February 2010 - ffmpeg-devel - lists.ffmpeg.org

Fwd: framebuffer device demuxer
by giliarde＠gmail.com 12 Mar '11

12 Mar '11

Hi, patch attached. Giliard wrote: > Dear developer, > > Recently I added a extension in the FFmpeg for grab the screen content > of devices which use framebuffer ( see http://www.directfb.org/ ). > I implemented this as a framebuffer device demuxer in the directory > "libavdevice". > I needed modify the files: libavdevice/alldevices.c, > libavdevice/Makefile and configure. > All similar to x11grab. > > If you has interest for included in the FFmpeg source code official, the > attachment contain the files modified and implemented. > First, thanks a lot for your work. Would it be possible for you to send a patch ? (diff -wup) of your tree and send it attached in ffmpeg-devel mailing list ? Thanks a lot. -- Baptiste COUDURIER GnuPG Key Id: 0x5C1ABAAA Key fingerprint 8D77134D20CC9220201FC5DB0AC9325C5C1ABAAA FFmpeg maintainer http://www.ffmpeg.org

9 39

Small memleak when use AAC codec.
by igor＠plusmobile.ru 11 Mar '11

11 Mar '11

I found small memory leak when using aac. This leak is 32 bytes per loop without MEMALIGN_HACK, and 48 bytes with MEMALIGN. Simple code: #include <iostream> extern "C" { #include "libavformat/avformat.h" #include "libavcodec/avcodec.h" } AVOutputFormat *OutFormat; AVFormatContext *FormatCtx; AVCodecContext *VideoCodecCtx, *AudioCodecCtx; AVCodec *VideoCodec, *AudioCodec; AVStream *VideoStream, *AudioStream; int main (int argc, char * const argv[]) { av_register_all(); while(1) { FormatCtx = avformat_alloc_context(); OutFormat = guess_format(NULL, "test.3gp", NULL); FormatCtx->oformat = OutFormat; strncpy(FormatCtx->filename, "test.3gp", sizeof(FormatCtx->filename)); VideoStream = av_new_stream(FormatCtx, 0); VideoCodecCtx = VideoStream->codec; VideoCodecCtx->codec_id = CODEC_ID_H263; VideoCodecCtx->codec_type = CODEC_TYPE_VIDEO; VideoCodecCtx->bit_rate = 16000; VideoCodecCtx->width = 176; VideoCodecCtx->height = 144; VideoCodecCtx->time_base.den = 10; VideoCodecCtx->time_base.num = 1; VideoCodecCtx->pix_fmt = PIX_FMT_YUV420P; VideoCodecCtx->flags |= CODEC_FLAG_QSCALE; VideoCodecCtx->gop_size = 20; VideoCodecCtx->global_quality = (float)5 * (float)FF_QP2LAMBDA; VideoStream->quality = (float)5; VideoStream->r_frame_rate.den = 10; VideoStream->r_frame_rate.num = 1; VideoCodecCtx->flags |= CODEC_FLAG_GLOBAL_HEADER; VideoCodec = avcodec_find_encoder(VideoCodecCtx->codec_id); if(avcodec_open(VideoCodecCtx, VideoCodec) < 0) { printf("Fail open video\n"); return 1; } AudioStream = av_new_stream(FormatCtx, 1); AudioCodecCtx = AudioStream->codec; AudioCodecCtx->codec_id = CODEC_ID_AAC; AudioCodecCtx->codec_type = CODEC_TYPE_AUDIO; AudioCodecCtx->bit_rate = 32000; AudioCodecCtx->sample_rate = 8000; AudioCodecCtx->channels = 1; AudioCodecCtx->time_base= (AVRational){1, 8000}; AudioCodecCtx->sample_fmt = SAMPLE_FMT_S16; AudioCodecCtx->flags |= CODEC_FLAG_GLOBAL_HEADER; AudioCodec = avcodec_find_encoder(AudioCodecCtx->codec_id); if(avcodec_open(AudioCodecCtx, AudioCodec)) { printf("Fail open audio\n"); return 2; } url_fopen(&FormatCtx->pb, "test.3gp",URL_WRONLY); av_write_header(FormatCtx); AVStream *st; av_write_trailer(FormatCtx); url_fclose(FormatCtx->pb); avcodec_close(VideoStream->codec); avcodec_close(AudioStream->codec); for(uint32_t i=0;i< FormatCtx->nb_streams;i++) { st = FormatCtx->streams[i]; if (st->parser) { av_parser_close(st->parser); av_free_packet(&st->cur_pkt); } av_metadata_free(&st->metadata); av_free(st->index_entries); av_free(st->codec->extradata); av_free(st->codec); av_free(st->priv_data); av_free(st); } av_free(FormatCtx); usleep(3000); } return 0; } I tried to understand exactly where the memory is lost, and Founded that the functions av_cold void ff_psy_preprocess_end(struct FFPsyPreprocessContext *ctx) don't free context, so if add av_free(ctx); mem leak goes away. Patch attached.

2 1

[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

[PATCH]Lagarith decoder.
by saintdev＠gmail.com 17 Jan '11

17 Jan '11

Here are the latest Lagarith patches. The only changes to the rangecoder from the previous patch are cosmetic (K&R formatting). The decoder includes Loren's floating point emulation. Thanks to everyone, and especially Loren and Reimar, for all their help. -- -Nathan Caldwell

9 26

[PATCH] Added integer 32 bits support to wavpack
by fenrir＠videolan.org 25 Dec '10

25 Dec '10

The attached patch completes the support for files with INT32INFO chunk with non zero sent_bits. -- fenrir

6 23