September 2008 - ffmpeg-devel

[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

[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

[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

[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

[PATCH] -fs parameter human friendly
by stefano.sabatini-lala＠poste.it 09 Jun '10

09 Jun '10

Hi all, addresses issue 551. It exports the av_strtod() function to eval.h and uses it in ffmpeg.c (minor bump missing for easying the maintainance of the patch). Regards. -- FFmpeg = Funny and Fiendish Meaningless Patchable Embarassing Guide

3 18

[PATCH] Implement PAFF in H.264
by heydowns＠borg.com 26 Mar '10

26 Mar '10

Attached is a patch that implements PAFF in H.264. I don't expect this to be accepted right away, due to size and complexity, but I want to get it out there for review and for people to try out and comment on. It is working stable for me on a number of samples, including ones that mix PAFF with progressive and/or MBAFF. Basic overview of how it works: 1. Fields are decoded and deinterlaced. 2. If non-paired fields are encountered, they are not output, but are used for reference if so marked. 3. Reference lists are kept as they were before, but now may also contain unmatched field pairs. Reference list construction splits field pairs out to fields as needed. There is a major shortcoming/known issue with this patch; it disables the error concealing calls for field pictures. This is because the error concealer needs slices to be contiguous regions of the picture buffers. I'd appreciate input on how best to augment the error concealer to handle interlaced slices (or if I am just missing something....) Also interested in hearing how this should be split into smaller patches (I know that at least the addition/changes to header documentation should come out). -Jeff

15 101

[PATCH] GSM-MS decoder and encoder
by mbardiaux＠mediaxim.be 01 Feb '10

01 Feb '10

-- Michel Bardiaux R&D Director T +32 [0] 2 790 29 41 F +32 [0] 2 790 29 02 E mailto:mbardiaux at mediaxim.be Mediaxim NV/SA Vorstlaan 191 Boulevard du Souverain Brussel 1160 Bruxelles http://www.mediaxim.com/

6 56

yuv2rgb_template.c uses MOVNTQ but not SFENCE?
by Reimar.Doeffinger＠stud.uni-karlsruhe.de 12 Jan '10

12 Jan '10

Can someone comment if this is right? It seems like a bug according to the documentation I have... Greetings, Reimar D?ffinger

3 5