Hi,
2014-07-10 8:47 GMT+02:00 Christophe Gisquet :
> The 4 first patches reduce memory allocations as, in a simplified view,
> only the previous ctb line is needed for the various prediction mechanisms.
> The actual required allocation is much smaller, but would require too many
> changes. And the
On Fri, Jun 20, 2014 at 12:01:12AM +0200, Carl Eugen Hoyos wrote:
> On Thursday 19 June 2014 01:32:58 am Michael Niedermayer wrote:
> > On Wed, Jun 18, 2014 at 07:02:59PM -0400, Dave Rice wrote:
> > > Hi all,
> > >
> > > On May 7, 2014, at 6:41 AM, Carl Eugen Hoyos wrote:
> > > > Nicolas George ns
On 10.07.2014 01:36, Lukasz Marek wrote:
On 08.07.2014 14:49, Lukasz Marek wrote:
On 7 July 2014 23:59, wm4
new patch attached.
Hi, any comments, if not I will sent pull request tomorrow including
libssh and ftp implementations (with minor changes)
___
On Fri, Jul 11, 2014 at 12:14:29AM +0100, Ben Avison wrote:
> The previous implementation targeted DTS Coherent Acoustics, which only
> requires nbits == 4 (fft16()). This case was (and still is) linked directly
> rather than being indirected through ff_fft_calc_vfp(), but now the full
> range from
I benchmarked the result by measuring the number of gperftools samples that
hit anywhere in the AAC decoder (starting from aac_decode_frame()) or
specifically in butterflies_float_c() / ff_butterflies_float_vfp() for the
same sample AAC stream:
Before After
I benchmarked the result by measuring the number of gperftools samples that
hit anywhere in the AAC decoder (starting from aac_decode_frame()) or
specifically in vector_fmul_window_c() / ff_vector_fmul_window_vfp() for the
same sample AAC stream:
Before After
The previous implementation targeted DTS Coherent Acoustics, which only
requires nbits == 4 (fft16()). This case was (and still is) linked directly
rather than being indirected through ff_fft_calc_vfp(), but now the full
range from radix-4 up to radix-65536 is available. This benefits other codecs
The previous implementation targeted DTS Coherent Acoustics, which only
requires mdct_bits == 6. This relatively small size lent itself to
unrolling the loops a small number of times, and encoding offsets
calculated at assembly time within the load/store instructions of each
iteration.
In the more
On Thu, Jul 10, 2014 at 1:26 PM, Michael Niedermayer wrote:
> On Thu, Jul 10, 2014 at 01:12:20PM -0700, Vignesh Venkatasubramanian wrote:
>> On Thu, Jul 10, 2014 at 12:58 PM, Michael Niedermayer
>> wrote:
>> > On Mon, Jul 07, 2014 at 01:41:34PM -0700, Vignesh Venkatasubramanian wrote:
>> >> This
On Thu, Jul 10, 2014 at 01:12:20PM -0700, Vignesh Venkatasubramanian wrote:
> On Thu, Jul 10, 2014 at 12:58 PM, Michael Niedermayer
> wrote:
> > On Mon, Jul 07, 2014 at 01:41:34PM -0700, Vignesh Venkatasubramanian wrote:
> >> This patch adds the ability to generate WebM DASH manifest XML using
>
On Thu, Jul 10, 2014 at 12:58 PM, Michael Niedermayer wrote:
> On Mon, Jul 07, 2014 at 01:41:34PM -0700, Vignesh Venkatasubramanian wrote:
>> This patch adds the ability to generate WebM DASH manifest XML using
>> ffmpeg. A sample command line would be as follows:
>>
>> ffmpeg \
>> -f webm_dash_
On Mon, Jul 07, 2014 at 01:41:34PM -0700, Vignesh Venkatasubramanian wrote:
> This patch adds the ability to generate WebM DASH manifest XML using
> ffmpeg. A sample command line would be as follows:
>
> ffmpeg \
> -f webm_dash_manifest -i video1.webm \
> -f webm_dash_manifest -i video2.webm \
On Tue, Jul 08, 2014 at 01:50:05PM -0300, James Almer wrote:
> And use the x86util ones instead, which are optimized for mmxext/sse2.
> About ~1% increase in performance on pre SSSE3 processors.
>
> Signed-off-by: James Almer
> ---
> libavfilter/x86/yadif-10.asm | 21 -
> 1 f
On Thu, Jul 10, 2014 at 04:08:21PM +0200, Christophe Gisquet wrote:
> 2014-07-10 14:53 GMT+02:00 Michael Niedermayer :
> > On Thu, Jul 10, 2014 at 06:47:54AM +, Christophe Gisquet wrote:
> >> When dealing with MVs, both components may be processed at a time.
> >
> > mixing memory access sizes c
On Thu, Jul 10, 2014 at 06:47:52AM +, Christophe Gisquet wrote:
> The merge list only needs to be derived up to the merge index.
> ---
> libavcodec/hevc_mvs.c | 35 ++-
> 1 file changed, 26 insertions(+), 9 deletions(-)
applied
thanks
[...]
--
Michael Gn
On Thu, Jul 10, 2014 at 07:53:25PM +0900, KO Myung-Hun wrote:
> ---
> configure |2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/configure b/configure
> index 658efb2..c5e854d 100755
> --- a/configure
> +++ b/configure
> @@ -3784,7 +3784,7 @@ elif enabled x86; then
>
> I am trying to parallel the h264 decoder in FFmpeg using wavefront
> processing.
How are you passing your CABAC state along your wavefront?
Kieran
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
Note: this is also used for skip blocks, so the speed improvement of
this patch should be around 1 or 2%. (amvp probably much less)
2014-07-10 8:47 GMT+02:00 Christophe Gisquet :
> The merge list only needs to be derived up to the merge index.
> ---
> libavcodec/hevc_mvs.c | 35 ++
2014-07-10 14:53 GMT+02:00 Michael Niedermayer :
> On Thu, Jul 10, 2014 at 06:47:54AM +, Christophe Gisquet wrote:
>> When dealing with MVs, both components may be processed at a time.
>
> mixing memory access sizes can lead to speed penalties on some cpus
> when such mixed accesses are close t
On Thu, Jul 10, 2014 at 06:47:53AM +, Christophe Gisquet wrote:
> When the candidate has been found, no need to derive others.
> ---
> libavcodec/hevc_mvs.c | 5 +
> 1 file changed, 5 insertions(+)
applied
thanks
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC7870
Hi,
On Thu, Jul 10, 2014 at 2:47 AM, Christophe Gisquet <
christophe.gisq...@gmail.com> wrote:
> -mergecandlist[nb_merge_cand].mv[0].x = l0_cand.mv
> [0].x;
> -mergecandlist[nb_merge_cand].mv[0].y = l0_cand.mv
> [0].y;
> -mergecandlist[nb_
Some improvements in lrc_probe.
Updated patch is attached.
Thanks.
>From 6ad46495be298fc7254676b141378295c2238e69 Mon Sep 17 00:00:00 2001
From: Star Brilliant
Date: Wed, 9 Jul 2014 14:24:05 +0800
Subject: [PATCH] AVFormat: LRC demuxer and muxer
---
Changelog| 1 +
doc/gener
On Thu, Jul 10, 2014 at 06:47:54AM +, Christophe Gisquet wrote:
> When dealing with MVs, both components may be processed at a time.
mixing memory access sizes can lead to speed penalties on some cpus
when such mixed accesses are close together
did you benchmark this ?
[...]
--
Michael
On Thu, Jul 10, 2014 at 06:47:56AM +, Christophe Gisquet wrote:
> ---
> libavcodec/hevc.c | 12 ++--
> 1 file changed, 6 insertions(+), 6 deletions(-)
applied
thanks
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Awnsering whenever a program halts
On Thu, Jul 10, 2014 at 06:47:55AM +, Christophe Gisquet wrote:
> ---
> libavcodec/hevc_mvs.c | 8
> 1 file changed, 4 insertions(+), 4 deletions(-)
applied
thx
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
No great genius has ever existed withou
On Thu, Jul 10, 2014 at 12:59:29PM +0200, Paul B Mahol wrote:
> On 7/10/14, Yu Xiaolei wrote:
> >
> > ---
> > configure | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/configure b/configure
> > index d17abb9..bd49b8b 100755
> > --- a/configure
> > +++ b/configure
>
On 7/10/14, Yu Xiaolei wrote:
>
> ---
> configure | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/configure b/configure
> index d17abb9..bd49b8b 100755
> --- a/configure
> +++ b/configure
> @@ -2136,7 +2136,7 @@ qdm2_decoder_select="mdct rdft mpegaudiodsp"
> ra_144_encod
---
configure |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/configure b/configure
index 658efb2..c5e854d 100755
--- a/configure
+++ b/configure
@@ -3784,7 +3784,7 @@ elif enabled x86; then
case $cpu in
i[345]86|pentium)
cpuflags="-march=$cpu"
Changes since last time:
I've made the changes to configure, and squashed the patches together.
Option changed from fribidi=1 (default 0) to text_shaping=1 (default 1).
(Ideas for better names are definitely welcome.)
Hopefully I've made the documentation more understandable.
No longer testing for
---
configure | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure b/configure
index d17abb9..bd49b8b 100755
--- a/configure
+++ b/configure
@@ -2136,7 +2136,7 @@ qdm2_decoder_select="mdct rdft mpegaudiodsp"
ra_144_encoder_select="audio_frame_queue lpc audiodsp"
ra_144_
On 7/9/14, Dave Rice wrote:
> Here is an updated patch based on the advice from Stefano and Timothy.
> Thanks much,
> Dave Rice
>
[...]
pushed
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
On 2014-07-10 03:36, Anssi Hannula wrote:
> 10.07.2014 02:40, Nicolas Martyanoff kirjoitti:
> > On 2014-07-10 01:32, Anssi Hannula wrote:
> >> avformat/hlsenc: cleaning
> >> - looks good to me, and there is no maintainer so I guess this is OK
> >> - well, actually, maybe you could use more than one
32 matches
Mail list logo
