me a favor to explain it in detail? > > Thanks in advance. > > your secerely > Speer Han > > From sunx pollux.usc.edu Mon Aug 26 01:30:15 2002 From: sunx pollux.usc.edu (sunx) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] FGS parameter file example Message-ID: Dear MPEG-4 users, Can anybody send me one copy of the parameter file used for MPEG-4 FGS coding? I am confused by those parameters defined in the parameter file. Thanks, Xiaoming From sunx pollux.usc.edu Mon Aug 26 12:45:43 2002 From: sunx pollux.usc.edu (sunx) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Subscribe to mpeg.nist.gov Message-ID: Dear MPEG-4 experts, My group subscribed to mpeg.nist.gov before, but later my advisor decided to cancel the subscription. Can anyone tell me how to subscribe to mpeg.nist.gov and how much is the cost for one year subscription? Thanks, Xiaoming From rkoenen intertrust.com Mon Aug 26 12:51:51 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] Subscribe to mpeg.nist.gov Message-ID: <3C124172E7FDD511B510000347426D59011B38EA@exchange.epr.com> As I already mentioned in a personal mail, M4IF does not deal with those issues. Please contact Peter Schirling, the head of the US MPEG delegation, as originally suggessted. Kind Regards, Rob > -----Original Message----- > From: sunx [mailto:sunx@pollux.usc.edu] > Sent: Monday, August 26, 2023 11:46 > To: MPEG-4 Technotes > Subject: [M4IF Technotes] Subscribe to mpeg.nist.gov > > > > Dear MPEG-4 experts, > > My group subscribed to mpeg.nist.gov before, but later my > advisor decided > to cancel the subscription. Can anyone tell me how to subscribe to > mpeg.nist.gov and how much is the cost for one year subscription? > > Thanks, > > Xiaoming > > > > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From jwalant.desai wipro.com Tue Aug 27 16:17:37 2002 From: jwalant.desai wipro.com (Jwalant Sumantrai Desai) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] H.264 Conformance testing Message-ID: Skipped content of type multipart/alternative-------------- next part -------------- **************************Disclaimer************************************ Information contained in this E-MAIL being proprietary to Wipro Limited is 'privileged' and 'confidential' and intended for use only by the individual or entity to which it is addressed. You are notified that any use, copying or dissemination of the information contained in the E-MAIL in any manner whatsoever is strictly prohibited. *************************************************************************** From nakata mpeg.rcast.u-tokyo.ac.jp Tue Aug 27 23:24:17 2002 From: nakata mpeg.rcast.u-tokyo.ac.jp (Mayumi (Nakata) Koike) Date: Wed Jul 30 14:08:38 2003 Subject: [M4IF Technotes] MPEG-4 bitstreams for conformance tests In-Reply-To: <000201c24a8e$259c31c0$5402a8c0@patni.com> Message-ID: <4.3.2-J.20020827221918.05ee6960@po.mpeg.rcast.u-tokyo.ac.jp> Dear Manu, Hi, here's the answer for your question. Host Name po.mpeg.rcast.u-tokyo.ac.jp User ID mpegbs Password Mp4-1999 Then, enter "public_htm" folder to find bitstreams. Please use FTP software to download. Best, Mayumi KOIKE At 15:46 02/08/23 +0530, Manu V Batura wrote: >Hi, > > From where can one procure the bitstreams donated by MPEG-4 Platform >Verification Bitstream Development Project of Japan? >Are they in public domain? > >Thanks & Regards, >Manu > > >_______________________________________________ >Technotes mailing list >Technotes@lists.m4if.org >http://lists.m4if.org/mailman/listinfo/technotes From rkoenen intertrust.com Wed Aug 28 13:32:55 2002 From: rkoenen intertrust.com (Rob Koenen) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] H.264 Conformance testing Message-ID: <3C124172E7FDD511B510000347426D59011B396B@exchange.epr.com> Jwalant, The H.264 / MPEG-4 AVC standard itself isn't finalized yet. Conformance always comes later than the standard, usually by 6-9 months. Rob -----Original Message----- From: Jwalant Sumantrai Desai [mailto:jwalant.desai@wipro.com] Sent: Tuesday, August 27, 2023 2:48 To: technotes@lists.m4if.org Subject: [M4IF Technotes] H.264 Conformance testing Hi all, I wish to know if the conformance testing criteria for H.264 has been finalized? If Yes, it would be helpful if someone could provide some pointers to availability of the compliance testing standard and test bitstreams for H.264. If No, what is the likely date for availability of the same. Thanks and rgds Jwalant -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/mp4-tech/attachments/20020828/7daa5201/attachment.html From praveen smartyantra.com Thu Aug 29 10:48:27 2002 From: praveen smartyantra.com (Praveen Patil) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] Testing H.263 Encoder Message-ID: <3D6DA093.000006.01080@sindhu> Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: image/gif Size: 494 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20020829/68afe9dd/attachment.gif -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: image/jpeg Size: 1431 bytes Desc: not available Url : /pipermail/mp4-tech/attachments/20020829/68afe9dd/attachment.jpeg From byan cse.cuhk.edu.hk Thu Aug 29 21:32:55 2002 From: byan cse.cuhk.edu.hk (Yan Bo) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] about the NTT DoCoMo software Message-ID: <200208291232.g7TCWru21351@cse.cuhk.edu.hk> Dear all, Can anyone tell me where to download the NTT DoCoMo software, which can insert many kinds of errors in the compressed video bitstream? Thanks and Regards! Yan Bo From Jeyaprakash.S lntinfotech.com Thu Aug 29 18:22:26 2002 From: Jeyaprakash.S lntinfotech.com (Jeyaprakash.S@lntinfotech.com) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] CELP Encoder Message-ID: Hi, I am having a doubt in ISO/IEC 14496-3:2001(E) Annex 3.B MPEG-4 CELP encoder tools. the n_lpc_analysis value for 8 KHz is given as (sbfrm_size/80) Can you please confirm whether it is sbfrm_size or frame_size. Warm Regards, Jeyaprakash S Jeyaprakash.S Software Engineer( Communications and Embedded Systems )L&T Infotech Ltd. Prestige Commercial Complex 4th Floor #2,Church Street Bangalore - 560 001, INDIA Off : +91 080 532 3734 /3736 Extn : 350 From mylanchen yahoo.com Thu Aug 29 06:11:43 2002 From: mylanchen yahoo.com (xin chen) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] questions about mpeg-4 file format Message-ID: <20020829121143.47319.qmail@web12101.mail.yahoo.com> Hi All, I read the related part in ISO/IEC 14496-1, and quicktime file format, but they are all concentrate on the meta-data part, that is the structure of moov atom. I want to know more about how the samples are stored (in mdat atom, right). Where to get documents? And about the SL layer, since the SL packet hearder is composed according to the SLconfigdescriptor, which is stored in ESDescriptor, my question is: the SL packet hearder are generated only when the mpeg4 file is transmitted, or is it stored already in the .mp4 (Because on the figure 1 of ISO/IEC 14496-1, the terminal get a file from storage medium will need to go through deliver layer and sync layer for decoding too.)? I think it's a waste to store the same inf. twice. Thank you very much. __________________________________________________ Do You Yahoo!? Yahoo! Finance - Get real-time stock quotes http://finance.yahoo.com From mjacklin geneva-link.ch Thu Aug 29 17:40:42 2002 From: mjacklin geneva-link.ch (Martin Jacklin) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] Doing any demos at IBC? Message-ID: Hi This message is for anyone, member or non-member, who is showing MPEG-4 applications at IBC in a couple of weeks, who hasn't yet provided me with their IBC demo description,... (This also applies if you are presenting a paper) Could you send it to me for some free publicity in our MPEG-4 at IBC exhibit guide? I can fit 30 words. I need it yesterday, but am holding open the gate specially for you, until tomorrow. /\/\artin Jacklin Hypermedium (M4IF marketing support and media relations) martin.jacklin@m4if.org +41 79 291 1882 gsm/sms/fax From richard sorenson.com Thu Aug 29 10:10:28 2002 From: richard sorenson.com (Richard Shields) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] questions about mpeg-4 file format In-Reply-To: <20020829121143.47319.qmail@web12101.mail.yahoo.com> Message-ID: <73FC8F68-BB61-11D6-8C4A-0030657E211C@sorenson.com> Hi xin chen, The mdat atom can contain any data in any order you see fit. There is no "framing" data stored in the mdat and thus it has no structure. It is the meta data in the moov atom that gives the data in the mdat atom structure but only indirectly. Typically you would write video or audio samples consecutively in decode order into the mdat but there are no rules stating that you have too. You have both audio and video tracks it's advantageous to write them in an interleaved fashion. The moov atom will indicate to the player how to locate the samples. SL Layer... I'll let the more educated among us to answer this one :-) -Richard On Thursday, August 29, 2002, at 06:11 AM, xin chen wrote: > Hi All, > I read the related part in ISO/IEC 14496-1, and > quicktime file format, but they are all concentrate on > the meta-data part, that is the structure of moov > atom. I want to know more about how the samples are > stored (in mdat atom, right). Where to get documents? > > And about the SL layer, since the SL packet hearder is > composed according to the SLconfigdescriptor, which is > stored in ESDescriptor, my question is: the SL packet > hearder are generated only when the mpeg4 file is > transmitted, or is it stored already in the .mp4 > (Because on the figure 1 of ISO/IEC 14496-1, the > terminal get a file from storage medium will need to > go through deliver layer and sync layer for decoding > too.)? I think it's a waste to store the same inf. > twice. > > Thank you very much. > > __________________________________________________ > Do You Yahoo!? > Yahoo! Finance - Get real-time stock quotes > http://finance.yahoo.com > _______________________________________________ > Technotes mailing list > Technotes@lists.m4if.org > http://lists.m4if.org/mailman/listinfo/technotes > From emre.aksu nokia.com Thu Aug 29 19:04:37 2002 From: emre.aksu nokia.com (emre.aksu@nokia.com) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] Visual Object Sequence Start Code Message-ID: Hello, I'd like to ask a question related to the Visual Object Sequence: In an MPEG-4 visual bitstream, can there be more than one Visual Object Sequence Start Code coming one after another without having a Visual Object Sequence End Code to terminate the previous one, i.e. : visual_object_sequence_start_code ... ... visual_object_sequence_start_code ... ... visual_object_sequence_end_code Looking forward to receiving an answer, Best Regards, Emre Baris Aksu Nokia Corp. From peter mltc.com.tw Fri Aug 30 11:01:21 2002 From: peter mltc.com.tw (Peter Hsu) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] Visual Object Sequence Start Code References: Message-ID: <003401c24fc9$2421a240$5001a8c0@mltc.com.tw> Hi, Emre Baris Aksu, Yes, you can include more than one Visual Object Sequence Start Code before a Visual Object Sequence End Code. Peter Hsu MicroLinks Technology Corp. 21F, No.55, Chung Cheng 3rd RD., Kaohsiung, Taiwan 800, R.O.C. Tel: 886-7-2225678 Ext. 534 Fax: 886-7-2225675 E-mail: peter@mltc.com.tw ----- Original Message ----- From: To: Sent: Thursday, August 29, 2023 11:04 PM Subject: [M4IF Technotes] Visual Object Sequence Start Code Hello, I'd like to ask a question related to the Visual Object Sequence: In an MPEG-4 visual bitstream, can there be more than one Visual Object Sequence Start Code coming one after another without having a Visual Object Sequence End Code to terminate the previous one, i.e. : visual_object_sequence_start_code ... ... visual_object_sequence_start_code ... ... visual_object_sequence_end_code Looking forward to receiving an answer, Best Regards, Emre Baris Aksu Nokia Corp. _______________________________________________ Technotes mailing list Technotes@lists.m4if.org http://lists.m4if.org/mailman/listinfo/technotes From neel2265 yahoo.com Fri Aug 30 00:08:30 2002 From: neel2265 yahoo.com (Neelakanth) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] Access unit reordering Message-ID: <20020830060830.57811.qmail@web40206.mail.yahoo.com> Hi All I am going through the ISMA Implementation Specification Version 1.0 (dated:28 August 2001). In appendix A of the document it is explained that "To reconstruct the original order, the RTP time stamp and the AU-Index-delta are used." This is possible if all the Access Units (AU's) are of same time duration. I would be obliged to you, if you let me know the strategy for reordering the AU's on the Client side if the Access Units (AU's) are of variable time duration. Thanks Neelakanth __________________________________________________ Do You Yahoo!? Yahoo! Finance - Get real-time stock quotes http://finance.yahoo.com From singer apple.com Fri Aug 30 17:32:08 2002 From: singer apple.com (Dave Singer) Date: Wed Jul 30 14:08:39 2003 Subject: [M4IF Technotes] questions about mpeg-4 file format In-Reply-To: <73FC8F68-BB61-11D6-8C4A-0030657E211C@sorenson.com> References: <73FC8F68-BB61-11D6-8C4A-0030657E211C@sorenson.com> Message-ID: Richard has it perfect for the media data. The access units are stored in the mdat in the placesyou compute from the chunk offsets and sample sizes, and the sample sizes give you, um, the sizes. The SL headers are not stored in the file. There is a pre-defined value of the SLConfig (2) that is used in the file format. It is expected that any hinter/streamer that delivers the data using SL headers will replace this with an appropriate config that matches what they do. The pre-defined value means essentially 'you're in a file, the info that would normally be in SL headers is in the meta-data'. At 09:10 -0600 8/29/02, Richard Shields wrote: >Hi xin chen, > >The mdat atom can contain any data in any order you see fit. There >is no "framing" data stored in the mdat and thus it has no >structure. It is the meta data in the moov atom that gives the data >in the mdat atom structure but only indirectly. Typically you would >write video or audio samples consecutively in decode order into the >mdat but there are no rules stating that you have too. You have >both audio and video tracks it's advantageous to write them in an >interleaved fashion. The moov atom will indicate to the player how >to locate the samples. > >SL Layer... I'll let the more educated among us to answer this one :-) > >-Richard > >On Thursday, August 29, 2002, at 06:11 AM, xin chen wrote: > >>Hi All, >>I read the related part in ISO/IEC 14496-1, and >>quicktime file format, but they are all concentrate on >>the meta-data part, that is the structure of moov >>atom. I want to know more about how the samples are >>stored (in mdat atom, right). Where to get documents? >> >>And about the SL layer, since the SL packet hearder is >>composed according to the SLconfigdescriptor, which is >>stored in ESDescriptor, my question is: the SL packet >>hearder are generated only when the mpeg4 file is >>transmitted, or is it stored already in the .mp4 >>(Because on the figure 1 of ISO/IEC 14496-1, the >>terminal get a file from storage medium will need to >>go through deliver layer and sync layer for decoding >>too.)? I think it's a waste to store the same inf. >>twice. >> >>Thank you very much. >> >>__________________________________________________ >>Do You Yahoo!? >>Yahoo! Finance - Get real-time stock quotes >>http://finance.yahoo.com >>_______________________________________________ >>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 -- David Singer Apple Computer/QuickTime