> -Original Message-
> From: ffmpeg-devel On Behalf Of James
> Almer
> Sent: Friday, August 21, 2020 10:43 PM
> To: ffmpeg-devel@ffmpeg.org
> Subject: Re: [FFmpeg-devel] [PATCH 1/2] avcodec: add av1 hardware
> acceleration decoder
>
> On 8/21/2020 2:29 AM, Fei Wang wrote:
> > This av1 d
On Fri, 2020-08-21 at 09:36 +0200, Max Dmitrichenko wrote:
> On Fri, Aug 21, 2020 at 9:29 AM Xiang, Haihao
> wrote:
>
> > On Fri, 2020-08-21 at 05:48 +, Soft Works wrote:
> > > > -Original Message-
> > > > From: ffmpeg-devel On Behalf Of
> > > > Haihao Xiang
> > > > Sent: Friday, Aug
On Fri, 2020-08-21 at 09:47 +0200, Thilo Borgmann wrote:
> >
> >
> > On Thu, 2020-08-20 at 02:26 +0200, Thilo Borgmann wrote:
> > > Am 20.08.20 um 02:13 schrieb Xiang, Haihao:
> > > > On Wed, 2020-08-19 at 13:53 +0200, Nicolas George wrote:
> > > > > Haihao Xiang (12020-08-19):
> > > > > > ---
>
Please review:
It help initialize chroma format and other info properly
Chroma format wasn't correct if I use below code:
avformat_find_stream_info(fmtc, NULL);
iVideoStream = av_find_best_stream(fmtc, AVMEDIA_TYPE_VIDEO, -1, -1,
NULL, 0);
eChromaFormat =
(AVPixelFormat)f
On Fri, 2020-08-21 at 20:11 +, Soft Works wrote:
> > -Original Message-
> > From: ffmpeg-devel On Behalf Of
> > Soft Works
> > Sent: Friday, August 21, 2020 9:45 AM
> > To: FFmpeg development discussions and patches > de...@ffmpeg.org>
> > Subject: Re: [FFmpeg-devel] [PATCH v2 1/2] qs
I'm new to the community and eager to help out with this work
as well. Is there a reason why we would build something to do
this rather than something like doxygen? It would mean that the
team can point any contributors to the documentation for doxygen
and the developers of ffmpeg will not need to
>
> What are the timestamps (and corresponding time bases) of the first
> frame you send on each input?
this frame is sent the first
fs->in[0]
pts = 0;
timscale=1/1000 but it doesn't matter as pts is 0
this frame is sent the second
fs->in[1]
pts =140
timscale=1/25
When I call av_buffersink_g
On 8/21/2020 2:29 AM, Fei Wang wrote:
> +static int update_reference_list (AVCodecContext *avctx)
> +{
> +AV1DecContext *s = avctx->priv_data;
> +AV1RawFrameHeader *header= &s->raw_frame_header;
> +int i;
> +int ret = 0;
> +
> +for (i = 0; i < 8; i++) {
Use AV1_NUM_REF_FRAMES i
> 在 2020年8月24日，09:59，Zane van Iperen 写道：
>
> On Sun, 09 Aug 2020 22:38:19 +1000
> "FFmpegQuickHacks"
> wrote:
>
>> Hi,
>>
>> The original patch is archived at:
>> http://ffmpeg.org/pipermail/ffmpeg-devel/2020-May/262634.html
>>
>> Here it is again:
>>
>> ---
>> src/documentation | 10 ++
On 2020-08-23 08:21, Nicolas George wrote:
Since the idea of documentation built in the libraries seems popular, I
have tried to outline an API to access it.…
See the attached file […`documentation.c` omitted…].
The idea would be to have the build system convert the documentation
into a C file
From: hwren
Signed-off-by: hbj
Signed-off-by: hwren
---
Changelog | 1 +
configure | 4 +
doc/decoders.texi | 21
doc/general.texi | 8 ++
libavcodec/Makefile| 1 +
libavcodec/allcodecs.c | 1 +
libavcodec/libuavs3d.c | 248 +
From: hwren
*** Latest changes ***
Supplement information in dec_frame, including pkt_pos, pkt_size,
coded_picture_number
and display_picture_number.
Will push if no more comments. Thanks.
*** Historical changes ***
=== Version1 ===
These patches are to supplement the third generation of Audi
From: hwren
Signed-off-by: hbj
Signed-off-by: hwren
---
libavcodec/codec_desc.c | 7 +++
libavcodec/codec_id.h | 1 +
2 files changed, 8 insertions(+)
diff --git a/libavcodec/codec_desc.c b/libavcodec/codec_desc.c
index 0ae6aee63b..61473178fc 100644
--- a/libavcodec/codec_desc.c
+++ b/l
From: hwren
Signed-off-by: hwren
---
libavcodec/Makefile | 2 +
libavcodec/avs3.c | 95 +
libavcodec/avs3.h | 52 +
3 files changed, 149 insertions(+)
create mode 100644 libavcodec/avs3.c
create mode 100644 libavcodec/av
From: hwren
Signed-off-by: hbj
Signed-off-by: hwren
---
libavcodec/Makefile | 1 +
libavcodec/avs3_parser.c | 180 +++
libavcodec/parsers.c | 1 +
3 files changed, 182 insertions(+)
create mode 100644 libavcodec/avs3_parser.c
diff --git a/liba
From: hwren
Signed-off-by: hbj
Signed-off-by: hwren
---
libavformat/Makefile | 1 +
libavformat/allformats.c | 1 +
libavformat/avs3dec.c| 68
libavformat/version.h| 2 +-
4 files changed, 71 insertions(+), 1 deletion(-)
create mode 1006
> -Original Message-
> From: ffmpeg-devel On Behalf Of James
> Almer
> Sent: Friday, August 21, 2020 11:01 PM
> To: ffmpeg-devel@ffmpeg.org
> Subject: Re: [FFmpeg-devel] [PATCH 1/2] avcodec: add av1 hardware
> acceleration decoder
>
> On 8/21/2020 11:47 AM, Paul B Mahol wrote:
> > On 8/
On 8/21/2020 2:29 AM, Fei Wang wrote:
> +static int vaapi_av1_start_frame(AVCodecContext *avctx,
> + av_unused const uint8_t *buffer,
> + av_unused uint32_t size)
> +{
> +AV1DecContext *s = avctx->priv_data;
> +const AV1RawSequ
On Sun, 09 Aug 2020 22:38:19 +1000
"FFmpegQuickHacks"
wrote:
> Hi,
>
> The original patch is archived at:
> http://ffmpeg.org/pipermail/ffmpeg-devel/2020-May/262634.html
>
> Here it is again:
>
> ---
> src/documentation | 10 ++
> 1 files changed, 6 insertions(+), 4 deletions(-)
>
> -Original Message-
> From: ffmpeg-devel On Behalf Of Fu, Ting
> Sent: 2020年8月21日 18:01
> To: FFmpeg development discussions and patches
> Subject: Re: [FFmpeg-devel] [PATCH V2] dnn: move output name from
> DNNModel.set_input_output to DNNModule.execute_model
>
>
>
> > -Original
> -Original Message-
> From: ffmpeg-devel On Behalf Of Mingyu
> Yin
> Sent: 2020年8月23日 23:12
> To: ffmpeg-devel@ffmpeg.org
> Subject: [FFmpeg-devel] [PATCH V5 2/2] dnn_backend_native_layer_mathbinary:
> add floormod support
>
> Signed-off-by: Mingyu Yin
> ---
> libavfilter/dnn/dnn_bac
On 8/23/2020 7:59 PM, Mark Thompson wrote:
> On 23/08/2020 20:23, James Almer wrote:
>> This makes them available for all frames within a Temporal Unit.
>>
>> Signed-off-by: James Almer
>> ---
>> libavcodec/cbs_av1_syntax_template.c | 43 +---
>> 1 file changed, 33 inser
Section 5.9.7 of the spec states
UpscaledWidth = RefUpscaledWidth[ ref_frame_idx[ i ] ]
FrameWidth= UpscaledWidth
FrameHeight = RefFrameHeight[ ref_frame_idx[ i ] ]
RenderWidth = RefRenderWidth[ ref_frame_idx[ i ] ]
RenderHeight = RefRenderHeight[ ref_frame_idx[ i ] ]
Signed-off-by: James Almer
---
libavcodec/cbs_av1.c | 12
libavcodec/cbs_av1.h | 5 +
2 files changed, 17 insertions(+)
diff --git a/libavcodec/cbs_av1.c b/libavcodec/cbs_av1.c
index 72ad18151e..726a2211da 100644
--- a/libavcodec/cbs_av1.c
+++ b/libavcodec/cbs_av1.c
@@ -1247,6
On 23/08/2020 20:23, James Almer wrote:
This makes them available for all frames within a Temporal Unit.
Signed-off-by: James Almer
---
libavcodec/cbs_av1_syntax_template.c | 43 +---
1 file changed, 33 insertions(+), 10 deletions(-)
Set LGTM.
Thanks,
- Mark
_
---
Setting HDR properties is a useful feature, but it's very unclear what we want
it to actually look like to the user. Not all encoders and decoders support
it, so it's essentially required that the implementation happen at the
bitstream filter level so that we can support all codecs in the s
---
libavcodec/h264_metadata_bsf.c | 225 +++--
1 file changed, 45 insertions(+), 180 deletions(-)
diff --git a/libavcodec/h264_metadata_bsf.c b/libavcodec/h264_metadata_bsf.c
index 7d6d871240..4243a8da98 100644
--- a/libavcodec/h264_metadata_bsf.c
+++ b/libavcodec/h26
---
libavcodec/vp9_metadata_bsf.c | 61 +--
1 file changed, 16 insertions(+), 45 deletions(-)
diff --git a/libavcodec/vp9_metadata_bsf.c b/libavcodec/vp9_metadata_bsf.c
index 00a5580c4d..1d82a401d5 100644
--- a/libavcodec/vp9_metadata_bsf.c
+++ b/libavcodec/vp9_met
This adds support for updating new extradata.
---
libavcodec/mpeg2_metadata_bsf.c | 94 +
1 file changed, 12 insertions(+), 82 deletions(-)
diff --git a/libavcodec/mpeg2_metadata_bsf.c b/libavcodec/mpeg2_metadata_bsf.c
index d0048c0e25..367aa1f9f5 100644
--- a/liba
---
libavcodec/av1_metadata_bsf.c | 176 --
1 file changed, 21 insertions(+), 155 deletions(-)
diff --git a/libavcodec/av1_metadata_bsf.c b/libavcodec/av1_metadata_bsf.c
index 3158ba995b..31fd4f0f41 100644
--- a/libavcodec/av1_metadata_bsf.c
+++ b/libavcodec/av1_me
---
libavcodec/h264_redundant_pps_bsf.c | 99 ++---
1 file changed, 21 insertions(+), 78 deletions(-)
diff --git a/libavcodec/h264_redundant_pps_bsf.c
b/libavcodec/h264_redundant_pps_bsf.c
index cf41abe96c..fb91227eec 100644
--- a/libavcodec/h264_redundant_pps_bsf.c
+++ b
---
libavcodec/h265_metadata_bsf.c | 191 -
1 file changed, 22 insertions(+), 169 deletions(-)
diff --git a/libavcodec/h265_metadata_bsf.c b/libavcodec/h265_metadata_bsf.c
index 504a75dac2..c3eadee92b 100644
--- a/libavcodec/h265_metadata_bsf.c
+++ b/libavcodec/h26
---
This series depends on the CBS unit table rewrite and following cleanups.
Since comments on that series seem to have dried up, I will apply it in a few
days if nothing further comes up.
libavcodec/Makefile | 2 +-
libavcodec/cbs_bsf.c | 161 +++
On Sun, Aug 23, 2020 at 04:02:23PM +0530, gautamr...@gmail.com wrote:
> From: Gautam Ramakrishnan
>
> This patch allows setting a compression ratio and to
> set multiple layers. The user has to input a compression
> ratio for each layer.
> The per layer compression ration can be set as follows:
>
On Sun, Aug 23, 2020 at 04:02:21PM +0530, gautamr...@gmail.com wrote:
> From: Gautam Ramakrishnan
>
> This patch fixes tag tree coding for JPEG2000
> encoder.
> ---
> libavcodec/j2kenc.c | 41 -
> libavcodec/jpeg2000.c | 1 +
> libavcodec/jpeg2000.h |
First about this discussion in general:
There is a reason this is tagged RFC and the other thread is called
proposal. I hope we can go on with the vivid discussions while keeping
in mind that it is not yet narrowed down and nothing is decided.
Always constraining all ideas, is not a good way to o
On Sun, 23 Aug 2020, Moritz Barsnick wrote:
Hej igen,
On Fri, Aug 21, 2020 at 12:19:06 +0300, Martin Storsjö wrote:
LGTM, with one small nit:
[...]
{"listen", "Listen for incoming connections", offsetof(pstruct,
options_field . listen),AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, .
Nicolas George:
> ff_set_common_formats() is currently only called after
> graph_check_validity(), guaranteeing that inputs and outputs
> are connected.
> If we want to support configuring partially-connected graphs,
> we will have a lot of redesign to do anyway.
>
> Fix CID 1466262 / 1466263.
>
Signed-off-by: James Almer
---
libavcodec/cbs_av1_syntax_template.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/libavcodec/cbs_av1_syntax_template.c
b/libavcodec/cbs_av1_syntax_template.c
index dedd549572..28d9ab9817 100644
--- a/libavcodec/cbs_av1_syntax_template.c
+++ b/libavcodec/cbs_
Simplifies code
Signed-off-by: James Almer
---
libavcodec/av1_parser.c | 28 ++--
1 file changed, 6 insertions(+), 22 deletions(-)
diff --git a/libavcodec/av1_parser.c b/libavcodec/av1_parser.c
index cd426a2b0f..1b9868d364 100644
--- a/libavcodec/av1_parser.c
+++ b/libav
This makes them available for all frames within a Temporal Unit.
Signed-off-by: James Almer
---
libavcodec/cbs_av1_syntax_template.c | 43 +---
1 file changed, 33 insertions(+), 10 deletions(-)
diff --git a/libavcodec/cbs_av1_syntax_template.c
b/libavcodec/cbs_av1_synta
Hendrik Leppkes (12020-08-23):
> I would at least want it to be optional.
Of course.
Also, it would be a new build-dep, it must be optional for that reason
too. Although it would make harder to eliminate the descriptions in the
source files like Alexander wants.
Regards,
--
Nicolas George
On Sun, Aug 23, 2020 at 6:48 PM Marton Balint wrote:
>
>
>
> On Sun, 23 Aug 2020, Nicolas George wrote:
>
> > Since the idea of documentation built in the libraries seems popular,
>
> I don't like the idea of compiling the full help text into the binaries.
> Libraries are already huge binaries, ad
On Sun, Aug 23, 2020 at 05:15:18PM +0200, Nicolas George wrote:
> Michael Niedermayer (12020-08-22):
> > iam not sure what extension you (plural) envission here but AVOption.help
> > can be extended without breaking API/ABI.
> > Consider if for example one wanted to add a filename, markdown or othe
Andreas Rheinhardt (12020-08-23):
> > No. If *f is freshly allocated, it has no owner yet and unref_fn(f) will
> > free it and set *f to NULL; av_freep(f) is then a no-op, so I removed
> > it. Keeping it would also be against the philosphy of this API (that it
> > cleans up after itself in case of
Andreas Rheinhardt:
> Nicolas George:
>> Andreas Rheinhardt (12020-08-09):
>>> The formats API deals with lists of channel layouts, sample rates,
>>> pixel formats and sample formats. These lists are refcounted in a way in
>>> which the list structure itself contains pointers to all of its owners.
On Sun, 16. Aug 17:42, Andriy Gelman wrote:
> On Sat, 15. Aug 22:34, Mark Thompson wrote:
> > On 17/07/2020 03:09, Andriy Gelman wrote:
> > > From: Andriy Gelman
> > >
> > > Return proper error when a frame buffers are full. This path is
> > > triggered on the DragonBoard 410c since the encoding
On 23/08/2020 18:31, James Almer wrote:
On 8/23/2020 1:06 PM, Mark Thompson wrote:
The OBU must be in both the spatial and temporal layers for the
operating point, not just one of them.
---
libavcodec/cbs_av1_syntax_template.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git
Marton Balint (12020-08-23):
> It is loaded (or at least mmapped) into memory unconditionally.
It is mmaped, and therefore only uses memory if it is used. Actually,
when it is used, it ends up requiring less memory than other solutions.
So I ask again: why do you consider it an issue?
>
On 8/23/2020 1:06 PM, Mark Thompson wrote:
> The OBU must be in both the spatial and temporal layers for the
> operating point, not just one of them.
> ---
> libavcodec/cbs_av1_syntax_template.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/libavcodec/cbs_av1_syntax_tem
On Sun, 23 Aug 2020, Nicolas George wrote:
Marton Balint (12020-08-23):
I don't like the idea of compiling the full help text into the binaries.
Libraries are already huge binaries, adding the help text increases the
binary size even more.
Why do you consider this an issue? It does not take
Marton Balint (12020-08-23):
> I don't like the idea of compiling the full help text into the binaries.
> Libraries are already huge binaries, adding the help text increases the
> binary size even more.
Why do you consider this an issue? It does not take more disk space in
the binaries than in sep
---
libavformat/http.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/libavformat/http.c b/libavformat/http.c
index 6c39da1a8b..3d25d652d3 100644
--- a/libavformat/http.c
+++ b/libavformat/http.c
@@ -577,7 +577,7 @@ static int http_open(URLContext *h, const char *uri, int
f
On Sun, 23 Aug 2020, Nicolas George wrote:
Since the idea of documentation built in the libraries seems popular,
I don't like the idea of compiling the full help text into the binaries.
Libraries are already huge binaries, adding the help text increases the
binary size even more. If we rea
Nicolas George:
> Andreas Rheinhardt (12020-08-09):
>> The formats API deals with lists of channel layouts, sample rates,
>> pixel formats and sample formats. These lists are refcounted in a way in
>> which the list structure itself contains pointers to all of its owners.
>> Furthermore, it is poss
Signed-off-by: Ting Fu
---
libavfilter/dnn/dnn_backend_native.c | 59 +++
libavfilter/dnn/dnn_backend_native.h | 5 ++
.../dnn/dnn_backend_native_layer_avgpool.c| 10 +++-
.../dnn/dnn_backend_native_layer_avgpool.h| 2 +-
.../dnn/dnn_backend_native_laye
Unify all error return as DNN_ERROR, in order to cease model executing
when return error in ff_dnn_execute_model_native layer_func.pf_exec
Signed-off-by: Ting Fu
---
V4:
Rename NetworkContext to NativeContext
Move pf_exec return DNN_ERROR from PATCH 2/2 to 1/2
libavfilter/dnn/dnn_backen
The OBU must be in both the spatial and temporal layers for the
operating point, not just one of them.
---
libavcodec/cbs_av1_syntax_template.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libavcodec/cbs_av1_syntax_template.c
b/libavcodec/cbs_av1_syntax_template.c
index 2c
On Sun, Aug 23, 2020 at 11:40 PM Fu, Ting wrote:
>
>
> > -Original Message-
> > From: ffmpeg-devel On Behalf Of Guo,
> > Yejun
> > Sent: Friday, August 21, 2020 06:57 PM
> > To: FFmpeg development discussions and patches
> > Subject: Re: [FFmpeg-devel] [PATCH V3 1/2] dnn/native: unify e
> -Original Message-
> From: ffmpeg-devel On Behalf Of Guo,
> Yejun
> Sent: Friday, August 21, 2020 06:57 PM
> To: FFmpeg development discussions and patches
> Subject: Re: [FFmpeg-devel] [PATCH V3 1/2] dnn/native: unify error return to
> DNN_ERROR
>
>
>
> > -Original Message
Alexander Strasser (12020-08-22):
> Wow the discussion gets up to speed quickly :)
Yes, it's nice for once.
> I think we should eradicate the copy in the source files and
> keep the description in the corresponding .md file. We shall
> build a tool that creates supplementary C files from the .md
> On Aug 23, 2020, at 4:51 PM, Andreas Rheinhardt
> wrote:
>
> Zhao Zhili:
>>
>>> On Aug 23, 2020, at 7:04 AM, Andreas Rheinhardt
>>> wrote:
>>>
>>> Casting const away directly before return is completely pointless
>>> here as avfilter_get_by_name() already returns a const AVFilter *.
>>>
Since the idea of documentation built in the libraries seems popular, I
have tried to outline an API to access it.
See the attached file.
The idea would be to have the build system convert the documentation
into a C file with initialization for one or several AVDocNode
structures.
Note that sinc
Michael Niedermayer (12020-08-22):
> iam not sure what extension you (plural) envission here but AVOption.help
> can be extended without breaking API/ABI.
> Consider if for example one wanted to add a filename, markdown or other
> text string. AVOption.help is a zero terminated string. you can add
Signed-off-by: Mingyu Yin
---
.../dnn/dnn_backend_native_layer_mathbinary.c | 143 +-
1 file changed, 73 insertions(+), 70 deletions(-)
diff --git a/libavfilter/dnn/dnn_backend_native_layer_mathbinary.c
b/libavfilter/dnn/dnn_backend_native_layer_mathbinary.c
index dd42c329a9..38
Signed-off-by: Mingyu Yin
---
libavfilter/dnn/dnn_backend_native_layer_mathbinary.c | 7 +++
libavfilter/dnn/dnn_backend_native_layer_mathbinary.h | 1 +
tests/dnn/dnn-layer-mathbinary-test.c | 5 +
tools/python/convert_from_tensorflow.py | 2 +-
tools/python
Signed-off-by: Nicolas George
---
doc/ffmpeg.texi | 10 ++
fftools/ffmpeg.h| 1 +
fftools/ffmpeg_filter.c | 2 ++
fftools/ffmpeg_opt.c| 3 +++
4 files changed, 16 insertions(+)
diff --git a/doc/ffmpeg.texi b/doc/ffmpeg.texi
index 267ddfe8b5..c054aedf3f 100644
--- a
Andreas Rheinhardt (12020-08-09):
> The formats API deals with lists of channel layouts, sample rates,
> pixel formats and sample formats. These lists are refcounted in a way in
> which the list structure itself contains pointers to all of its owners.
> Furthermore, it is possible for a list to be
On Sun, Aug 23, 2020 at 03:04:10PM +0200, Marton Balint wrote:
>
>
> On Sun, 23 Aug 2020, lance.lmw...@gmail.com wrote:
>
> > On Sun, Aug 23, 2020 at 10:17:01AM +0200, Marton Balint wrote:
> > >
> > >
> > > On Sun, 23 Aug 2020, lance.lmw...@gmail.com wrote:
> > >
> > > > On Sat, Aug 22, 2020
Andreas Rheinhardt (12020-08-09):
> The query_formats function of the channelmap filter tries to allocate
> a list of channel layouts which on success are attached to more permanent
> objects (an AVFilterLink) for storage afterwards. If attaching succeeds,
> the link becomes one of the common owner
On 8/23/2020 8:36 AM, Mark Thompson wrote:
> On 23/08/2020 03:02, James Almer wrote:
>> Same logic as tile_cols and tile_rows, this saves CBS users having to
>> recalculate these values, and makes them available for all frames
>> within a
>> Temporal Unit.
>>
>> Signed-off-by: James Almer
>> ---
>
On 6/12/20, Gavin Smith wrote:
>
> On 12/06/2020 16:58, Gavin Smith wrote:
>> On 12/06/2020 12:59, Timo Rothenpieler wrote:
>>> On 12.06.2020 00:31, Gavin Smith wrote:
On 11/06/2020 22:03, Timo Rothenpieler wrote:
> On 11.06.2020 22:27, Gavin Smith wrote:
>> This is to address tr
On Sun, 23 Aug 2020, lance.lmw...@gmail.com wrote:
On Sun, Aug 23, 2020 at 10:17:01AM +0200, Marton Balint wrote:
On Sun, 23 Aug 2020, lance.lmw...@gmail.com wrote:
> On Sat, Aug 22, 2020 at 07:27:30PM +0200, Marton Balint wrote:
> >
> >
> > On Sat, 22 Aug 2020, lance.lmw...@gmail.com wr
On Mon, Aug 17, 2020 at 16:21:23 -0700, Philip Langdale wrote:
> The latest builds of glslang introduce new libraries that need to be
> linked for all symbols to be fully resolved.
>
> Presumably, requiring these new libraries breaks building against
> older glslang releases? What range are we comm
My GSoC project is “ABR meets FFmpeg” which aim to add an ABR module to
FFmpeg.
Report in brief:
Describe my work briefly
- Implement ABR module for FFmpeg: ffabr protocol.
- Use ffabr protocol in hls, support a switch mechanism.
- Test the hls with abr in ffplay.
What is done
- Added a inter
From: spartazhc
Add abr option, ffplay can play hls using abr by:
ffplay -i http://xxx/master.m3u8 -abr
Structure ABRList is added to save stream type and index, it is
used to allow packet_queue_put function to put pkt which from same
type(for example: video pkt) but different stream index to qu
From: spartazhc
For example, ffabr:crypto+http://xxx.m3u8
Signed-off-by: spartazhc
---
libavformat/hls.c | 16
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/libavformat/hls.c b/libavformat/hls.c
index 4e760f8e8a..37a5a017b1 100644
--- a/libavformat/hls.c
+++ b
From: spartazhc
HLS use abr may switch streams, ffplay should allow pkt whic same type
(for example: video pkt) but different stream index to queue.
But in the begining, avformat_find_stream_info() will add pkts to
packet_buffer from all streams to find info. This will cause problem
because thes
From: spartazhc
Add an AVDictionary option "abr_initial", which could be used to send
message to ffplay. Currently, the first entry "abr_init_duration" is
added.
Signed-off-by: spartazhc
---
doc/demuxers.texi | 3 +++
libavformat/hls.c | 14 ++
2 files changed, 17 insertions(+)
d
From: spartazhc
Add abr module for hls/dash.
Signed-off-by: spartazhc
v1 fixed:
1. add an "ff" prefix to the protocol name to mark it internal.
2. use 1.2f for float constant 1.2.
3. simplify abr_seek for we just need AVSEEK_SIZE only.
v2 fixed:
1. fix error return
2. simplify abr_seek
v3 fi
From: spartazhc
add ffabr protocol to the default_whitelist of http and https
Signed-off-by: spartazhc
---
libavformat/http.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/libavformat/http.c b/libavformat/http.c
index 6c39da1a8b..88e69114a2 100644
--- a/libavformat/ht
From: spartazhc
When abr is enable, it will take over the task to call http to
download segments, and will return a switch-request for hls to
switch streams.
For reason not to waste segments that have been downloaded,
switch will become effective after old segments is used out.
Abr cannot work wi
Hej igen,
On Fri, Aug 21, 2020 at 12:19:06 +0300, Martin Storsjö wrote:
> LGTM, with one small nit:
[...]
> > {"listen", "Listen for incoming connections",
> > offsetof(pstruct, options_field . listen),AV_OPT_TYPE_INT, { .i64 = 0
> > }, 0, 1, .flags = TLS_OPTFL }, \
> > +{"h
On 23/08/2020 03:02, James Almer wrote:
Same logic as tile_cols and tile_rows, this saves CBS users having to
recalculate these values, and makes them available for all frames within a
Temporal Unit.
Signed-off-by: James Almer
---
Not a fan of adding more derived values to "raw" structs, but mu
On 8/21/20, Paul B Mahol wrote:
> Overall speed changes for 1920x1080, yuv422p10le, 60fps from: 0.19x to
> 0.343x
>
> Signed-off-by: Paul B Mahol
Will apply soon.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo
From: Gautam Ramakrishnan
This patch fixes tag tree coding for JPEG2000
encoder.
---
libavcodec/j2kenc.c | 41 -
libavcodec/jpeg2000.c | 1 +
libavcodec/jpeg2000.h | 1 +
3 files changed, 26 insertions(+), 17 deletions(-)
diff --git a/libavcodec/j2ken
From: Gautam Ramakrishnan
This patch makes the ff_jpeg2000_cleanup
function take in an extra parameter which
indicates whether it is called from the
encoder or decoder.
---
libavcodec/j2kenc.c | 2 +-
libavcodec/jpeg2000.c| 2 +-
libavcodec/jpeg2000.h| 2 +-
libavcodec/jpeg2000dec.c
From: Gautam Ramakrishnan
This patch makes the tag_tree_zero() and tag_tree_size()
functions non static and callable from other files.
---
libavcodec/jpeg2000.c | 12 ++--
libavcodec/jpeg2000.h | 3 +++
2 files changed, 9 insertions(+), 6 deletions(-)
diff --git a/libavcodec/jpeg2000.c
From: Gautam Ramakrishnan
This patch allows setting a compression ratio and to
set multiple layers. The user has to input a compression
ratio for each layer.
The per layer compression ration can be set as follows:
-layer_rates "r1,r2,...rn"
for to create 'n' layers.
---
libavcodec/j2kenc.c
Michael Niedermayer (12020-08-23):
> > +if (buf->pts == AV_NOPTS_VALUE) {
> > +av_log(ctx, AV_LOG_ERROR, "Frame without timestamp on input %d\n",
> > in_no);
> > +return AVERROR(EINVAL);
> > +}
>
> what if the timestamp is "AV_NOPTS_VALUE + 1" ?
> would that work or overfl
Andreas Rheinhardt (12020-08-23):
> The result of the last check in code like
>
> if (p)
> foo
> else
> p = av_mallocz(sizeof(*p));
>
> if (p->ptr)
> bar
> else
> bar2
>
> is known in advance if the else branch of the first check was taken
> because av_mallocz() returns zeroed bu
Andreas Rheinhardt (12020-08-23):
> Parsing labeled outputs involves a check for an already known match
> (a labeled input with the same name) to pair them together. If yes,
> it is attempted to create a link between the two filters; in this case
> the AVFilterInOuts have fulfilled their purpose an
Andreas Rheinhardt (12020-08-23):
> The AVFilterInOuts normally get freed in init_output_filter() when
> the corresponding streams get created; yet if an error happens before
> one reaches said point, they leak. Therefore this commit makes
> ffmpeg_cleanup free them, too.
>
> Fixes ticket #8267.
>
Andreas Rheinhardt (12020-08-23):
> parse_filter() did not check the return value of av_get_token() for
> success; in case name (the name of a filter) was NULL, one got a
> segfault in av_strlcpy() (called from create_filter()).
>
> Signed-off-by: Andreas Rheinhardt
> ---
> libavfilter/graphpars
Andreas Rheinhardt (12020-08-23):
> This happened in parse_link_name() if there was a '[' without matching
> ']'. While this is not undefined behaviour (pointer arithmetic one
> beyond the end of an array works fine as long as there are no accesses),
> it is potentially dangerous. It currently isn'
ff_set_common_formats() is currently only called after
graph_check_validity(), guaranteeing that inputs and outputs
are connected.
If we want to support configuring partially-connected graphs,
we will have a lot of redesign to do anyway.
Fix CID 1466262 / 1466263.
Signed-off-by: Nicolas George
-
The AVFilterInOuts normally get freed in init_output_filter() when
the corresponding streams get created; yet if an error happens before
one reaches said point, they leak. Therefore this commit makes
ffmpeg_cleanup free them, too.
Fixes ticket #8267.
Signed-off-by: Andreas Rheinhardt
---
fftool
Parsing labeled outputs involves a check for an already known match
(a labeled input with the same name) to pair them together. If yes,
it is attempted to create a link between the two filters; in this case
the AVFilterInOuts have fulfilled their purpose and are freed. Yet if
creating the link fail
The result of the last check in code like
if (p)
foo
else
p = av_mallocz(sizeof(*p));
if (p->ptr)
bar
else
bar2
is known in advance if the else branch of the first check was taken
because av_mallocz() returns zeroed buffers. Therefore the above snippet
can be simplified by moving
parse_filter() did not check the return value of av_get_token() for
success; in case name (the name of a filter) was NULL, one got a
segfault in av_strlcpy() (called from create_filter()).
Signed-off-by: Andreas Rheinhardt
---
libavfilter/graphparser.c | 7 +++
1 file changed, 7 insertions(+
1 - 100 of 107 matches
Mail list logo
