Re: [FFmpeg-devel] [PATCH 2/2] aacenc: add SIMD optimizations for abs_pow34 and quantization

On Sat, Oct 8, 2016 at 5:20 PM, Rostislav Pehlivanov wrote: > +cglobal aac_quantize_bands, 8, 8, 7, out, in, scaled, size, Q34, is_signed, > maxval, rounding [...] > +movdm4, is_signedd movd is SSE2. Can be worked around by moving it through the stack though. [...]

Re: [FFmpeg-devel] [PATCH] mov: Evaluate the movie display matrix

On Fri, Oct 07, 2016 at 10:38:22PM -0400, Vittorio Giovara wrote: > On Fri, Oct 7, 2016 at 8:38 PM, Michael Niedermayer > wrote: > > On Fri, Oct 07, 2016 at 03:31:46PM -0400, Vittorio Giovara wrote: > >> This matrix needs to be applied after all others have (currently only

[FFmpeg-devel] [PATCH v2] aacenc: add SIMD optimizations for abs_pow34 and quantization

Performance improvements: quant_bands: with: 681 decicycles in quant_bands, 8388453 runs,155 skips without: 1190 decicycles in quant_bands, 8388386 runs,222 skips Around 42% for the function Twoloop coder: abs_pow34: with/without: 7.82s/8.17s Around 4% for the entire encoder Both:

Re: [FFmpeg-devel] [PATCH] hwcontext: add a QSV implementation

On 9/28/2016 3:18 PM, James Almer wrote: > On 9/19/2016 7:00 AM, Ivan Uskov wrote: >> This should be a good step to make qsv branches of ffmpeg and libav >> closer. >> LGTM. > > This has now been merged. > > Could i ask you or Nablet to look at commits > a0524d9b1e1bb0012207584f067096df7792df6c

[FFmpeg-devel] [PATCH 1/2] aacenc: use the decoder's lcg PRNG

Using lfg was an overkill in this case where the random numbers were only used for encoder descisions. Should increase result uniformity between different FPUs and gives a slight speedup. Signed-off-by: Rostislav Pehlivanov --- libavcodec/aaccoder.c | 8 +++-

[FFmpeg-devel] [PATCH] avformat/utils: Update codec_id before using it in the parser init

Fixes assertion failure Fixes: input.avi Found-by: 连一汉 Signed-off-by: Michael Niedermayer --- libavformat/utils.c | 21 +++-- 1 file changed, 11 insertions(+), 10 deletions(-) diff --git a/libavformat/utils.c b/libavformat/utils.c

[FFmpeg-devel] [PATCH 0/2] aacenc: performance improvements

Should improve encoding performance a bit with twoloop and alot with the fast coder. Rostislav Pehlivanov (2): aacenc: use the decoder's lcg PRNG aacenc: add SIMD optimizations for abs_pow34 and quantization libavcodec/aaccoder.c| 30 +-- libavcodec/aaccoder_trellis.h

Re: [FFmpeg-devel] [PATCH] avformat/udp: deprecate local_port option

On Sat, Oct 08, 2016 at 03:35:02PM +0800, Steven Liu wrote: > > udp.c | 19 ++- > version.h |3 +++ > 2 files changed, 21 insertions(+), 1 deletion(-) > bf9c64b4a0fd9a1d998c11376c5c5c2a90389caf > 0001-avformat-udp-deprecate-local_port-option.patch > From

Re: [FFmpeg-devel] [PATCH] avformat/udp: deprecate local_port option

2016-10-09 2:11 GMT+08:00 Michael Niedermayer : > On Sat, Oct 08, 2016 at 03:35:02PM +0800, Steven Liu wrote: > > > > > udp.c | 19 ++- > > version.h |3 +++ > > 2 files changed, 21 insertions(+), 1 deletion(-) > >

Re: [FFmpeg-devel] [PATCH 1/2] aacenc: use the decoder's lcg PRNG

On Sat, Oct 08, 2016 at 04:20:30PM +0100, Rostislav Pehlivanov wrote: > Using lfg was an overkill in this case where the random numbers > were only used for encoder descisions. Should increase result > uniformity between different FPUs and gives a slight speedup. > > Signed-off-by: Rostislav

Re: [FFmpeg-devel] [PATCH v2] aacenc: add SIMD optimizations for abs_pow34 and quantization

On Sat, Oct 08, 2016 at 06:42:28PM +0100, Rostislav Pehlivanov wrote: > Performance improvements: > > quant_bands: > with: 681 decicycles in quant_bands, 8388453 runs,155 skips > without: 1190 decicycles in quant_bands, 8388386 runs,222 skips > Around 42% for the function > > Twoloop

[FFmpeg-devel] [PATCH 2/2] avformat/movenc: Check frame rate in mov_write_uuidprof_tag()

Signed-off-by: Michael Niedermayer --- libavformat/movenc.c | 11 +-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/libavformat/movenc.c b/libavformat/movenc.c index 2c155eb..d7c7158 100644 --- a/libavformat/movenc.c +++ b/libavformat/movenc.c @@

[FFmpeg-devel] [PATCH 1/2] avformat/movenc: Switch mov_write_uuidprof_tag() to avg_frame_rate

Using the stream timebase simply overflows Fix integer overflow in psp framerate computation Signed-off-by: Michael Niedermayer --- libavformat/movenc.c| 3 +-- tests/ref/fate/copy-psp | 2 +- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git

[FFmpeg-devel] [PATCH] lavfi/pan: renormalize negative gain coefficients properly

The parser for the outdef will accept a negative value for the first named channel's gain. As negative values effectively only invert the phase of the signal, and not negate the level, the gains' absolute values must be used to correctly accumulate the levels. Signed-off-by: Moritz Barsnick

[FFmpeg-devel] [PATCH] lavfi/pan: allow negative gain parameters also for other inputs than the first named

Expands the parser to also accept the separator '-' in addition to '+', and take the negative sign into consideration. Signed-off-by: Moritz Barsnick --- doc/filters.texi | 2 +- libavfilter/af_pan.c | 8 +--- 2 files changed, 6 insertions(+), 4 deletions(-) diff

[FFmpeg-devel] [PATCH] lavfi/pan: better handling of negative gains

While trying to use pan to subtract two channels (actually to see whether a two-channel file I have is effectively mono), I noticed that "pan=mono|c0=-0.5*c0+0.5*c1" would work, while "pan=mono|c0=0.5*c0-0.5*c1" would cause the parser to choke, expecting a '+'. Using this (second) patch, you

[FFmpeg-devel] [PATCH] avformat/udp: deprecate local_port option

0001-avformat-udp-deprecate-local_port-option.patch Description: Binary data ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] lavfi/metadata: fix metadata deletion if comparison returns false

On 10/6/16, Marton Balint wrote: > Signed-off-by: Marton Balint > --- > libavfilter/f_metadata.c | 4 +--- > 1 file changed, 1 insertion(+), 3 deletions(-) > lgtm ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org

[FFmpeg-devel] [PATCH]lavf/mxfenc: Do not print a useless error message

Hi! Attached patch removes a warning that is always shown if the default mxf muxer is used. Please comment, Carl Eugen From 07718cad8cefbafd18355f91b9548118308c42bd Mon Sep 17 00:00:00 2001 From: Carl Eugen Hoyos Date: Sat, 8 Oct 2016 11:23:41 +0200 Subject: [PATCH]

[FFmpeg-devel] [PATCH 2/2] libavformat/segment: reindent

Signed-off-by: Simon Thelen --- libavformat/segment.c | 14 +++--- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/libavformat/segment.c b/libavformat/segment.c index 82872fa..98e7bcc 100644 --- a/libavformat/segment.c +++ b/libavformat/segment.c @@

[FFmpeg-devel] [PATCH 1/2] avformat/segment: add delete segment_list flag

Works in the same manner as the hls segment_delete flag. Signed-off-by: Simon Thelen --- doc/muxers.texi | 5 +++ libavformat/segment.c | 89 +++ 2 files changed, 94 insertions(+) diff --git a/doc/muxers.texi