On 17/02/2018 00:57, wm4 wrote:
On Fri, 16 Feb 2018 18:02:05 +0100
Luca Barbato wrote:
Unbreaks the rate-control behaviour.
---
libavcodec/libx265.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavcodec/libx265.c b/libavcodec/libx265.c
index
Set probably Ok.
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
On Fri, Feb 16, 2018 at 9:15 PM, Luca Barbato wrote:
> On 17/02/2018 00:57, wm4 wrote:
>
>> On Fri, 16 Feb 2018 18:02:05 +0100
>> Luca Barbato wrote:
>>
>> Unbreaks the rate-control behaviour.
>>>
>>
Would be nice that the commit message is expanded,
On Fri, Feb 16, 2018 at 12:02 PM, Luca Barbato wrote:
> And not the often misleading timebase one.
> ---
> libavcodec/libopenh264enc.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/libavcodec/libopenh264enc.c b/libavcodec/libopenh264enc.c
> index
On Fri, Feb 16, 2018 at 12:02 PM, Luca Barbato wrote:
> In preparation of using it in the encoders.
> ---
> libavcodec/avcodec.h | 3 ++-
> libavcodec/utils.c | 5 +
> 2 files changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/libavcodec/avcodec.h
On Fri, Feb 16, 2018 at 12:02 PM, Luca Barbato wrote:
> Unbreaks the rate-control behaviour.
>
Same comment as for the x265 patch.
This seems to properly use x264 API.
---
> libavcodec/libx264.c | 8 ++--
> 1 file changed, 6 insertions(+), 2 deletions(-)
>
> diff --git
Unbreak the rate-control behaviour.
---
libavcodec/nvenc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavcodec/nvenc.c b/libavcodec/nvenc.c
index a7171e4b17..53cb5f974e 100644
--- a/libavcodec/nvenc.c
+++ b/libavcodec/nvenc.c
@@ -946,8 +946,8 @@ static int
Unbreaks the rate-control behaviour.
---
libavcodec/libx265.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavcodec/libx265.c b/libavcodec/libx265.c
index fd5452193b..73aff2caef 100644
--- a/libavcodec/libx265.c
+++ b/libavcodec/libx265.c
@@ -116,8 +116,8 @@ static
Unbreaks the rate-control behaviour.
---
libavcodec/libx264.c | 8 ++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/libavcodec/libx264.c b/libavcodec/libx264.c
index 3dc53aaf38..b3f91a1f07 100644
--- a/libavcodec/libx264.c
+++ b/libavcodec/libx264.c
@@ -573,8 +573,12 @@
Now the framerate field is always populated.
---
libavcodec/qsvenc.c | 9 ++---
1 file changed, 2 insertions(+), 7 deletions(-)
diff --git a/libavcodec/qsvenc.c b/libavcodec/qsvenc.c
index 24d9ec4d71..bb1494949e 100644
--- a/libavcodec/qsvenc.c
+++ b/libavcodec/qsvenc.c
@@ -402,13 +402,8 @@
And not the often misleading timebase one.
---
libavcodec/libopenh264enc.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavcodec/libopenh264enc.c b/libavcodec/libopenh264enc.c
index d075cb0854..f779c46196 100644
--- a/libavcodec/libopenh264enc.c
+++
So it can be used by the rate control as a better estimate of it.
Before the time_base was used and that confused some rate-control
algorithms or cause setup failures in some of the encoders.
---
avtools/avconv.c | 7 +--
avtools/avconv.h | 2 ++
avtools/avconv_opt.c | 10 --
In preparation of using it in the encoders.
---
libavcodec/avcodec.h | 3 ++-
libavcodec/utils.c | 5 +
2 files changed, 7 insertions(+), 1 deletion(-)
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 7eaa0c9277..f86ae0b8d6 100644
--- a/libavcodec/avcodec.h
+++
All this is triggered by a bug report about x265 rate-control behaving
erratically. Since it was already a pain point for QSV I decided to
give it a stab.
Now if the input has an average framerate you won't be forced to
manually set it in order to have the encoder behave as expected.
(I'll add
It seems to improve the compatibility with the js demuxers.
---
libavformat/hlsenc.c | 25 -
1 file changed, 20 insertions(+), 5 deletions(-)
diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index c84dd82e7d..2fd3f19fcb 100644
--- a/libavformat/hlsenc.c
+++
On 16/02/2018 18:28, Diego Biurrun wrote:
On Fri, Feb 16, 2018 at 06:02:05PM +0100, Luca Barbato wrote:
Unbreaks the rate-control behaviour.
How does this unbreak what?
If your timebase is 1/1000 and your actual frame rate is 30, it gets
extra creative in deciding how many bits allocate
On Fri, Feb 16, 2018 at 06:02:05PM +0100, Luca Barbato wrote:
> Unbreaks the rate-control behaviour.
How does this unbreak what?
Diego
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
On Fri, Feb 16, 2018 at 06:02:04PM +0100, Luca Barbato wrote:
> So it can be used by the rate control as a better estimate of it.
> Before the time_base was used and that confused some rate-control
> algorithms or cause setup failures in some of the encoders.
cause_d
No opinion on the patch
These are supported by llvm-dlltool.
---
configure | 4
1 file changed, 4 insertions(+)
diff --git a/configure b/configure
index 06fb839a18..1c35f9dc64 100755
--- a/configure
+++ b/configure
@@ -3894,6 +3894,10 @@ case $target_os in
LIBTARGET=i386
if enabled x86_64; then
This reverts commit 67c72f08a4707c18a67a4734660e3a23cc9488b6.
While the linker produced import libraries might work with MSVC in
simple test cases, they don't if e.g. linking to multiple GNU ld
produced import libraries at the same time. The ones produced by
dlltool work fine though.
This issue
On Fri, 16 Feb 2018 18:02:05 +0100
Luca Barbato wrote:
> Unbreaks the rate-control behaviour.
> ---
> libavcodec/libx265.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/libx265.c b/libavcodec/libx265.c
> index fd5452193b..73aff2caef
21 matches
Mail list logo