Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
The second link in my last message is wrong, the correct one with the summary of the media features for Intel GPUs is: https://github.com/intel/media-driver/blob/master/docs/media_features.md On Sun, 11 Dec 2022 at 18:44, wrote: > My question is not related to post-processing or color-space > transformation. It is about how to hardware encode using a pixel format > that provides high color accuracy for screen capture, differently from what > NV12 is capable of offering. > > Regarding hardware encoding support, accordingly to this: > > https://github.com/intel/media-driver/blob/master/README.md > > Hardware support has been added since Ice Lake (2019, 10th generation) to > HEVC and VP9 codecs for 4:4:4 chroma for both 8 and 10 bits modes and also > for both encoding and decoding. > > This other page gives more details: > > https://github.com/intel/media-driver/blob/master/README.md > > Accordingly to the above link , the following 4:4:4 formats are supported > for hardware decoding when using HEVC or VP9: AYUV (8 bits) and Y410 (10 > bits) and the following 4:4:4 formats are supported for hardware encoding > when using HEVC or VP9: AYUV and RGB (8 bits) and Y410 (10 bits). > > I've found out that gstreamer in my system (ubuntu version 1.20.1) is > able to perform 4:4:4 10 bits encoding when using the Y410 format and the > following command (ffprobe for the generated file reports pixel format > as yuv444p10le): > > > gst-launch-1.0 ximagesrc display-name=:0.0 show-pointer=false > use-damage=false num-buffers=2400 ! video/x-raw,framerate=20/1 ! > videoconvert ! video/x-raw,format=RGBx ! vaapipostproc ! > video/x-raw,format=Y410 ! vaapih265enc rate-control=cqp quality-level=1 > quality-factor=18 ! video/x-h265 ! h265parse ! matroskamux ! filesink > location=desktop.mkv > > > and hardware decoding of the above generated file can be achieved with the > following command: > > > gst-play-1.0 desktop.mkv > > > Unfortunately, gstreamer version I have has no support for the AYUV pixel > format with hardware acceleration, therefore I could not test 8 bits > hardware encoding and decoding, but at least for 10 bits it is working > properly and colors looks just great. > > === > > I believe it would be great to add support in ffmpeg for hardware encoding > and decoding for at least the following 4:4:4 pixel formats: AYUV, RGB > and Y410 as this is available in Intel hardware since 2019. > > > > On Sat, 10 Dec 2022 at 17:00, Carl Eugen Hoyos wrote: > > > I wanted to write "no hardware encoders support full chroma" and > > "no hardware encoders support RGB" - Wenbin's answer seems > > to indicate that I was wrong but in any case the qsv code in FFmpeg > > currently neither allows full-chroma encoding nor RGB encoding. > > > > I don't know if mxf (the library used for qsv encoding) support > > full-chroma and RGB. > > > > No idea how your question could be related to post-processing > > or colour-space transformation. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
My question is not related to post-processing or color-space transformation. It is about how to hardware encode using a pixel format that provides high color accuracy for screen capture, differently from what NV12 is capable of offering. Regarding hardware encoding support, accordingly to this: https://github.com/intel/media-driver/blob/master/README.md Hardware support has been added since Ice Lake (2019, 10th generation) to HEVC and VP9 codecs for 4:4:4 chroma for both 8 and 10 bits modes and also for both encoding and decoding. This other page gives more details: https://github.com/intel/media-driver/blob/master/README.md Accordingly to the above link , the following 4:4:4 formats are supported for hardware decoding when using HEVC or VP9: AYUV (8 bits) and Y410 (10 bits) and the following 4:4:4 formats are supported for hardware encoding when using HEVC or VP9: AYUV and RGB (8 bits) and Y410 (10 bits). I've found out that gstreamer in my system (ubuntu version 1.20.1) is able to perform 4:4:4 10 bits encoding when using the Y410 format and the following command (ffprobe for the generated file reports pixel format as yuv444p10le): gst-launch-1.0 ximagesrc display-name=:0.0 show-pointer=false use-damage=false num-buffers=2400 ! video/x-raw,framerate=20/1 ! videoconvert ! video/x-raw,format=RGBx ! vaapipostproc ! video/x-raw,format=Y410 ! vaapih265enc rate-control=cqp quality-level=1 quality-factor=18 ! video/x-h265 ! h265parse ! matroskamux ! filesink location=desktop.mkv and hardware decoding of the above generated file can be achieved with the following command: gst-play-1.0 desktop.mkv Unfortunately, gstreamer version I have has no support for the AYUV pixel format with hardware acceleration, therefore I could not test 8 bits hardware encoding and decoding, but at least for 10 bits it is working properly and colors looks just great. === I believe it would be great to add support in ffmpeg for hardware encoding and decoding for at least the following 4:4:4 pixel formats: AYUV, RGB and Y410 as this is available in Intel hardware since 2019. On Sat, 10 Dec 2022 at 17:00, Carl Eugen Hoyos wrote: > I wanted to write "no hardware encoders support full chroma" and > "no hardware encoders support RGB" - Wenbin's answer seems > to indicate that I was wrong but in any case the qsv code in FFmpeg > currently neither allows full-chroma encoding nor RGB encoding. > > I don't know if mxf (the library used for qsv encoding) support > full-chroma and RGB. > > No idea how your question could be related to post-processing > or colour-space transformation. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
Am Do., 8. Dez. 2022 um 20:32 Uhr schrieb : > Stream #0:0: Video: hevc (Rext), gbrp Completely unrelated: Do all your video players play above file? > ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose > -f x11grab -draw_mouse 0 -framerate 20 -probesize 42M > -video_size 1920x1200 -thread_queue_size 1000 -vcodec rawvideo > -i ":0.0" -an -vf 'hwupload=extra_hw_frames=64,scale_qsv=format=bgra' > -c:v hevc_qsv -preset veryslow -global_quality 17 -look_ahead 1 > -y /tmp/output.mkv (-vcodec rawvideo looks wrong) When asking for support on this mailing list, always provide the command line you tested together with the complete, uncut console output. > As you can see, in the first example, libx265 SW encoder based, the output > is generated with the "gbrp" colorspace while in the second example, > hevc_qsv HW encoder based, the output is generated with the "yuvj420p" > pixel format. I wanted to write "no hardware encoders support full chroma" and "no hardware encoders support RGB" - Wenbin's answer seems to indicate that I was wrong but in any case the qsv code in FFmpeg currently neither allows full-chroma encoding nor RGB encoding. I don't know if mxf (the library used for qsv encoding) support full-chroma and RGB. No idea how your question could be related to post-processing or colour-space transformation. Carl Eugen ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
On Fri, 9 Dec 2022 at 09:16, Chen, Wenbin < wenbin.chen-at-intel@ffmpeg.org> wrote: > > > Output of 'ffmpeg -h filter=vpp_qsv' is bellow > > > > > > > > > Filter vpp_qsv > > > Quick Sync Video VPP. > > > Inputs: > > >#0: default (video) > > > Outputs: > > >#0: default (video) > > > vpp_qsv AVOptions: > > >deinterlace ..FV... deinterlace mode: 0=off, > 1=bob, > > > 2=advanced (from 0 to 2) (default 0) > > > bob 1..FV... Bob deinterlace mode. > > > advanced2..FV... Advanced deinterlace > mode. > > >denoise ..FV... denoise level [0, 100] > (from 0 > > > to 100) (default 0) > > >detail..FV... enhancement level [0, > 100] (from > > > 0 to 100) (default 0) > > >framerate..FV... output framerate (from 0 > to > > > DBL_MAX) (default 0/1) > > >procamp ..FV... Enable ProcAmp (from 0 > to 1) > > > (default 0) > > >hue ..FV... ProcAmp hue (from -180 > to 180) > > > (default 0) > > >saturation ..FV... ProcAmp saturation (from > 0 to > > > 10) (default 1) > > >contrast..FV... ProcAmp contrast (from 0 > to 10) > > > (default 1) > > >brightness ..FV... ProcAmp brightness (from > -100 to > > > 100) (default 0) > > >transpose ..FV... set transpose direction > (from -1 > > > to 6) (default -1) > > > cclock_hflip0..FV... rotate counter-clockwise > with > > > horizontal flip > > > clock 1..FV... rotate clockwise > > > cclock 2..FV... rotate counter-clockwise > > > clock_hflip 3..FV... rotate clockwise with > horizontal > > > flip > > > reversal4..FV... rotate by half-turn > > > hflip 5..FV... flip horizontally > > > vflip 6..FV... flip vertically > > >cw ..FV... set the width crop area > > > expression (default "iw") > > >ch ..FV... set the height crop area > > > expression (default "ih") > > >cx ..FV... set the x crop area > expression > > > (default "(in_w-out_w)/2") > > >cy ..FV... set the y crop area > expression > > > (default "(in_h-out_h)/2") > > >w ..FV... Output video width > (default > > > "cw") > > >width ..FV... Output video width > (default > > > "cw") > > >h ..FV... Output video height > (default > > > "w*ch/cw") > > >height ..FV... Output video height > (default > > > "w*ch/cw") > > >format ..FV... Output pixel format > (default > > > "same") > > >async_depth ..FV... Internal parallelization > depth, > > > the higher the value the higher the latency. (from 0 to INT_MAX) > (default 0) > > >scale_mode..FV... scale mode: 0=auto, > 1=low power, > > > 2=high quality (from 0 to 2) (default 0) > > > > > > > > > I tried the following command with vpp_qsv filter but pixel > > format/colorspace > > > of the output is the same: > > > > > > > > > ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose -f > x11grab > > > -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 > > > -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -vf > > > 'hwupload=extra_hw_frames=64,vpp_qsv=format=bgra' -c:v hevc_qsv - > > preset > > > veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv > > > > > > > > > The ffprobe result for the above command is: > > > > > > > > > Input #0, matroska,webm, from '/tmp/output.mkv': > > > Metadata: > > > ENCODER : Lavf59.27.100 > > > Duration: 00:00:04.85, start: 0.00, bitrate: 50 kb/s > > > Stream #0:0: Video: hevc (Main), yuvj420p(pc, gbr/unknown/unknown, > > > progressive), 1920x1200 [SAR 1:1 DAR 8:5], 20 fps, 20 tbr, 1k tbn > > > Metadata: > > > ENCODER : Lavc59.37.100 hevc_qsv > > > DURATION: 00:00:04.85000 > > > > > > > > > On Thu, 8 Dec 2022 at 17:06, Dennis Mungai wrote: > > In case of RGB input, VPL/MSDK uses 420 chroma format as default for the > purpose of backward compatibility. > I will send patch to enable 444 encoding for RGB input, then you can get > the same output as libx265 SW. > > Thanks > Wenbin > > > > > > > > Hey there, > > > > > > > > I'd recommend using the vpp_qsv filter instead. > > > > > > > > Please provide the output of: > > > > > > > > ffmpeg -h filter=vpp_qsv > > > > > > > > And I'll generate a command line you can use to test and report back. > > Perhaps we can try out libplacebo. That vulkan-based filter should have support for the output format(s)
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
> > Output of 'ffmpeg -h filter=vpp_qsv' is bellow > > > > > > Filter vpp_qsv > > Quick Sync Video VPP. > > Inputs: > >#0: default (video) > > Outputs: > >#0: default (video) > > vpp_qsv AVOptions: > >deinterlace ..FV... deinterlace mode: 0=off, > > 1=bob, > > 2=advanced (from 0 to 2) (default 0) > > bob 1..FV... Bob deinterlace mode. > > advanced2..FV... Advanced deinterlace mode. > >denoise ..FV... denoise level [0, 100] (from 0 > > to 100) (default 0) > >detail..FV... enhancement level [0, 100] > > (from > > 0 to 100) (default 0) > >framerate..FV... output framerate (from 0 to > > DBL_MAX) (default 0/1) > >procamp ..FV... Enable ProcAmp (from 0 to 1) > > (default 0) > >hue ..FV... ProcAmp hue (from -180 to 180) > > (default 0) > >saturation ..FV... ProcAmp saturation (from 0 to > > 10) (default 1) > >contrast..FV... ProcAmp contrast (from 0 to > > 10) > > (default 1) > >brightness ..FV... ProcAmp brightness (from -100 > > to > > 100) (default 0) > >transpose ..FV... set transpose direction (from > > -1 > > to 6) (default -1) > > cclock_hflip0..FV... rotate counter-clockwise with > > horizontal flip > > clock 1..FV... rotate clockwise > > cclock 2..FV... rotate counter-clockwise > > clock_hflip 3..FV... rotate clockwise with > > horizontal > > flip > > reversal4..FV... rotate by half-turn > > hflip 5..FV... flip horizontally > > vflip 6..FV... flip vertically > >cw ..FV... set the width crop area > > expression (default "iw") > >ch ..FV... set the height crop area > > expression (default "ih") > >cx ..FV... set the x crop area expression > > (default "(in_w-out_w)/2") > >cy ..FV... set the y crop area expression > > (default "(in_h-out_h)/2") > >w ..FV... Output video width (default > > "cw") > >width ..FV... Output video width (default > > "cw") > >h ..FV... Output video height (default > > "w*ch/cw") > >height ..FV... Output video height (default > > "w*ch/cw") > >format ..FV... Output pixel format (default > > "same") > >async_depth ..FV... Internal parallelization > > depth, > > the higher the value the higher the latency. (from 0 to INT_MAX) (default 0) > >scale_mode..FV... scale mode: 0=auto, 1=low > > power, > > 2=high quality (from 0 to 2) (default 0) > > > > > > I tried the following command with vpp_qsv filter but pixel > format/colorspace > > of the output is the same: > > > > > > ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose -f x11grab > > -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 > > -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -vf > > 'hwupload=extra_hw_frames=64,vpp_qsv=format=bgra' -c:v hevc_qsv - > preset > > veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv > > > > > > The ffprobe result for the above command is: > > > > > > Input #0, matroska,webm, from '/tmp/output.mkv': > > Metadata: > > ENCODER : Lavf59.27.100 > > Duration: 00:00:04.85, start: 0.00, bitrate: 50 kb/s > > Stream #0:0: Video: hevc (Main), yuvj420p(pc, gbr/unknown/unknown, > > progressive), 1920x1200 [SAR 1:1 DAR 8:5], 20 fps, 20 tbr, 1k tbn > > Metadata: > > ENCODER : Lavc59.37.100 hevc_qsv > > DURATION: 00:00:04.85000 > > > > > > On Thu, 8 Dec 2022 at 17:06, Dennis Mungai wrote: In case of RGB input, VPL/MSDK uses 420 chroma format as default for the purpose of backward compatibility. I will send patch to enable 444 encoding for RGB input, then you can get the same output as libx265 SW. Thanks Wenbin > > > > > Hey there, > > > > > > I'd recommend using the vpp_qsv filter instead. > > > > > > Please provide the output of: > > > > > > ffmpeg -h filter=vpp_qsv > > > > > > And I'll generate a command line you can use to test and report back. > > ___ > > ffmpeg-user mailing list > > ffmpeg-user@ffmpeg.org > > https://ffmpeg.org/mailman/listinfo/ffmpeg-user > > > > To unsubscribe, visit link above, or email > > ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe". ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
Output of 'ffmpeg -h filter=vpp_qsv' is bellow Filter vpp_qsv Quick Sync Video VPP. Inputs: #0: default (video) Outputs: #0: default (video) vpp_qsv AVOptions: deinterlace ..FV... deinterlace mode: 0=off, 1=bob, 2=advanced (from 0 to 2) (default 0) bob 1..FV... Bob deinterlace mode. advanced2..FV... Advanced deinterlace mode. denoise ..FV... denoise level [0, 100] (from 0 to 100) (default 0) detail..FV... enhancement level [0, 100] (from 0 to 100) (default 0) framerate..FV... output framerate (from 0 to DBL_MAX) (default 0/1) procamp ..FV... Enable ProcAmp (from 0 to 1) (default 0) hue ..FV... ProcAmp hue (from -180 to 180) (default 0) saturation ..FV... ProcAmp saturation (from 0 to 10) (default 1) contrast..FV... ProcAmp contrast (from 0 to 10) (default 1) brightness ..FV... ProcAmp brightness (from -100 to 100) (default 0) transpose ..FV... set transpose direction (from -1 to 6) (default -1) cclock_hflip0..FV... rotate counter-clockwise with horizontal flip clock 1..FV... rotate clockwise cclock 2..FV... rotate counter-clockwise clock_hflip 3..FV... rotate clockwise with horizontal flip reversal4..FV... rotate by half-turn hflip 5..FV... flip horizontally vflip 6..FV... flip vertically cw ..FV... set the width crop area expression (default "iw") ch ..FV... set the height crop area expression (default "ih") cx ..FV... set the x crop area expression (default "(in_w-out_w)/2") cy ..FV... set the y crop area expression (default "(in_h-out_h)/2") w ..FV... Output video width (default "cw") width ..FV... Output video width (default "cw") h ..FV... Output video height (default "w*ch/cw") height ..FV... Output video height (default "w*ch/cw") format ..FV... Output pixel format (default "same") async_depth ..FV... Internal parallelization depth, the higher the value the higher the latency. (from 0 to INT_MAX) (default 0) scale_mode..FV... scale mode: 0=auto, 1=low power, 2=high quality (from 0 to 2) (default 0) I tried the following command with vpp_qsv filter but pixel format/colorspace of the output is the same: ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose -f x11grab -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -vf 'hwupload=extra_hw_frames=64,vpp_qsv=format=bgra' -c:v hevc_qsv -preset veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv The ffprobe result for the above command is: Input #0, matroska,webm, from '/tmp/output.mkv': Metadata: ENCODER : Lavf59.27.100 Duration: 00:00:04.85, start: 0.00, bitrate: 50 kb/s Stream #0:0: Video: hevc (Main), yuvj420p(pc, gbr/unknown/unknown, progressive), 1920x1200 [SAR 1:1 DAR 8:5], 20 fps, 20 tbr, 1k tbn Metadata: ENCODER : Lavc59.37.100 hevc_qsv DURATION: 00:00:04.85000 On Thu, 8 Dec 2022 at 17:06, Dennis Mungai wrote: > Hey there, > > I'd recommend using the vpp_qsv filter instead. > > Please provide the output of: > > ffmpeg -h filter=vpp_qsv > > And I'll generate a command line you can use to test and report back. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
On Thu, 8 Dec 2022 at 22:32, wrote: > I'm having trouble selecting a pixel format/colorspace while trying to > encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces. I > have an i7-1165G7 with integrated graphics and I'm using ffmpeg 5.1.1 with > ubuntu > > I'm able to select the pixel format/colorspace with the SW encoder, for > instance, if I use the following command: > > > ffmpeg -f x11grab -draw_mouse 0 -framerate 20 -probesize 42M -video_size > 1920x1200 -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -c:v hevc > -preset veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv > > > I get the correct pixel format and with ffprobe I have the following: > > > Input #0, matroska,webm, from '/tmp/output.mkv': > Metadata: > ENCODER : Lavf59.27.100 > Duration: 00:00:02.90, start: 0.00, bitrate: 92 kb/s > Stream #0:0: Video: hevc (Rext), gbrp(pc, gbr/unknown/unknown, > progressive), 1920x1200, 20 fps, 20 tbr, 1k tbn > Metadata: > ENCODER : Lavc59.37.100 libx265 > DURATION: 00:00:02.9 > > > The video from the above example is correctly encoded and colors are > exactly what I need them to be. > > On the other hand, when I try to use the HW encoder, the pixel format is > wrong. For instance: > > > ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose -f x11grab > -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 > -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -vf > 'hwupload=extra_hw_frames=64,scale_qsv=format=bgra' -c:v hevc_qsv -preset > veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv > > > ffprobe generates: > > > Input #0, matroska,webm, from '/tmp/output.mkv': > Metadata: > ENCODER : Lavf59.27.100 > Duration: 00:00:04.40, start: 0.00, bitrate: 55 kb/s > Stream #0:0: Video: hevc (Main), yuvj420p(pc, gbr/unknown/unknown, > progressive), 1920x1200 [SAR 1:1 DAR 8:5], 20 fps, 20 tbr, 1k tbn > Metadata: > ENCODER : Lavc59.37.100 hevc_qsv > DURATION: 00:00:04.4 > > > As you can see, in the first example, libx265 SW encoder based, the output > is generated with the "gbrp" colorspace while in the second example, > hevc_qsv HW encoder based, the output is generated with the "yuvj420p" > pixel format. > > QUESTION: How can I properly control the pixel format/colorspace of the > generated output using the HW based encoders such as QSV or VAAPI with > my i7-1165G7 integrated graphics card ? I would like a format that is 24 or > 32bits instead of the 12bits I'm getting > > Hey there, I'd recommend using the vpp_qsv filter instead. Please provide the output of: ffmpeg -h filter=vpp_qsv And I'll generate a command line you can use to test and report back. ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
[FFmpeg-user] Trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces
I'm having trouble selecting a pixel format/colorspace while trying to encode a video using ffmpeg and HW accelerated vaapi or qsv interfaces. I have an i7-1165G7 with integrated graphics and I'm using ffmpeg 5.1.1 with ubuntu I'm able to select the pixel format/colorspace with the SW encoder, for instance, if I use the following command: ffmpeg -f x11grab -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -c:v hevc -preset veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv I get the correct pixel format and with ffprobe I have the following: Input #0, matroska,webm, from '/tmp/output.mkv': Metadata: ENCODER : Lavf59.27.100 Duration: 00:00:02.90, start: 0.00, bitrate: 92 kb/s Stream #0:0: Video: hevc (Rext), gbrp(pc, gbr/unknown/unknown, progressive), 1920x1200, 20 fps, 20 tbr, 1k tbn Metadata: ENCODER : Lavc59.37.100 libx265 DURATION : 00:00:02.9 The video from the above example is correctly encoded and colors are exactly what I need them to be. On the other hand, when I try to use the HW encoder, the pixel format is wrong. For instance: ffmpeg -init_hw_device qsv=hw -filter_hw_device hw -v verbose -f x11grab -draw_mouse 0 -framerate 20 -probesize 42M -video_size 1920x1200 -thread_queue_size 1000 -vcodec rawvideo -i ":0.0" -an -vf 'hwupload=extra_hw_frames=64,scale_qsv=format=bgra' -c:v hevc_qsv -preset veryslow -global_quality 17 -look_ahead 1 -y /tmp/output.mkv ffprobe generates: Input #0, matroska,webm, from '/tmp/output.mkv': Metadata: ENCODER : Lavf59.27.100 Duration: 00:00:04.40, start: 0.00, bitrate: 55 kb/s Stream #0:0: Video: hevc (Main), yuvj420p(pc, gbr/unknown/unknown, progressive), 1920x1200 [SAR 1:1 DAR 8:5], 20 fps, 20 tbr, 1k tbn Metadata: ENCODER : Lavc59.37.100 hevc_qsv DURATION : 00:00:04.4 As you can see, in the first example, libx265 SW encoder based, the output is generated with the "gbrp" colorspace while in the second example, hevc_qsv HW encoder based, the output is generated with the "yuvj420p" pixel format. QUESTION: How can I properly control the pixel format/colorspace of the generated output using the HW based encoders such as QSV or VAAPI with my i7-1165G7 integrated graphics card ? I would like a format that is 24 or 32bits instead of the 12bits I'm getting ___ ffmpeg-user mailing list ffmpeg-user@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-user To unsubscribe, visit link above, or email ffmpeg-user-requ...@ffmpeg.org with subject "unsubscribe".
