Re: [FFmpeg-devel] [PATCH v3 2/3] avcodec/sanm: fobj left/top are signed

On Mon, 17 Feb 2025, Manuel Lauss wrote: The left and top parameters of an FOBJ are signed values. Signed-off-by: Manuel Lauss --- v3: change the bytestream accessor to signed too get_xxxu means "unchecked" here (as in not checking for buffer overread), not unsigned. Bytestream API alway

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

Niklas Haas (HE12025-02-17): > For the record, this approach seems to work well - adding a const char *label > to the AVFilterPad struct and then setting it during > avfilter_graph_segment_create_filters() recovers the full label names without > any need for shenanigans. Thanks for looking into it

Re: [FFmpeg-devel] [PATCH 05/11] chromaber_vulkan: fix use of texture()

On 17/02/2025 19:31, Lynne wrote: While the rest of texture() uses were fine, this one was not. --- libavfilter/vf_chromaber_vulkan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavfilter/vf_chromaber_vulkan.c b/libavfilter/vf_chromaber_vulkan.c index 65b53afd64..f8c

Re: [FFmpeg-devel] [PATCH 08/11] gblur_vulkan: port to imageLoad()

On 17/02/2025 19:31, Lynne wrote: --- libavfilter/vf_gblur_vulkan.c | 21 ++--- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/libavfilter/vf_gblur_vulkan.c b/libavfilter/vf_gblur_vulkan.c index 101c8ffd8c..bc4c2217b5 100644 --- a/libavfilter/vf_gblur_vulkan.c

Re: [FFmpeg-devel] [PATCH] avfilter/interlace_vulkan: add interlace_vulkan filter

On 17/02/2025 17:19, Niklas Haas wrote: From: Niklas Haas This is a Vulkan-accelerated version of the existing interlace filter. --- configure | 1 + doc/filters.texi | 2 +- libavfilter/Makefile | 1 + libavfilter/allfilters.c

Re: [FFmpeg-devel] [PATCH v5 1/6] libavcodec: Add generic metadata injection using AV_PKT_DATA_METADATA_UPDATE

Le lun. 17 févr. 2025 à 13:10, Lynne a écrit : > > On 17/02/2025 17:19, Romain Beauxis wrote: > > libavcodec/decode.c: intercept `AV_PKT_DATA_METADATA_UPDATE` packet extra > > data, > > attach them to the next decoded frame. > > > > The metadata needs to be cached because there is no guarantee th

[FFmpeg-devel] [PATCH 2/2] vulkan: fix crash in ff_vk_shader_free

The issue is if a shader fails to compile, shd->desc_layout would not exist, but nb_descriptor_sets would. --- libavutil/vulkan.c | 10 ++ 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/libavutil/vulkan.c b/libavutil/vulkan.c index d3ada59334..a868a3b90d 100644 --- a/libavut

[FFmpeg-devel] [PATCH 1/2] avgblur_vulkan: fix duplicated variable error when planes=0

--- libavfilter/vf_avgblur_vulkan.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/libavfilter/vf_avgblur_vulkan.c b/libavfilter/vf_avgblur_vulkan.c index 5a7012531b..156278dd78 100644 --- a/libavfilter/vf_avgblur_vulkan.c +++ b/libavfilter/vf_avgblur_vulkan.c @@ -127,6 +127

Re: [FFmpeg-devel] [PATCH v5 1/6] libavcodec: Add generic metadata injection using AV_PKT_DATA_METADATA_UPDATE

On 17/02/2025 17:19, Romain Beauxis wrote: libavcodec/decode.c: intercept `AV_PKT_DATA_METADATA_UPDATE` packet extra data, attach them to the next decoded frame. The metadata needs to be cached because there is no guarantee that each packet will generate a decoded frame immediately. `AV_PKT_DAT

[FFmpeg-devel] [PATCH v5 1/6] libavcodec: Add generic metadata injection using AV_PKT_DATA_METADATA_UPDATE

libavcodec/decode.c: intercept `AV_PKT_DATA_METADATA_UPDATE` packet extra data, attach them to the next decoded frame. The metadata needs to be cached because there is no guarantee that each packet will generate a decoded frame immediately. `AV_PKT_DATA_METADATA_UPDATE` does not seem used in `lib

[FFmpeg-devel] [PATCH v2 2/2] libavformat/rtsp: Free memory allocated for temporary variables while processing sdp info

--- libavformat/rtsp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavformat/rtsp.c b/libavformat/rtsp.c index 0c65f8d1a4..da733938bc 100644 --- a/libavformat/rtsp.c +++ b/libavformat/rtsp.c @@ -478,6 +478,7 @@ static void sdp_parse_line(AVFormatContext *s, SDPParseState *s1,

[FFmpeg-devel] [PATCH v2 1/2] libavformat/rtsp: Make source specific multicast work for rtsp streams

by first changing the RTSPSource to track the destination address obtained from the source filter. For each RTSPStream, only add the source filter from the sdp if sdp_ip string matches source-filter's destination address. Before issuing the setup request, change the lower_transport to multicast if

[FFmpeg-devel] [PATCH 02/11] bwdif_vulkan: convert to storage images

texture() uses bilinear scaling; imageLoad() accesses the image directly. The reason why texture() was used throughout Vulkan filters is that back when they were written, they were targetting old Intel hardware, which had a texel cache only for sampled images. These days, GPUs have a generic cache

[FFmpeg-devel] [PATCH 08/11] gblur_vulkan: port to imageLoad()

--- libavfilter/vf_gblur_vulkan.c | 21 ++--- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/libavfilter/vf_gblur_vulkan.c b/libavfilter/vf_gblur_vulkan.c index 101c8ffd8c..bc4c2217b5 100644 --- a/libavfilter/vf_gblur_vulkan.c +++ b/libavfilter/vf_gblur_vulkan.c @@

[FFmpeg-devel] [PATCH 11/11] overlay_vulkan: port to imageLoad()

--- libavfilter/vf_overlay_vulkan.c | 26 +++--- 1 file changed, 11 insertions(+), 15 deletions(-) diff --git a/libavfilter/vf_overlay_vulkan.c b/libavfilter/vf_overlay_vulkan.c index 7d5c201cb5..3c6fb96014 100644 --- a/libavfilter/vf_overlay_vulkan.c +++ b/libavfilter/vf_over

[FFmpeg-devel] [PATCH 10/11] blend_vulkan: port to imageLoad()

--- libavfilter/vf_blend_vulkan.c | 23 +-- 1 file changed, 9 insertions(+), 14 deletions(-) diff --git a/libavfilter/vf_blend_vulkan.c b/libavfilter/vf_blend_vulkan.c index 58a8f24d5c..57cf3c696b 100644 --- a/libavfilter/vf_blend_vulkan.c +++ b/libavfilter/vf_blend_vulkan.c @

[FFmpeg-devel] [PATCH 09/11] transpose_vulkan: port to imageLoad()

--- libavfilter/vf_transpose_vulkan.c | 20 +++- 1 file changed, 7 insertions(+), 13 deletions(-) diff --git a/libavfilter/vf_transpose_vulkan.c b/libavfilter/vf_transpose_vulkan.c index 0169804fbb..3fe2d11cb2 100644 --- a/libavfilter/vf_transpose_vulkan.c +++ b/libavfilter/vf_tr

[FFmpeg-devel] [PATCH 05/11] chromaber_vulkan: fix use of texture()

While the rest of texture() uses were fine, this one was not. --- libavfilter/vf_chromaber_vulkan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavfilter/vf_chromaber_vulkan.c b/libavfilter/vf_chromaber_vulkan.c index 65b53afd64..f8cfab1480 100644 --- a/libavfilter/vf_ch

[FFmpeg-devel] [PATCH 07/11] gblur_vulkan: port to imageLoad()

--- libavfilter/vf_gblur_vulkan.c | 20 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/libavfilter/vf_gblur_vulkan.c b/libavfilter/vf_gblur_vulkan.c index 58251c4a17..101c8ffd8c 100644 --- a/libavfilter/vf_gblur_vulkan.c +++ b/libavfilter/vf_gblur_vulkan.c @@ -

[FFmpeg-devel] [PATCH 06/11] flip_vulkan: port to imageLoad()

--- libavfilter/vf_flip_vulkan.c | 23 --- 1 file changed, 8 insertions(+), 15 deletions(-) diff --git a/libavfilter/vf_flip_vulkan.c b/libavfilter/vf_flip_vulkan.c index 2cf793cfe8..3e2aed0fda 100644 --- a/libavfilter/vf_flip_vulkan.c +++ b/libavfilter/vf_flip_vulkan.c @@ -40

[FFmpeg-devel] [PATCH 03/11] nlmeans_vulkan: switch to imageLoad()

--- libavfilter/vf_nlmeans_vulkan.c | 56 ++--- 1 file changed, 24 insertions(+), 32 deletions(-) diff --git a/libavfilter/vf_nlmeans_vulkan.c b/libavfilter/vf_nlmeans_vulkan.c index 67360dc745..22a2a73eae 100644 --- a/libavfilter/vf_nlmeans_vulkan.c +++ b/libavfilter/

[FFmpeg-devel] [PATCH 04/11] avgblur_vulkan: port to imageLoad()

--- libavfilter/vf_avgblur_vulkan.c | 19 +++ 1 file changed, 7 insertions(+), 12 deletions(-) diff --git a/libavfilter/vf_avgblur_vulkan.c b/libavfilter/vf_avgblur_vulkan.c index e291d5d8e9..5a7012531b 100644 --- a/libavfilter/vf_avgblur_vulkan.c +++ b/libavfilter/vf_avgblur_vulk

[FFmpeg-devel] [PATCH 01/11] vulkan_filter: use GENERAL image layout when no sampler is given

This allows using the generic filter code with STORAGE image inputs. --- libavfilter/vulkan_filter.c | 21 ++--- 1 file changed, 14 insertions(+), 7 deletions(-) diff --git a/libavfilter/vulkan_filter.c b/libavfilter/vulkan_filter.c index bdbebb3cb2..ecb24711f3 100644 --- a/libavf

[FFmpeg-devel] [PATCH v5 5/6] Parse comments from secondary chained ogg/opus streams

libavformat/oggparseopus.c: Parse comments from secondary chained streams header packet. This is the same processing as with ogg/flac except that is done in the `opus_header` handler. This is because, in the case of opus streams, the header is currently parsed again on secondary chained streams s

[FFmpeg-devel] [PATCH v5 4/6] tests: Add chained ogg/flac stream dump test.

Implements fate test for decoding chainged ogg/flac metadata. * Demuxing is not modified * Only change is added frame metadata. Samples are here: https://www.dropbox.com/scl/fo/xrtrna2rxr1j354hrtymq/AGwemlxHYecBLNmQ8Fsy--4?rlkey=lzilr4m9w4gfdqygoe172vvy8&dl=0 Before changes: Stream ID: 0, codec

Re: [FFmpeg-devel] [PATCH v5 0/6] Properly decode ogg metadata in ogg/flac and ogg/opus chained bitstreams

Le lun. 17 févr. 2025 à 10:19, Romain Beauxis a écrit : > > This is a series of patches to allow proper decoding of ogg metadata in > chained > `ogg/flac` and `ogg/opus` streams. I forgot to add: this patch set is also available for review at: https://code.ffmpeg.org/FFmpeg/FFmpeg/pulls/6 > ##

[FFmpeg-devel] [PATCH v5 2/6] tests: Add stream dump test API util.

This is the main test utility used in subsequent new FATE tests. It dumps DTS/PTS of packets and decoded frames along with their metadata. --- tests/Makefile| 1 + tests/api/Makefile| 2 +- tests/api/api-dump-stream-meta-test.c | 169 +++

Re: [FFmpeg-devel] [PATCH] avfilter/dnn: add zero-shot image classification using CLIP models

Hello Yejun Guo, yes i can do that and submit it in another patch. Do you also have some feedback for the clip backend? I have already made some small changes (cuda accel and new preprocessing) that i will submit along with the other patch, but i would like to hear your thoughts. Thanks _

[FFmpeg-devel] [PATCH] avfilter/interlace_vulkan: add interlace_vulkan filter

From: Niklas Haas This is a Vulkan-accelerated version of the existing interlace filter. --- configure | 1 + doc/filters.texi | 2 +- libavfilter/Makefile | 1 + libavfilter/allfilters.c | 1 + libavfilter/vf_interlace_vulka

[FFmpeg-devel] [PATCH v5 6/6] tests: Add chained ogg/opus stream dump test.

Implements fate test for decoding chainged ogg/opus metadata. * Demuxing is not modified * Only change is added frame metadata. Samples are here: https://www.dropbox.com/scl/fo/xrtrna2rxr1j354hrtymq/AGwemlxHYecBLNmQ8Fsy--4?rlkey=lzilr4m9w4gfdqygoe172vvy8&dl=0 Before changes: Stream ID: 0, codec

[FFmpeg-devel] [PATCH v5 3/6] Parse ogg/flac comments in new ogg packets

libavformat/oggparseflac.c: Parse ogg/flac comments in new ogg packets, add them to ogg stream new_metadata. This is pretty straight forward and taps into a mechanism already in place in libavformat/oggdec.c --- libavformat/oggparseflac.c | 28 1 file changed, 28 in

[FFmpeg-devel] [PATCH v5 0/6] Properly decode ogg metadata in ogg/flac and ogg/opus chained bitstreams

This is a series of patches to allow proper decoding of ogg metadata in chained `ogg/flac` and `ogg/opus` streams. ## Changes since last version: - Opus processing moved back to `opus_header` to make sure that opus demuxing is not impacted by those changes. - Generic decoding method implemented

Re: [FFmpeg-devel] [PATCH 1/2] libavformat/rtsp: Make source specific multicast work for rtsp streams

I'm not sure how this happened. I'll check my Thunderbird settings. For now, I have attached the patches. On Mon, Feb 17, 2025 at 3:46 AM Michael Niedermayer wrote: > On Fri, Feb 14, 2025 at 12:08:24PM -0500, Rashad Tatum wrote: > > by first changing the RTSPSource to track the destination addre

[FFmpeg-devel] [PATCH 2/2] avformat/iamf_parse: ensure there's at most one of each parameter types in audio elements

Should prevent potential memory leaks on invalid files. Signed-off-by: James Almer --- libavformat/iamf_parse.c | 14 +++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/libavformat/iamf_parse.c b/libavformat/iamf_parse.c index f71ea5315b..db40ae37ab 100644 --- a/libavf

[FFmpeg-devel] [PATCH 1/2] avformat/iamf_parse: add missing constrains for num_parameters in audio_element_oub()

Fixes ticket #11475. Signed-off-by: James Almer --- libavformat/iamf_parse.c | 6 ++ 1 file changed, 6 insertions(+) diff --git a/libavformat/iamf_parse.c b/libavformat/iamf_parse.c index 316093b35f..f71ea5315b 100644 --- a/libavformat/iamf_parse.c +++ b/libavformat/iamf_parse.c @@ -733,6 +

[FFmpeg-devel] [PATCH] avfilter/libplacebo: use a transparent default fillcolor

From: Niklas Haas This will leave the background transparent when the target format contains an alpha channel, without affecting the behavior on formats without one. --- doc/filters.texi| 2 +- libavfilter/vf_libplacebo.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

On Mon, 17 Feb 2025 13:42:47 +0100 Nicolas George wrote: > Niklas Haas (HE12025-02-17): > > From: Niklas Haas > > > > This developer tool is especially handy when debugging filter graph > > auto-negotiation, although it can be useful in whatever scenario to > > get a canonical dump of the fully s

[FFmpeg-devel] [PATCH 2/2] Added support for direct RGB input to AMF encoder

--- libavcodec/amfenc.c | 7 +++ libavutil/hwcontext_amf.c | 2 ++ 2 files changed, 9 insertions(+) diff --git a/libavcodec/amfenc.c b/libavcodec/amfenc.c index 0fa3123cac..63b83a625b 100644 --- a/libavcodec/amfenc.c +++ b/libavcodec/amfenc.c @@ -117,6 +117,13 @@ const enum AVPixelForma

[FFmpeg-devel] [PATCH 1/2] Enable 10-bit output for RGB input in AMF HEVC and AV1

encoders --- libavcodec/amfenc.h | 1 + libavcodec/amfenc_av1.c | 46 +++--- libavcodec/amfenc_hevc.c | 84 +--- 3 files changed, 84 insertions(+), 47 deletions(-) diff --git a/libavcodec/amfenc.h b/libavcodec/amfenc.h index c9d8a74fd1..

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

On Mon, 17 Feb 2025 13:40:49 + Soft Works wrote: > Well, then the question would be when or whether this future will become a > reality? > > Or maybe the solution rather lies in the past? > > IIRC, I had already submitted a patchset with a complete solution: > > - It includes absolutely all

Re: [FFmpeg-devel] [RFC] Experiment: enable github pull requests

Le 16 février 2025 21:20:59 GMT+02:00, Soft Works a écrit : > > >> -Original Message- >> From: ffmpeg-devel On Behalf Of Rémi >> Denis-Courmont >> Sent: Sonntag, 16. Februar 2025 18:59 >> To: ffmpeg-devel@ffmpeg.org >> Subject: Re: [FFmpeg-devel] [RFC] Experiment: enable github pull re

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

> -Original Message- > From: ffmpeg-devel On Behalf Of > Nicolas George > Sent: Montag, 17. Februar 2025 13:43 > To: FFmpeg development discussions and patches de...@ffmpeg.org> > Subject: Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add - > print_filter_graph option > > Niklas H

[FFmpeg-devel] ffmpeg developer contracting request

I am looking for an ffmpeg developer to help create some utility software. A terminal application to do the following: - parse an AVID ALE file and extract metadata - receive frame extraction / blanking infromation and  burn-in flags via args - batch process a folder of exported MOVs (re-wra

[FFmpeg-devel] [PATCH v3 3/3] avcodec/sanm: add smush codec23 decoder

This codec alternatingly skips and changes existing pixels. A second 16bit parameter in the FOBJ header indicates how to do the pixel changes: either by specifying a LUT in the codec datastream or by adding a constant value to the pixel. Signed-off-by: Manuel Lauss --- v3: updates due to changes

[FFmpeg-devel] [PATCH v3 2/3] avcodec/sanm: fobj left/top are signed

The left and top parameters of an FOBJ are signed values. Signed-off-by: Manuel Lauss --- v3: change the bytestream accessor to signed too v2: no changes libavcodec/sanm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libavcodec/sanm.c b/libavcodec/sanm.c index a4f0a28

[FFmpeg-devel] [PATCH v3 1/3] avcodec/sanm: ignore unknown codecs in FOBJs

Don't error out, just ignore unknown codec numbers and pretend decode succeeded. This is useful for older LucasArts titles which stack a lot of different FOBJs with different codecs into a single frame. Signed-off-by: Manuel Lauss --- v3: no changes v2: Mark frame as corrupt, suggested by Marton

Re: [FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

Niklas Haas (HE12025-02-17): > From: Niklas Haas > > This developer tool is especially handy when debugging filter graph > auto-negotiation, although it can be useful in whatever scenario to > get a canonical dump of the fully settled filter graph. > > To make the result slightly more useful, we

[FFmpeg-devel] [PATCH] fftools/ffmpeg_filter: add -print_filter_graph option

From: Niklas Haas This developer tool is especially handy when debugging filter graph auto-negotiation, although it can be useful in whatever scenario to get a canonical dump of the fully settled filter graph. To make the result slightly more useful, we omit buffersrc/buffersink filters and inst

Re: [FFmpeg-devel] [PATCH 1/2] tests/checkasm/sw_rgb: Added {yuyv, uyvy}toyuv{420, 422} test cases

On Tue, 11 Feb 2025, Krzysztof Pyrkosz via ffmpeg-devel wrote: Splitting the previous patch into two. I noticed that on my x86 box, one of the newly added tests fail: MMXEXT: uyvytoyuv420_mmxext (sw_rgb.c:126) yuyvtoyuv420_mmxext (sw_rgb.c:126) - sw_rgb.uyvytoyuv [FAILED] SSE2, AV

Re: [FFmpeg-devel] [PATCH] swscale/aarch64/rgb2rgb_neon: Implemented {yuyv, uyvy}toyuv{420, 422}

On Thu, 13 Feb 2025, Krzysztof Pyrkosz via ffmpeg-devel wrote: This patch succesfully passes the github pipeline. The previous one, which adds tests fails only the first check on linux x86, probably because of that mmx issue. The tiny patch in the second email chain (the one about right shift b

Re: [FFmpeg-devel] [PATCH 1/2] tests/checkasm/sw_rgb: Added {yuyv, uyvy}toyuv{420, 422} test cases

On Mon, 17 Feb 2025, Martin Storsjö wrote: On Tue, 11 Feb 2025, Krzysztof Pyrkosz via ffmpeg-devel wrote: Splitting the previous patch into two. I noticed that on my x86 box, one of the newly added tests fail: MMXEXT: uyvytoyuv420_mmxext (sw_rgb.c:126) yuyvtoyuv420_mmxext (sw_rgb.c:126) -

Re: [FFmpeg-devel] [PATCH] swscale/aarch64/rgb24toyv12: skip early right shift by 2

On Tue, 11 Feb 2025, Krzysztof Pyrkosz via ffmpeg-devel wrote: It's a minor improvement that shaves off 5-8% from the execution time. Instead of shifting by 2 right away and by 7 soon after, shift by 9 one time. Times before and after: A78: rgb24toyv12_16_200_neon:

Re: [FFmpeg-devel] [PATCH 1/2] libavformat/rtsp: Make source specific multicast work for rtsp streams

On Fri, Feb 14, 2025 at 12:08:24PM -0500, Rashad Tatum wrote: > by first changing the RTSPSource to track the destination address > obtained from the source filter. For each RTSPStream, only add the source > filter from the sdp if sdp_ip string matches source-filter's destination > address. > > Be