similar to: Understanding the Theora Bitstream

Hi, Sorry for taking so long with this, but I finally committed a first pass of my own VP3 decoder to the ffmpeg codebase. You can browse the source here: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/ffmpeg/ffmpeg/libavcodec/ vp3.c and vp3data.h are the files in question. Attached is a message I sent to the ffmpeg-devel list about the new decoder and its present shortcomings.

Hi, Some of you may recall that I got the VP3 decoder working with the xine multimedia player some time ago and posted a patch for anyone else to try. I really messed up when I did that since I generated the patch against some random CVS snapshot of xine-lib. I did better this time and generated the patch against the latest release of xine-lib (1.0-beta4, released today). If you

Hi, Since a number of people are eager to try out the native VP3 decoder in xine, I have made a patch available, bz2-compressed: http://www.pcisys.net/~melanson/codecs/xine-vp3-decoder-patch.bz2 To work with this patch: 1) Get the latest xine-lib from CVS. Visit this page for more details: http://sourceforge.net/cvs/?group_id=9655 The module name is xine-lib, but you will also need

Hi, So it took a little while to get here but VP3 through ffmpeg is finally starting to achieve viewable results. It partially works right now. Keyframes only. And monochrome. Oh, and the picture is upside down (something I did not realize about VP3 until I let it fly just now). Anyway, here is a picture of ffmpeg's VP3 decoder in action using xine:

[standard disclaimer about my mail format] in ftp.vp3.com/theora user: vp3 pass: vp3dev theora_dbm_5-28.zip this implements a bitstream change; the header now contains a compressed huffman tree rather than the frequency counts (as discussed) pardon my inability to use diff correctly. The change in toplevel.c is trivial (new function names & params) huffman.c is the important one. I

./configure: line 524: syntax error near unexpected token `AM_INIT_AUTOMAKE(libtheora,0.0)' ./configure: line 524: `AM_INIT_AUTOMAKE(libtheora,0.0)' I'm sure I'm doing something pro-stupid, but --- my setup works fine for ogg & vorbis. autoconf ver 2.13 automake (GNU automake) 1.4-p4 if that matters hey let's put tool version numbers into the README -dan

Hi, Check it out-- another update in the space of a week: http://home.pcisys.net/~melanson/codecs/vp3-format.txt v0.4: March 2, 2004 - renamed and expanded section "Initializing The Quantization Matrices" - outlined section "Reconstructing The Frame" - moved Theora Differences Appendix to its own section entitled "Theora Specification" - added Appendix:

Hi, I have been on a documentation kick recently. I finally updated my VP3 format document tonight: http://www.pcisys.net/~melanson/codecs/vp3-format.txt v0.2: October 9, 2003 - expanded section "Reversing the DC Prediction" - added Appendix B: Theora Differences Feel free to look it over and point out any glaring mistakes. I can't wait to see how much sense it makes to

Hi, I don't know how much of the old VP3 codebase is incorporated into Theora. I wrote ffmpeg's VP3 decoder from scratch but eventually ported in the original VP3 IDCT function. A BSD conflict cropped up that I thought I would pass along. -- -Mike Melanson <p>---------- Forwarded message ---------- Date: Wed, 10 Sep 2003 13:06:37 -0700 (PDT) From: Steven M. Schultz

the files produced by windows compile vs. Linux are indeed significantly different in size (about 140K vs. 160) However I'm not convinced all the default parameters are set the same way in the two example files. This needs to be verified to see if there really is a bug, or can the two versions produce byte-equal output streams? <p> ___ Dan Miller (++,) Founder, On2 Technologies

Hi, Can anyone tell me the significant differences between VP30 and VP31 (or are you allowed to)? Are there different token and quantizer tables? Are there algorithmic differences? I have a VP30-encoded file that plays with the VP3 VfW DLL but not with the open source decoder. Thanks... -- -Mike Melanson --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project

Hi, Quick note: In case anyone has not heard, Nullsoft, the people behind Winamp, have seen fit to develop their own multimedia container format. It can hold any kind of data, but their primary codecs are MP3 for audio and VP3 for video: http://www.nullsoft.com/nsv/ I and a few other people are still working out the details of the format:

Hi, I am making good progress on understanding, documenting, and re-implementing VP3. I have found some more deadwood in the decoder code and I just wanted to verify that it really has no place: * Version 3.3 tables: There are 2 sets of VLCs: One for v3.1 and one for v3.3. The v3.1 set is always selected. Are the v3.3 tables in production, or will they ever be? * DC Search Points: VP3

Hi, Is there anything special I need to know about VP3's IDCT? I mean besides the fact that there are separate IDCTs to handle sparse coefficient matrices. Are the IDCT functions mathematically equivalent to any textbook IDCT functions? Thanks... -- -Mike Melanson --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To

Dan, In reviewing the code, derf came across the DCSearchPoints array in QuadCodeDisplayFragments(), encode.c:696. It looks like it specifies a search order for nearby coded fragments for prediction if there isn't a nearest neighbor. However, the upper search bound is defined as *0*, so the search loop (line 917) is never executed and it always falls back to the last coded fragment.

Hi, So I have almost completed a new VP3 decoder implementation. In the course of doing so, I have encountered something odd about macroblock coding. As a quick overview, VP3 has a notion of fragments (8x8 pixels) and superblocks (32x32 pixels, 4x4 fragments). These apply to each individual plane (e.g., a 64x64 video will have 4 Y superblocks and 1+1 C superblocks, 64 Y fragments

Hi, There seem to be 2 sets of block coding patterns in the VP3x/Theora algorithm. Based on the comments in the code, the first pattern is used if encoder version < 6 and the second pattern is used in encoder version >= 6. None of the code to support the second pattern is ever called. Is there an encoder with a version >= 6? The code in question is

I try to merge VP3's mmx decoder into theora. http://kyoto.cool.ne.jp/vp3/developers/theora-alpha3-MMXd-src.zip You can see the change by searching the keyword "_UsingMMX_" in all lib folder's file. From VP3???YO vp3@go8.enjoy.ne.jp

I hope this list is small enough for not beeing banned when introducing oneself. So... Hi, I just subscribed to the theora-lists. I'm from the XVID project, so I know a deal about video and MPEG-4 in particular, but nothing about Vorbis/Theora etc., That's why I'm here, because I'm interested in patent-free alternatives. I guess there is no documentation of the current status

Hi, I have been working with VP3 for some time now. Dan Miller from On2 just told me about this list. A few months ago, I managed to get On2's VP3 source snapshot to compile under Linux, a feat that many have apparently attempted and quickly abandoned. I compiled a small console app and the results of the experiment are explained here: