August 2009 - ffmpeg-devel - lists.ffmpeg.org

dvdsub encoder: respect palette when it makes sense
by nicolas.george＠normalesup.org 05 Sep '10

05 Sep '10

Hi. The current version of the dvdsub encoder uses the alpha channel of the palette to weight the most important colors, and then completely ignores the palette. In particular, the alpha values it puts in the stream are hardcoded arbitrary values. The attached patch fix that to a certain extent: when there is exactly one rectangle and the palette has exactly 4 colors, the colors are kept in their original order and the alpha channel of the palette is copied to the stream. With this patch, it becomes possible to use the dvdsub encoder to produce a VOBSUB pair of file that mplayer can understand. Regards, -- Nicolas George

2 11

[PATCH] MMX2/SSSE3 VC1 loop filter
by lessen42＠gmail.com 11 Jul '10

11 Jul '10

Hi, Overall 17% faster decode on my Penryn, including the first function to use SSE4 instructions in ffmpeg! (which shave an entire 2 clocks off of vc1_h_loop_filter8 for me) One thing I don't understand is why the PSIGNW_SRA_MMX macro is necessary for correct results. I know that psignw isn't equivalent to ((a ^ b) - b), but the only difference I'm aware of is when b is 0, psignw sets a to 0 as well. It's probably a stupidly simple case that I'm missing... 700 dezicycles in vc1_v_loop_filter4_mmx2, 1048506 runs, 70 skips. 639 dezicycles in vc1_v_loop_filter4_ssse3, 1048447 runs, 129 skips 977 dezicycles in vc1_h_loop_filter4_mmx2, 2097069 runs, 83 skips. 951 dezicycles in vc1_h_loop_filter4_ssse3, 2097040 runs, 112 skips 1116 dezicycles in vc1_v_loop_filter8_mmx2, 33552803 runs, 1629 skips 677 dezicycles in vc1_v_loop_filter8_ssse3, 33552817 runs, 1615 skips 1648 dezicycles in vc1_h_loop_filter8_mmx2, 33552806 runs, 1626 skips 1158 dezicycles in vc1_h_loop_filter8_ssse3, 33552878 runs, 1554 skips 1137 dezicycles in vc1_h_loop_filter8_sse4, 33553447 runs, 985 skips

7 21

[PATCH] Suppress some error messages while there are previous images yet
by rnissl＠gmx.de 05 Jul '10

05 Jul '10

Hi, when showing the samples contained in subdirectory rnissl, a bunch of error messages appears which seem to be legal as those still images are no IDR frames / fields. The attached patch suppresses these error messages as long as there hasn't been any previous image yet. Bye. -- Dipl.-Inform. (FH) Reinhard Nissl mailto:rnissl at gmx.de

2 1

[PATCH] Altivec split-radix FFT
by lorenm＠u.washington.edu 01 Jul '10

01 Jul '10

1.6x faster than altivec radix-2 as measured by fft-test on a G4. 2.0x as measured by gprof. I failed to get oprofile or POWERPC_PERF to measure anything. This is my first time writing ppc asm, so any optimization hints are appreciated. I used raw asm rather than intrinsics because gcc adds a ginormous overhead to each function call. Is there anything I need to do to make it work on ppc64, if it doesn't already? --Loren Merritt

10 43

[PATCH 1/6] flacdec: change variable-length array to fixed length
by mans＠mansr.com 30 Jun '10

30 Jun '10

pred_order can never exceed 64, so always allocating that amount is safe and not very wasteful. --- libavcodec/flacdec.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/libavcodec/flacdec.c b/libavcodec/flacdec.c index 4ff7fae..643ed24 100644 --- a/libavcodec/flacdec.c +++ b/libavcodec/flacdec.c @@ -358,7 +358,7 @@ static int decode_subframe_lpc(FLACContext *s, int channel, int pred_order) { int i, j; int coeff_prec, qlevel; - int coeffs[pred_order]; + int coeffs[64]; int32_t *decoded = s->decoded[channel]; /* warm up samples */ -- 1.6.4

5 24

[PATCH] Default to using libraries when enabled
by lessen42＠gmail.com 25 Jun '10

25 Jun '10

Hi, This changes the order of initialization so that external libraries that are enabled are used by default over native codecs. My main motivation in moving these is to be able to have the ogg and mkv muxers default to encoding via libvorbis if it's enabled; currently, specifying CODEC_ID_VORBIS will use the native encoder regardless, which is much worse quality. However, this also affects faad and xvid. IMO, faad should be preferred anyway until the native decoder supports SBR, then dropped soon after. My only argument for xvid is that it's probably what users want to use when they --enable-libxvid.

11 37

IEC61937 compatible muxer
by bartek.wolowiec＠gmail.com 16 Jun '10

16 Jun '10

Hi, Here is IEC61937 compatible muxer. It can be used for sending compressed data over S/PDIF link. It's currently not very useful, but can be tested by: ./ffmpeg -y -i test_48khz.dts -acodec copy -f spdif test.spdif && ./ffmpeg -ar 48000 -f s16le -ac 2 -i test.spdif -f alsa plughw:0 -- Bartlomiej Wolowiec

7 23

[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] Export large crop values via AVCodecContext
by lessen42＠gmail.com 13 Mar '10

13 Mar '10

Hi, As far as I can tell, it's impractical to crop large values internally, since buffers returned from decode_frame must have the same dimensions as AVCodecContext, which also must be the coded dimensions (or up to 15 less) in order for get_buffer to allocate buffers of the needed size. Thus, the only solution I see is to export these values and leave it to individual applications that care about such offsets to deal with it themselves.

3 6