[FFmpeg-devel] [PATCH] hwcontext_opencl: choose the first device if multiple devices are available

From: Haihao Xiang This makes '-init_hw_device opencl' work in a multiple-device system. Signed-off-by: Haihao Xiang --- doc/ffmpeg.texi | 5 + libavutil/hwcontext_opencl.c | 9 - 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/doc/ffmpeg.texi

[FFmpeg-devel] [PATCH 3/3] libavfilter/vf_dnn_detect: Use class confidence to filt boxes

From: Wenbin Chen Use class confidence instead of box_score to filt boxes, which is more accurate. Class confidence is obtained by multiplying class probability distribution and box_score. Signed-off-by: Wenbin Chen --- libavfilter/vf_dnn_detect.c | 6 +++--- 1 file changed, 3 insertions(+),

[FFmpeg-devel] [PATCH 2/3] libavfilter/dnn_interface: use dims to represent shapes

From: Wenbin Chen For detect and classify output, width and height make no sence, so change width, height to dims to represent the shape of tensor. Use layout and dims to get width, height and channel. Signed-off-by: Wenbin Chen --- libavfilter/dnn/dnn_backend_openvino.c | 80

[FFmpeg-devel] [PATCH 1/3] libavfilter/dnn_bakcend_openvino: Add automatic input/output detection

From: Wenbin Chen Now when using openvino backend, user doesn't need to set input/output names in command line. Model ports will be automatically detected. For example: ffmpeg -i input.png -vf \ dnn_detect=dnn_backend=openvino:model=model.xml:input=image:\ output=detection_out -y output.png

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/d3d12va_decode: check existance before assigning a new index

On Do, 2024-01-11 at 14:31 +0800, tong1.wu-at-intel@ffmpeg.org wrote: > From: Tong Wu > > Fixes #10759. > > It can happen in H.264, MPEG2, VC1 that the current frame resource > memory is already in ref_resource. For example, for a interlaced frame, > the same curr memory is passed twice.

Re: [FFmpeg-devel] [PATCH v2 1/2] lavc/d3d12va: remove unused variables

On Ma, 2024-01-15 at 10:34 +0800, tong1.wu-at-intel@ffmpeg.org wrote: > From: Tong Wu > > Signed-off-by: Tong Wu > --- >  libavcodec/d3d12va_av1.c    | 18 +- >  libavcodec/d3d12va_decode.c |  9 ++--- >  libavcodec/d3d12va_h264.c   |  4 >  libavcodec/d3d12va_hevc.c  

[FFmpeg-devel] [PATCH v2 0/2] avcodec/vvcdec: fix valgrind issue for deblock

change since v1: Rebased on top of master. Nuo Mi (2): avcodec/vvcdec: deblock, fix uninitialized values avcodec/vvcdec: misc, constify hor_ctu_edge libavcodec/vvc/vvc_filter_template.c | 242 ++- 1 file changed, 122 insertions(+), 120 deletions(-) -- 2.25.1

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/vvcdec: deblock, fix uninitialized values

On Tue, Jan 16, 2024 at 10:38 PM James Almer wrote: > On 1/8/2024 10:00 PM, Nuo Mi wrote: > > > > > > On Sun, Jan 7, 2024 at 12:43 PM Nuo Mi > > wrote: > > > > see > > > https://fate.ffmpeg.org/report.cgi?slot=x86_64-archlinux-gcc-valgrind=20240105201935 > < >

[FFmpeg-devel] [PATCH v2 1/2] avcodec/vvcdec: deblock, fix uninitialized values

see https://fate.ffmpeg.org/report.cgi?slot=x86_64-archlinux-gcc-valgrind=20240105201935 If tc is zero, the max_len_q, max_len_p are uninitialized. Reported-by: James Almer --- libavcodec/vvc/vvc_filter_template.c | 236 ++- 1 file changed, 119 insertions(+), 117

[FFmpeg-devel] [PATCH v2 2/2] avcodec/vvcdec: misc, constify hor_ctu_edge

--- libavcodec/vvc/vvc_filter_template.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/libavcodec/vvc/vvc_filter_template.c b/libavcodec/vvc/vvc_filter_template.c index 845c8a4181..b7eaef5125 100644 --- a/libavcodec/vvc/vvc_filter_template.c +++

[FFmpeg-devel] [PATCH] avfilter/vf_xfade: Update to support transition between two points in single video stream.

This patch replaces my previous patch from Jan. 4 2024, which had bugs. Signed-off-by: Aidan O'Connor --- libavfilter/vf_xfade.c | 260 - 1 file changed, 207 insertions(+), 53 deletions(-) diff --git a/libavfilter/vf_xfade.c b/libavfilter/vf_xfade.c

Re: [FFmpeg-devel] [PATCH 4/8] fftools/ffmpeg_demux: set options on codec parameters rather than decoder

On 1/16/2024 4:49 PM, Anton Khirnov wrote: Quoting James Almer (2024-01-06 15:22:10) This does not fix the issue of the guessed channel layout not being present on the output stream, Well, it wasn't supposed to fix anything. This patch exists as a prerequisite for the next one. I know,

Re: [FFmpeg-devel] [PATCH] avformat/avcodec: Add DTS-UHD demuxer and parser, movenc support.

Updated to latest master changes.--- Begin Message --- Signed-off-by: Roy Funderburk --- Changelog | 1 + configure | 1 + doc/general_contents.texi | 1 + libavformat/Makefile | 1 + libavformat/allformats.c | 1 + libavformat/dtshddec.c| 2

Re: [FFmpeg-devel] [PATCH 6/8] fftools/ffmpeg_demux: add demuxing thread private data

Quoting Stefano Sabatini (2024-01-06 12:44:27) > On date Friday 2024-01-05 17:42:49 +0100, Anton Khirnov wrote: > > To be used for data that never needs to be visible outside of the > > demuxer thread, similarly as was previously done for other components. > > --- > > fftools/ffmpeg_demux.c | 67

Re: [FFmpeg-devel] [PATCH 5/8] fftools/ffmpeg_demux: only allocate the decoder when actually decoding

Quoting Stefano Sabatini (2024-01-06 12:34:53) > On date Friday 2024-01-05 17:42:48 +0100, Anton Khirnov wrote: > > It is not needed otherwise. > > --- > > fftools/ffmpeg_demux.c | 24 > > 1 file changed, 12 insertions(+), 12 deletions(-) > > > > diff --git

Re: [FFmpeg-devel] [PATCH 4/8] fftools/ffmpeg_demux: set options on codec parameters rather than decoder

Quoting James Almer (2024-01-06 15:22:10) > This does not fix the issue of the guessed channel layout not being > present on the output stream, Well, it wasn't supposed to fix anything. This patch exists as a prerequisite for the next one. -- Anton Khirnov

Re: [FFmpeg-devel] [PATCH 3/3] lavc/svq1enc: R-V V ssd_int8_vs_int16

Okay, I updated it in the reply Rémi Denis-Courmont 于2024年1月17日周三 02:04写道： > +vsetvli t0, a2, e8, m2, tu, ma > +vle8.v v0, (a0) > +sub a2, a2, t0 > +vsetvli zero, t0, e16, m4, tu, ma > +vle16.v v8, (a1) > +vsetvli

[FFmpeg-devel] [PATCH] libavformat/libgme: avoid infinite reads of looping files

From: Rebecca Kelly This fixes an issue where certain VGM/VGZ files (and possibly other formats handled by libgme) will report a duration of "N/A" when ffprobed, and will produce an infinite-length audio file when converted with ffmpeg. The behaviour here follows the recommendations in the

Re: [FFmpeg-devel] [PATCH 3/3] lavc/svq1enc: R-V V ssd_int8_vs_int16

Le sunnuntaina 7. tammikuuta 2024, 10.36.23 EET flow gg a écrit : > Alright, I learned a bit more, so should we not consider the internal > implementation? You asked what the reason was for your counter-intuitive observations, and I provided a plausible hypothesis. Nothing more ,nothing less.

Re: [FFmpeg-devel] [PATCH 3/3] lavc/svq1enc: R-V V ssd_int8_vs_int16

+vsetvli t0, a2, e8, m2, tu, ma +vle8.v v0, (a0) +sub a2, a2, t0 +vsetvli zero, t0, e16, m4, tu, ma +vle16.v v8, (a1) +vsetvli zero, t0, e8, m2, tu, ma +vwsub.wv v16, v8, v0 +vsetvli zero,

[FFmpeg-devel] [PATCH 3/3] lavc/h264pred: R-V V pred16x16_dc_8

From 8c5fdbfea42e9ad6ba6e1df5e4ea3c583d59537a Mon Sep 17 00:00:00 2001 From: sunyuechi Date: Tue, 16 Jan 2024 23:57:53 +0800 Subject: [PATCH 3/3] lavc/h264pred: R-V V pred16x16_dc_8 C908 pred16x16_dc_8_c: 2.5 pred16x16_dc_8_rvv_i32: 1.7 --- libavcodec/riscv/h264pred_init.c | 2 ++

[FFmpeg-devel] [PATCH 2/3] lavc/h264pred: R-V V pred16x16_horizontal_8

From 806f84ea5557c4652e48451decc4c679c9485472 Mon Sep 17 00:00:00 2001 From: sunyuechi Date: Tue, 16 Jan 2024 23:56:33 +0800 Subject: [PATCH 2/3] lavc/h264pred: R-V V pred16x16_horizontal_8 C908 pred16x16_horizontal_8_c: 3.0 pred16x16_horizontal_8_rvv_i32: 2.5 ---

[FFmpeg-devel] [PATCH 1/3] lavc/h264pred: R-V V pred16x16_vertical_8

From eaac50d41b3398ef39d1026a7d84480860a1c41e Mon Sep 17 00:00:00 2001 From: sunyuechi Date: Tue, 16 Jan 2024 23:55:33 +0800 Subject: [PATCH 1/3] lavc/h264pred: R-V V pred16x16_vertical_8 C908 pred16x16_vertical_8_c: 1.5 pred16x16_vertical_8_rvv_i32: 1.0 --- libavcodec/h264pred.c|

Re: [FFmpeg-devel] Hardware purchase request: AVX512-capable laptop

Jan 16, 2024, 11:06 by kier...@obe.tv: >> >> A ticket doesn't have durability. >> > > A Ryzen 5 vs Ryzen 7 in the same laptop chassis doesn't change its > durability, it only doubles the laptop's price. > It doubles the number of registers and expands the number of instructions available. The

Re: [FFmpeg-devel] [PATCH 1/4] avcodec/mdec: DC reading for STRv1 is like STRv2

These email issues made me forgot to tell an important thing... That new FPS patch is meant to be applied right after this: avformat/psxstr: fix unknown sector type 00/80 a96d556baaf7cb71babad96238eadb532d02f3a6 i.e. not over the previous FPS patch attempt: avformat/psxstr: basic FPS detection

Re: [FFmpeg-devel] [PATCH 1/4] avcodec/mdec: DC reading for STRv1 is like STRv2

Lol, I told you I was struggling with emClient, it decided to send .eml as .url... Here are the patches, hopefully! [FFmpeg-devel] [PATCH 1/2] avformat/psxstr: infer video FPS using sector LBA https://ffmpeg.org/pipermail/ffmpeg-devel/2024-January/319701.html [FFmpeg-devel] [PATCH 2/2]

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/vvcdec: deblock, fix uninitialized values

On 1/8/2024 10:00 PM, Nuo Mi wrote: On Sun, Jan 7, 2024 at 12:43 PM Nuo Mi > wrote: see https://fate.ffmpeg.org/report.cgi?slot=x86_64-archlinux-gcc-valgrind=20240105201935

[FFmpeg-devel] [PATCH 2/2] avformat/psxstr: update fate checksums

Signed-off-by: aybe --- tests/ref/fate/psx-str-demux | 178 +-- 1 file changed, 89 insertions(+), 89 deletions(-) diff --git a/tests/ref/fate/psx-str-demux b/tests/ref/fate/psx-str-demux index 18a21b0bb6..0a31c44323 100644 --- a/tests/ref/fate/psx-str-demux +++

[FFmpeg-devel] [PATCH 1/2] avformat/psxstr: infer video FPS using sector LBA

Signed-off-by: aybe --- libavformat/psxstr.c | 44 ++-- 1 file changed, 42 insertions(+), 2 deletions(-) diff --git a/libavformat/psxstr.c b/libavformat/psxstr.c index 306a690f52..c718fb6daf 100644 --- a/libavformat/psxstr.c +++ b/libavformat/psxstr.c

Re: [FFmpeg-devel] [PATCH 1/4] avcodec/mdec: DC reading for STRv1 is like STRv2

Modern email clients like to do stuff they're told not to... The link I meant previously: https://github.com/aybe/FFmpeg-PSX-STR-tests -- Original Message -- >From "aybe aybe" >mailto:aybe.one-at-hotmail@ffmpeg.org>> To "FFmpeg development discussions and patches"

Re: [FFmpeg-devel] [PATCH 1/4] avcodec/mdec: DC reading for STRv1 is like STRv2

The new patch appears to work, stuff is synchronized, except: - 30 FPS video still seen as 150FPS by VLC - 15 FPS video emits tons of "[vost#0:0/rawvideo @ 0269b580] Clipping frame in rate conversion by 0.199989" As for the corrupt samples, been able to find the original sources: - the

[FFmpeg-devel] [PATCH v4 2/2] fate: Add tests for aoverlay filter

--- tests/fate/filter-audio.mak| 22 ++ tests/ref/fate/filter-aoverlay-crossfade-d | 224 ++ tests/ref/fate/filter-aoverlay-crossfade-t | 202 tests/ref/fate/filter-aoverlay-default | 259 +

[FFmpeg-devel] [PATCH v4 1/2] avfilter: add audio overlay filter

Co-authored-by: Paul B Mahol Signed-off-by: Harshit Karwal --- doc/filters.texi | 40 +++ libavfilter/Makefile | 1 + libavfilter/af_aoverlay.c | 538 ++ libavfilter/allfilters.c | 1 + 4 files changed, 580 insertions(+) create mode

[FFmpeg-devel] [PATCH v4 0/2] GSoC 2023: Add Audio Overlay Filter

Includes some fixes authored by Paul over the v3 patch I sent earlier, and FATE tests for the filter. Harshit Karwal (2): avfilter: add audio overlay filter fate: Add tests for aoverlay filter doc/filters.texi | 40 ++ libavfilter/Makefile

Re: [FFmpeg-devel] Hardware purchase request: AVX512-capable laptop

On Tue, 16 Jan 2024, 11:50 Paul B Mahol, wrote: > On Tue, Jan 16, 2024 at 11:06 AM Kieran Kunhya wrote: > > > > > > > A ticket doesn't have durability. > > > > > > > A Ryzen 5 vs Ryzen 7 in the same laptop chassis doesn't change its > > durability, it only doubles the laptop's price. > > > >

Re: [FFmpeg-devel] Hardware purchase request: AVX512-capable laptop

On Tue, Jan 16, 2024 at 11:06 AM Kieran Kunhya wrote: > > > > A ticket doesn't have durability. > > > > A Ryzen 5 vs Ryzen 7 in the same laptop chassis doesn't change its > durability, it only doubles the laptop's price. > Are you CEO of FFmpeg or FFlabs now? > > Kieran >

Re: [FFmpeg-devel] Hardware purchase request: AVX512-capable laptop

> > A ticket doesn't have durability. > A Ryzen 5 vs Ryzen 7 in the same laptop chassis doesn't change its durability, it only doubles the laptop's price. Kieran ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org