January 2009 - ffmpeg-devel - lists.ffmpeg.org

[PATCH] fixup mpegts_get_pcr to be more reliable
by elupus＠ecce.se 14 Sep '11

14 Sep '11

Hi, Here are a few patches i created while trying to figure out why a file "locked" up in mpegts_get_pcr on seek. I don't seem to learn that patches to the mpegts demuxer have a tendency to get lost in the void. Anyway... 0001: The code for backwords pcr searching could never be reached, so remove it. 0002: Respect the maximum stream position that the function is allowed to process. 0003: Make sure the return value of url_seek is respected 0004: Reindent after the above 0005: Make mpegts_get_pcr resync to stream if the position it is asked to look for a timestamp isn't on a mpegts syncword. ---- The real issue I found out later is that this particlar stream doesn't contain any packets with pid = pcr_pid, so it reads the whole file to attempt to find a starting timestamp for av_gen_search. Are the pid's for the pcr supposed to be listed in PMT? I haven't checked this yet, but i suspect the given pid isn't available in there either. Maybe it should be reset to -1 if the pid isn't found? Regards Joakim

2 1

[PATCH] lavfi test pattern generator
by stefano.sabatini-lala＠poste.it 26 Jun '11

26 Jun '11

Hi all, useful for testing, it is based on the test pattern generator by Nicolas George: http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2007-October/036838.html basically I just stripped the audio signal generator and adapted few things to the lavfi framework, so most credit goes to Nicolas. And it would plenty cool to find some way to get ffmpeg/ffplay take as input such a generator, but I still can't see how to do that. What I can do now is to take as input /dev/null, then overlay the test_pattern over it, but this is rather ugly. Regards. -- FFmpeg = Freak and Fierce Mysterious Proud Erratic Gadget

7 24

[FWD] [PATCH] immediate decoding of dvd stills
by nicola_sabbi＠fastwebnet.it 24 Jun '11

24 Jun '11

Some time ago someone posted this small patch to permit decoding dvd still images without setting LOW_DELAY; unfortunately it seems the patch was forgotten Is it ok to commit? If not, what's the right way to proceed?

5 17

[Toy] MS Video 1 encoder
by kostya.shishkov＠gmail.com 04 May '11

04 May '11

Here's a result of sleepless day :). With a small patch to ELBG it was possible to implement simple 15-bit MSVC encoder. At least it can encode more or less recognizable video. Since I have no intentions nor time to polish it, feel free to use it as you like - finish, rework into RPZA encoder, add link to this post to "Interesting patches" on Multimedia Wiki and forget it, whatever. (I should be working on AAC encoder anyway)

5 5

[PATCH] DCA-in-WAV detection
by kostya.shishkov＠gmail.com 26 Jan '11

26 Jan '11

Here is my attempt on patch discussed on Febryary 2006 Thread called "[PATCH] Fix DTS in WAV file playback."

7 12

[PATCH] Apple Video Encoder (rpza)
by ffmpeg.php＠gmail.com 24 Jan '11

24 Jan '11

Attached is a patch to add support for Apple Video encoding (rpza). The quality settings are defined at compile time right now since I wasn't really sure how to map ffmpeg's numerous video options to the capabilities of rpza. It supports all rpza encoding opcodes but the 4-color block encoding doesn't try to do multiblock runs yet. -Todd

9 20

[PATCH]wmv3 encoder
by fortin＠nerim.net 23 Jan '11

23 Jan '11

Hi, The following patch adds a wmv3 encoder (i removed the vc1 encoder because it doesn't generate an elementary stream). I don't think i can split the patch without breaking compilation between two patches. Please comment (particularly the vc1enc.h include in mpegvideo.c). Thanks, Denis

5 6

[Toy] LucasArts SMUSH demuxer and decoder
by kostya.shishkov＠gmail.com 23 Jan '11

23 Jan '11

Again, for the record and not for review. I just took very old patch and extended it to decode palettized LucasArts FMVs from the games like Full Throttle, Curse of Monkey Island, The Dig, etc. I would be glad if somebody pick and finish it (there are minor glitches during decoding and fade in/fade out scenes are rendered as madly flipping frames instead; oh, and the code style is rather bad).

5 6

[PATCH] Chinese AVS encoder
by stefan.gehrer＠gmx.de 23 Jan '11

23 Jan '11

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, as announced before here is my encoder for Chinese AVS. Features that it does support: - - Intra macroblocks - - P_SKIP macroblocks - - P_16x16 macroblocks - - one or two reference frames - - constant quantizer encoding - - motion estimation using hexagon search and qpel refinement Features that it does not support: - - P_16x8, P_8x16 and P_8x8 macroblocks - - B macroblocks/frames - - rate control - - anything else you might think of To test it, here is a sample command line ./ffmpeg -i in.avi -vcodec cavs -strict -2 -cqp 30 -refs 2 -psnr out.avi - -strict -2 is required for the encoder's experimental nature, I have not thought much about its standard conformance, just that it decodes with my decoder. - -cqp n with n in the range 0 to 63 is required as only fixed quantizer encoding is supported - -refs 2 is optional, one reference frame would be the default - -psnr is optional Michael, regarding your question on cvslog (sorry for crossposting, I wasn't subscribed to cvslog at the time): >> Author: stefang >> Date: Sat Jul 7 09:14:58 2007 >> New Revision: 9518 >> >> Log: >> move dequantization into it's own inline function > >why inline? same question for the other inline functions >is there some speed gain from duplicating them in the object files >between encoder and decoder? To be honest I did not think much about object size, it just seemed to be a convenient way to share functions this way instead of having to create ff_cavs_foobar names and without creating function call overhead that was not there before. If you think object size would be a problem I can of course move some functions back from cavs.h to cavs.c. Stefan -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGocyhYpYOlpT3vNMRArJrAKC5aQQ40GEmegoLh8u0b0HCc0f60QCgzgc4 tMea5NiHbnlzxikJYvkl9Q4= =0R8q -----END PGP SIGNATURE-----

5 10

Proper parsing of DTS-HD MA streams
by robert.mcnamara＠gmail.com 23 Jan '11

23 Jan '11

Good evening, I am writing to revive a discussion on proper parsing of DTS-HD MA in dca_parser.c and mpegts.c. There were discussions of both issues in June/July of this year. However, it looks as though the discussion lost steam, someone went on vacation, adequate samples weren't provided in time, etc. Most recent Fox and Sony blu-ray disks are coming out with DTS-HD MA audio tracks, and it would be nice to at least be able to pull out the DTS core. I have uploaded a sample to upload.mplayerhq.hu as /DTS-HD-MA/bond_sample_dtshdma.m2ts. Two issues are at play here. The first is the fact that the mpegts.c parser is not able to parse the m2ts and find the DTS-HD track, a discussion begun (and ultimately dying) here: http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2008-June/048387.html The second issue is the fact that DTS-HD tracks are not playing properly in dca_parser.c because of an apparent variable frame size. Discussion here: http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2008-June/048498.html I have uploaded a sample, tested the former patch and found it to more or less work (although I understand it needs to be done a certain way to be committed) and have uploaded a sample as requested in that long-ago thread. I stand ready to provide other samples, samples of greater size, and otherwise do whatever I can to assist in solving this problem. Thanks a lot. Robert

4 6