Re: [FFmpeg-devel] [PATCH] doc/filters: Add CUDA Video Filters section for CUDA-based and CUDA+NPP based filters.

[Gyan Doshi <ffmpeg@gyani.pro>]

On 2025-03-17 12:45 am, Danil Iashchenko wrote:
> Hi Gyan and Michael,
> Thank you for reviewing the patch and providing feedback!
> I've addressed all the issues and resubmitting the patch (built and tested with Texinfo 7.1.1).
>
> Per Gyan's suggestion, I'm resubmitting since Patchwork was down when I originally sent it on the 10th.
>
> Please let me know if there's anything else I can clarify or improve.
> Thanks again!

Generally, looks fine. I've a couple of minor gripes but I'll adjust the 
commit msg and apply this.
We can then address the minor points.

Regards,
Gyan


>
> ---
>   doc/filters.texi | 1353 ++++++++++++++++++++++++----------------------
>   1 file changed, 713 insertions(+), 640 deletions(-)
>
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 0ba7d3035f..37b8674756 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -8619,45 +8619,6 @@ Set planes to filter. Default is first only.
>   
>   This filter supports the all above options as @ref{commands}.
>   
> -@section bilateral_cuda
> -CUDA accelerated bilateral filter, an edge preserving filter.
> -This filter is mathematically accurate thanks to the use of GPU acceleration.
> -For best output quality, use one to one chroma subsampling, i.e. yuv444p format.
> -
> -The filter accepts the following options:
> -@table @option
> -@item sigmaS
> -Set sigma of gaussian function to calculate spatial weight, also called sigma space.
> -Allowed range is 0.1 to 512. Default is 0.1.
> -
> -@item sigmaR
> -Set sigma of gaussian function to calculate color range weight, also called sigma color.
> -Allowed range is 0.1 to 512. Default is 0.1.
> -
> -@item window_size
> -Set window size of the bilateral function to determine the number of neighbours to loop on.
> -If the number entered is even, one will be added automatically.
> -Allowed range is 1 to 255. Default is 1.
> -@end table
> -@subsection Examples
> -
> -@itemize
> -@item
> -Apply the bilateral filter on a video.
> -
> -@example
> -./ffmpeg -v verbose \
> --hwaccel cuda -hwaccel_output_format cuda -i input.mp4  \
> --init_hw_device cuda \
> --filter_complex \
> -" \
> -[0:v]scale_cuda=format=yuv444p[scaled_video];
> -[scaled_video]bilateral_cuda=window_size=9:sigmaS=3.0:sigmaR=50.0" \
> --an -sn -c:v h264_nvenc -cq 20 out.mp4
> -@end example
> -
> -@end itemize
> -
>   @section bitplanenoise
>   
>   Show and measure bit plane noise.
> @@ -9243,58 +9204,6 @@ Only deinterlace frames marked as interlaced.
>   The default value is @code{all}.
>   @end table
>   
> -@section bwdif_cuda
> -
> -Deinterlace the input video using the @ref{bwdif} algorithm, but implemented
> -in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
> -and/or nvenc.
> -
> -It accepts the following parameters:
> -
> -@table @option
> -@item mode
> -The interlacing mode to adopt. It accepts one of the following values:
> -
> -@table @option
> -@item 0, send_frame
> -Output one frame for each frame.
> -@item 1, send_field
> -Output one frame for each field.
> -@end table
> -
> -The default value is @code{send_field}.
> -
> -@item parity
> -The picture field parity assumed for the input interlaced video. It accepts one
> -of the following values:
> -
> -@table @option
> -@item 0, tff
> -Assume the top field is first.
> -@item 1, bff
> -Assume the bottom field is first.
> -@item -1, auto
> -Enable automatic detection of field parity.
> -@end table
> -
> -The default value is @code{auto}.
> -If the interlacing is unknown or the decoder does not export this information,
> -top field first will be assumed.
> -
> -@item deint
> -Specify which frames to deinterlace. Accepts one of the following
> -values:
> -
> -@table @option
> -@item 0, all
> -Deinterlace all frames.
> -@item 1, interlaced
> -Only deinterlace frames marked as interlaced.
> -@end table
> -
> -The default value is @code{all}.
> -@end table
> -
>   @section ccrepack
>   
>   Repack CEA-708 closed captioning side data
> @@ -9408,48 +9317,6 @@ ffmpeg -f lavfi -i color=c=black:s=1280x720 -i video.mp4 -shortest -filter_compl
>   @end example
>   @end itemize
>   
> -@section chromakey_cuda
> -CUDA accelerated YUV colorspace color/chroma keying.
> -
> -This filter works like normal chromakey filter but operates on CUDA frames.
> -for more details and parameters see @ref{chromakey}.
> -
> -@subsection Examples
> -
> -@itemize
> -@item
> -Make all the green pixels in the input video transparent and use it as an overlay for another video:
> -
> -@example
> -./ffmpeg \
> -    -hwaccel cuda -hwaccel_output_format cuda -i input_green.mp4  \
> -    -hwaccel cuda -hwaccel_output_format cuda -i base_video.mp4 \
> -    -init_hw_device cuda \
> -    -filter_complex \
> -    " \
> -        [0:v]chromakey_cuda=0x25302D:0.1:0.12:1[overlay_video]; \
> -        [1:v]scale_cuda=format=yuv420p[base]; \
> -        [base][overlay_video]overlay_cuda" \
> -    -an -sn -c:v h264_nvenc -cq 20 output.mp4
> -@end example
> -
> -@item
> -Process two software sources, explicitly uploading the frames:
> -
> -@example
> -./ffmpeg -init_hw_device cuda=cuda -filter_hw_device cuda \
> -    -f lavfi -i color=size=800x600:color=white,format=yuv420p \
> -    -f lavfi -i yuvtestsrc=size=200x200,format=yuv420p \
> -    -filter_complex \
> -    " \
> -        [0]hwupload[under]; \
> -        [1]hwupload,chromakey_cuda=green:0.1:0.12[over]; \
> -        [under][over]overlay_cuda" \
> -    -c:v hevc_nvenc -cq 18 -preset slow output.mp4
> -@end example
> -
> -@end itemize
> -
>   @section chromanr
>   Reduce chrominance noise.
>   
> @@ -10427,38 +10294,6 @@ For example to convert the input to SMPTE-240M, use the command:
>   colorspace=smpte240m
>   @end example
>   
> -@section colorspace_cuda
> -
> -CUDA accelerated implementation of the colorspace filter.
> -
> -It is by no means feature complete compared to the software colorspace filter,
> -and at the current time only supports color range conversion between jpeg/full
> -and mpeg/limited range.
> -
> -The filter accepts the following options:
> -
> -@table @option
> -@item range
> -Specify output color range.
> -
> -The accepted values are:
> -@table @samp
> -@item tv
> -TV (restricted) range
> -
> -@item mpeg
> -MPEG (restricted) range
> -
> -@item pc
> -PC (full) range
> -
> -@item jpeg
> -JPEG (full) range
> -
> -@end table
> -
> -@end table
> -
>   @section colortemperature
>   Adjust color temperature in video to simulate variations in ambient color temperature.
>   
> @@ -18988,84 +18823,6 @@ testsrc=s=100x100, split=4 [in0][in1][in2][in3];
>   
>   @end itemize
>   
> -@anchor{overlay_cuda}
> -@section overlay_cuda
> -
> -Overlay one video on top of another.
> -
> -This is the CUDA variant of the @ref{overlay} filter.
> -It only accepts CUDA frames. The underlying input pixel formats have to match.
> -
> -It takes two inputs and has one output. The first input is the "main"
> -video on which the second input is overlaid.
> -
> -It accepts the following parameters:
> -
> -@table @option
> -@item x
> -@item y
> -Set expressions for the x and y coordinates of the overlaid video
> -on the main video.
> -
> -They can contain the following parameters:
> -
> -@table @option
> -
> -@item main_w, W
> -@item main_h, H
> -The main input width and height.
> -
> -@item overlay_w, w
> -@item overlay_h, h
> -The overlay input width and height.
> -
> -@item x
> -@item y
> -The computed values for @var{x} and @var{y}. They are evaluated for
> -each new frame.
> -
> -@item n
> -The ordinal index of the main input frame, starting from 0.
> -
> -@item pos
> -The byte offset position in the file of the main input frame, NAN if unknown.
> -Deprecated, do not use.
> -
> -@item t
> -The timestamp of the main input frame, expressed in seconds, NAN if unknown.
> -
> -@end table
> -
> -Default value is "0" for both expressions.
> -
> -@item eval
> -Set when the expressions for @option{x} and @option{y} are evaluated.
> -
> -It accepts the following values:
> -@table @option
> -@item init
> -Evaluate expressions once during filter initialization or
> -when a command is processed.
> -
> -@item frame
> -Evaluate expressions for each incoming frame
> -@end table
> -
> -Default value is @option{frame}.
> -
> -@item eof_action
> -See @ref{framesync}.
> -
> -@item shortest
> -See @ref{framesync}.
> -
> -@item repeatlast
> -See @ref{framesync}.
> -
> -@end table
> -
> -This filter also supports the @ref{framesync} options.
> -
>   @section owdenoise
>   
>   Apply Overcomplete Wavelet denoiser.
> @@ -21516,287 +21273,6 @@ If the specified expression is not valid, it is kept at its current
>   value.
>   @end table
>   
> -@anchor{scale_cuda}
> -@section scale_cuda
> -
> -Scale (resize) and convert (pixel format) the input video, using accelerated CUDA kernels.
> -Setting the output width and height works in the same way as for the @ref{scale} filter.
> -
> -The filter accepts the following options:
> -@table @option
> -@item w
> -@item h
> -Set the output video dimension expression. Default value is the input dimension.
> -
> -Allows for the same expressions as the @ref{scale} filter.
> -
> -@item interp_algo
> -Sets the algorithm used for scaling:
> -
> -@table @var
> -@item nearest
> -Nearest neighbour
> -
> -Used by default if input parameters match the desired output.
> -
> -@item bilinear
> -Bilinear
> -
> -@item bicubic
> -Bicubic
> -
> -This is the default.
> -
> -@item lanczos
> -Lanczos
> -
> -@end table
> -
> -@item format
> -Controls the output pixel format. By default, or if none is specified, the input
> -pixel format is used.
> -
> -The filter does not support converting between YUV and RGB pixel formats.
> -
> -@item passthrough
> -If set to 0, every frame is processed, even if no conversion is necessary.
> -This mode can be useful to use the filter as a buffer for a downstream
> -frame-consumer that exhausts the limited decoder frame pool.
> -
> -If set to 1, frames are passed through as-is if they match the desired output
> -parameters. This is the default behaviour.
> -
> -@item param
> -Algorithm-Specific parameter.
> -
> -Affects the curves of the bicubic algorithm.
> -
> -@item force_original_aspect_ratio
> -@item force_divisible_by
> -Work the same as the identical @ref{scale} filter options.
> -
> -@item reset_sar
> -Works the same as the identical @ref{scale} filter option.
> -
> -@end table
> -
> -@subsection Examples
> -
> -@itemize
> -@item
> -Scale input to 720p, keeping aspect ratio and ensuring the output is yuv420p.
> -@example
> -scale_cuda=-2:720:format=yuv420p
> -@end example
> -
> -@item
> -Upscale to 4K using nearest neighbour algorithm.
> -@example
> -scale_cuda=4096:2160:interp_algo=nearest
> -@end example
> -
> -@item
> -Don't do any conversion or scaling, but copy all input frames into newly allocated ones.
> -This can be useful to deal with a filter and encode chain that otherwise exhausts the
> -decoders frame pool.
> -@example
> -scale_cuda=passthrough=0
> -@end example
> -@end itemize
> -
> -@anchor{scale_npp}
> -@section scale_npp
> -
> -Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
> -format conversion on CUDA video frames. Setting the output width and height
> -works in the same way as for the @var{scale} filter.
> -
> -The following additional options are accepted:
> -@table @option
> -@item format
> -The pixel format of the output CUDA frames. If set to the string "same" (the
> -default), the input format will be kept. Note that automatic format negotiation
> -and conversion is not yet supported for hardware frames
> -
> -@item interp_algo
> -The interpolation algorithm used for resizing. One of the following:
> -@table @option
> -@item nn
> -Nearest neighbour.
> -
> -@item linear
> -@item cubic
> -@item cubic2p_bspline
> -2-parameter cubic (B=1, C=0)
> -
> -@item cubic2p_catmullrom
> -2-parameter cubic (B=0, C=1/2)
> -
> -@item cubic2p_b05c03
> -2-parameter cubic (B=1/2, C=3/10)
> -
> -@item super
> -Supersampling
> -
> -@item lanczos
> -@end table
> -
> -@item force_original_aspect_ratio
> -Enable decreasing or increasing output video width or height if necessary to
> -keep the original aspect ratio. Possible values:
> -
> -@table @samp
> -@item disable
> -Scale the video as specified and disable this feature.
> -
> -@item decrease
> -The output video dimensions will automatically be decreased if needed.
> -
> -@item increase
> -The output video dimensions will automatically be increased if needed.
> -
> -@end table
> -
> -One useful instance of this option is that when you know a specific device's
> -maximum allowed resolution, you can use this to limit the output video to
> -that, while retaining the aspect ratio. For example, device A allows
> -1280x720 playback, and your video is 1920x800. Using this option (set it to
> -decrease) and specifying 1280x720 to the command line makes the output
> -1280x533.
> -
> -Please note that this is a different thing than specifying -1 for @option{w}
> -or @option{h}, you still need to specify the output resolution for this option
> -to work.
> -
> -@item force_divisible_by
> -Ensures that both the output dimensions, width and height, are divisible by the
> -given integer when used together with @option{force_original_aspect_ratio}. This
> -works similar to using @code{-n} in the @option{w} and @option{h} options.
> -
> -This option respects the value set for @option{force_original_aspect_ratio},
> -increasing or decreasing the resolution accordingly. The video's aspect ratio
> -may be slightly modified.
> -
> -This option can be handy if you need to have a video fit within or exceed
> -a defined resolution using @option{force_original_aspect_ratio} but also have
> -encoder restrictions on width or height divisibility.
> -
> -@item reset_sar
> -Works the same as the identical @ref{scale} filter option.
> -
> -@item eval
> -Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
> -
> -@table @samp
> -@item init
> -Only evaluate expressions once during the filter initialization or when a command is processed.
> -
> -@item frame
> -Evaluate expressions for each incoming frame.
> -
> -@end table
> -
> -@end table
> -
> -The values of the @option{w} and @option{h} options are expressions
> -containing the following constants:
> -
> -@table @var
> -@item in_w
> -@item in_h
> -The input width and height
> -
> -@item iw
> -@item ih
> -These are the same as @var{in_w} and @var{in_h}.
> -
> -@item out_w
> -@item out_h
> -The output (scaled) width and height
> -
> -@item ow
> -@item oh
> -These are the same as @var{out_w} and @var{out_h}
> -
> -@item a
> -The same as @var{iw} / @var{ih}
> -
> -@item sar
> -input sample aspect ratio
> -
> -@item dar
> -The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
> -
> -@item n
> -The (sequential) number of the input frame, starting from 0.
> -Only available with @code{eval=frame}.
> -
> -@item t
> -The presentation timestamp of the input frame, expressed as a number of
> -seconds. Only available with @code{eval=frame}.
> -
> -@item pos
> -The position (byte offset) of the frame in the input stream, or NaN if
> -this information is unavailable and/or meaningless (for example in case of synthetic video).
> -Only available with @code{eval=frame}.
> -Deprecated, do not use.
> -@end table
> -
> -@section scale2ref_npp
> -
> -Use the NVIDIA Performance Primitives (libnpp) to scale (resize) the input
> -video, based on a reference video.
> -
> -See the @ref{scale_npp} filter for available options, scale2ref_npp supports the same
> -but uses the reference video instead of the main input as basis. scale2ref_npp
> -also supports the following additional constants for the @option{w} and
> -@option{h} options:
> -
> -@table @var
> -@item main_w
> -@item main_h
> -The main input video's width and height
> -
> -@item main_a
> -The same as @var{main_w} / @var{main_h}
> -
> -@item main_sar
> -The main input video's sample aspect ratio
> -
> -@item main_dar, mdar
> -The main input video's display aspect ratio. Calculated from
> -@code{(main_w / main_h) * main_sar}.
> -
> -@item main_n
> -The (sequential) number of the main input frame, starting from 0.
> -Only available with @code{eval=frame}.
> -
> -@item main_t
> -The presentation timestamp of the main input frame, expressed as a number of
> -seconds. Only available with @code{eval=frame}.
> -
> -@item main_pos
> -The position (byte offset) of the frame in the main input stream, or NaN if
> -this information is unavailable and/or meaningless (for example in case of synthetic video).
> -Only available with @code{eval=frame}.
> -@end table
> -
> -@subsection Examples
> -
> -@itemize
> -@item
> -Scale a subtitle stream (b) to match the main video (a) in size before overlaying
> -@example
> -'scale2ref_npp[b][a];[a][b]overlay_cuda'
> -@end example
> -
> -@item
> -Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
> -@example
> -[logo-in][video-in]scale2ref_npp=w=oh*mdar:h=ih/10[logo-out][video-out]
> -@end example
> -@end itemize
> -
>   @section scale_vt
>   
>   Scale and convert the color parameters using VTPixelTransferSession.
> @@ -22243,23 +21719,6 @@ Keep the same chroma location (default).
>   @end table
>   @end table
>   
> -@section sharpen_npp
> -Use the NVIDIA Performance Primitives (libnpp) to perform image sharpening with
> -border control.
> -
> -The following additional options are accepted:
> -@table @option
> -
> -@item border_type
> -Type of sampling to be used ad frame borders. One of the following:
> -@table @option
> -
> -@item replicate
> -Replicate pixel values.
> -
> -@end table
> -@end table
> -
>   @section shear
>   Apply shear transform to input video.
>   
> @@ -24417,47 +23876,6 @@ The command above can also be specified as:
>   transpose=1:portrait
>   @end example
>   
> -@section transpose_npp
> -
> -Transpose rows with columns in the input video and optionally flip it.
> -For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
> -
> -It accepts the following parameters:
> -
> -@table @option
> -
> -@item dir
> -Specify the transposition direction.
> -
> -Can assume the following values:
> -@table @samp
> -@item cclock_flip
> -Rotate by 90 degrees counterclockwise and vertically flip. (default)
> -
> -@item clock
> -Rotate by 90 degrees clockwise.
> -
> -@item cclock
> -Rotate by 90 degrees counterclockwise.
> -
> -@item clock_flip
> -Rotate by 90 degrees clockwise and vertically flip.
> -@end table
> -
> -@item passthrough
> -Do not apply the transposition if the input geometry matches the one
> -specified by the specified value. It accepts the following values:
> -@table @samp
> -@item none
> -Always apply transposition. (default)
> -@item portrait
> -Preserve portrait geometry (when @var{height} >= @var{width}).
> -@item landscape
> -Preserve landscape geometry (when @var{width} >= @var{height}).
> -@end table
> -
> -@end table
> -
>   @section trim
>   Trim the input so that the output contains one continuous subpart of the input.
>   
> @@ -26644,64 +26062,6 @@ filter").
>   It accepts the following parameters:
>   
>   
> -@table @option
> -
> -@item mode
> -The interlacing mode to adopt. It accepts one of the following values:
> -
> -@table @option
> -@item 0, send_frame
> -Output one frame for each frame.
> -@item 1, send_field
> -Output one frame for each field.
> -@item 2, send_frame_nospatial
> -Like @code{send_frame}, but it skips the spatial interlacing check.
> -@item 3, send_field_nospatial
> -Like @code{send_field}, but it skips the spatial interlacing check.
> -@end table
> -
> -The default value is @code{send_frame}.
> -
> -@item parity
> -The picture field parity assumed for the input interlaced video. It accepts one
> -of the following values:
> -
> -@table @option
> -@item 0, tff
> -Assume the top field is first.
> -@item 1, bff
> -Assume the bottom field is first.
> -@item -1, auto
> -Enable automatic detection of field parity.
> -@end table
> -
> -The default value is @code{auto}.
> -If the interlacing is unknown or the decoder does not export this information,
> -top field first will be assumed.
> -
> -@item deint
> -Specify which frames to deinterlace. Accepts one of the following
> -values:
> -
> -@table @option
> -@item 0, all
> -Deinterlace all frames.
> -@item 1, interlaced
> -Only deinterlace frames marked as interlaced.
> -@end table
> -
> -The default value is @code{all}.
> -@end table
> -
> -@section yadif_cuda
> -
> -Deinterlace the input video using the @ref{yadif} algorithm, but implemented
> -in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
> -and/or nvenc.
> -
> -It accepts the following parameters:
> -
> -
>   @table @option
>   
>   @item mode
> @@ -27172,6 +26532,719 @@ value.
>   
>   @c man end VIDEO FILTERS
>   
> +@chapter CUDA Video Filters
> +@c man begin CUDA Video Filters
> +
> +To enable CUDA and/or NPP filters please refer to configuration guidelines for @ref{CUDA} and for @ref{CUDA NPP} filters.
> +
> +Running CUDA filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
> +@table @option
> +
> +@item -init_hw_device cuda[=@var{name}][:@var{device}[,@var{key=value}...]]
> +Initialise a new hardware device of type @var{cuda} called @var{name}, using the
> +given device parameters.
> +
> +@item -filter_hw_device @var{name}
> +Pass the hardware device called @var{name} to all filters in any filter graph.
> +
> +@end table
> +
> +For more detailed information see @url{https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options}
> +
> +@itemize
> +@item
> +Example of initializing second CUDA device on the system and running scale_cuda and bilateral_cuda filters.
> +@example
> +./ffmpeg -hwaccel cuda -hwaccel_output_format cuda -i input.mp4 -init_hw_device cuda:1 -filter_complex \
> +"[0:v]scale_cuda=format=yuv444p[scaled_video];[scaled_video]bilateral_cuda=window_size=9:sigmaS=3.0:sigmaR=50.0" \
> +-an -sn -c:v h264_nvenc -cq 20 out.mp4
> +@end example
> +@end itemize
> +
> +Since CUDA filters operate exclusively on GPU memory, frame data must sometimes be uploaded (@ref{hwupload}) to hardware surfaces associated with the appropriate CUDA device before processing, and downloaded (@ref{hwdownload}) back to normal memory afterward, if required. Whether @ref{hwupload} or @ref{hwdownload} is necessary depends on the specific workflow:
> +
> +@itemize
> +@item If the input frames are already in GPU memory (e.g., when using @code{-hwaccel cuda} or @code{-hwaccel_output_format cuda}), explicit use of @ref{hwupload} is not needed, as the data is already in the appropriate memory space.
> +@item If the input frames are in CPU memory (e.g., software-decoded frames or frames processed by CPU-based filters), it is necessary to use @ref{hwupload} to transfer the data to GPU memory for CUDA processing.
> +@item If the output of the CUDA filters needs to be further processed by software-based filters or saved in a format not supported by GPU-based encoders, @ref{hwdownload} is required to transfer the data back to CPU memory.
> +@end itemize
> +Note that @ref{hwupload} uploads data to a surface with the same layout as the software frame, so it may be necessary to add a @ref{format} filter immediately before @ref{hwupload} to ensure the input is in the correct format. Similarly, @ref{hwdownload} may not support all output formats, so an additional @ref{format} filter may need to be inserted immediately after @ref{hwdownload} in the filter graph to ensure compatibility.
> +
> +@anchor{CUDA}
> +@section CUDA
> +Below is a description of the currently available Nvidia CUDA video filters.
> +
> +Prerequisites:
> +@itemize
> +@item Install Nvidia CUDA Toolkit
> +@end itemize
> +
> +Note: If FFmpeg detects the Nvidia CUDA Toolkit during configuration, it will enable CUDA filters automatically without requiring any additional flags. If you want to explicitly enable them, use the following options:
> +
> +@itemize
> +@item Configure FFmpeg with @code{--enable-cuda-nvcc --enable-nonfree}.
> +@item Configure FFmpeg with @code{--enable-cuda-llvm}. Additional requirement: @code{llvm} lib must be installed.
> +@end itemize
> +
> +@subsection bilateral_cuda
> +CUDA accelerated bilateral filter, an edge preserving filter.
> +This filter is mathematically accurate thanks to the use of GPU acceleration.
> +For best output quality, use one to one chroma subsampling, i.e. yuv444p format.
> +
> +The filter accepts the following options:
> +@table @option
> +@item sigmaS
> +Set sigma of gaussian function to calculate spatial weight, also called sigma space.
> +Allowed range is 0.1 to 512. Default is 0.1.
> +
> +@item sigmaR
> +Set sigma of gaussian function to calculate color range weight, also called sigma color.
> +Allowed range is 0.1 to 512. Default is 0.1.
> +
> +@item window_size
> +Set window size of the bilateral function to determine the number of neighbours to loop on.
> +If the number entered is even, one will be added automatically.
> +Allowed range is 1 to 255. Default is 1.
> +@end table
> +@subsubsection Examples
> +
> +@itemize
> +@item
> +Apply the bilateral filter on a video.
> +
> +@example
> +./ffmpeg -v verbose \
> +-hwaccel cuda -hwaccel_output_format cuda -i input.mp4  \
> +-init_hw_device cuda \
> +-filter_complex \
> +" \
> +[0:v]scale_cuda=format=yuv444p[scaled_video];
> +[scaled_video]bilateral_cuda=window_size=9:sigmaS=3.0:sigmaR=50.0" \
> +-an -sn -c:v h264_nvenc -cq 20 out.mp4
> +@end example
> +
> +@end itemize
> +
> +@subsection bwdif_cuda
> +
> +Deinterlace the input video using the @ref{bwdif} algorithm, but implemented
> +in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
> +and/or nvenc.
> +
> +It accepts the following parameters:
> +
> +@table @option
> +@item mode
> +The interlacing mode to adopt. It accepts one of the following values:
> +
> +@table @option
> +@item 0, send_frame
> +Output one frame for each frame.
> +@item 1, send_field
> +Output one frame for each field.
> +@end table
> +
> +The default value is @code{send_field}.
> +
> +@item parity
> +The picture field parity assumed for the input interlaced video. It accepts one
> +of the following values:
> +
> +@table @option
> +@item 0, tff
> +Assume the top field is first.
> +@item 1, bff
> +Assume the bottom field is first.
> +@item -1, auto
> +Enable automatic detection of field parity.
> +@end table
> +
> +The default value is @code{auto}.
> +If the interlacing is unknown or the decoder does not export this information,
> +top field first will be assumed.
> +
> +@item deint
> +Specify which frames to deinterlace. Accepts one of the following
> +values:
> +
> +@table @option
> +@item 0, all
> +Deinterlace all frames.
> +@item 1, interlaced
> +Only deinterlace frames marked as interlaced.
> +@end table
> +
> +The default value is @code{all}.
> +@end table
> +
> +@subsection chromakey_cuda
> +CUDA accelerated YUV colorspace color/chroma keying.
> +
> +This filter works like normal chromakey filter but operates on CUDA frames.
> +for more details and parameters see @ref{chromakey}.
> +
> +@subsubsection Examples
> +
> +@itemize
> +@item
> +Make all the green pixels in the input video transparent and use it as an overlay for another video:
> +
> +@example
> +./ffmpeg \
> +    -hwaccel cuda -hwaccel_output_format cuda -i input_green.mp4  \
> +    -hwaccel cuda -hwaccel_output_format cuda -i base_video.mp4 \
> +    -init_hw_device cuda \
> +    -filter_complex \
> +    " \
> +        [0:v]chromakey_cuda=0x25302D:0.1:0.12:1[overlay_video]; \
> +        [1:v]scale_cuda=format=yuv420p[base]; \
> +        [base][overlay_video]overlay_cuda" \
> +    -an -sn -c:v h264_nvenc -cq 20 output.mp4
> +@end example
> +
> +@item
> +Process two software sources, explicitly uploading the frames:
> +
> +@example
> +./ffmpeg -init_hw_device cuda=cuda -filter_hw_device cuda \
> +    -f lavfi -i color=size=800x600:color=white,format=yuv420p \
> +    -f lavfi -i yuvtestsrc=size=200x200,format=yuv420p \
> +    -filter_complex \
> +    " \
> +        [0]hwupload[under]; \
> +        [1]hwupload,chromakey_cuda=green:0.1:0.12[over]; \
> +        [under][over]overlay_cuda" \
> +    -c:v hevc_nvenc -cq 18 -preset slow output.mp4
> +@end example
> +
> +@end itemize
> +
> +@subsection colorspace_cuda
> +
> +CUDA accelerated implementation of the colorspace filter.
> +
> +It is by no means feature complete compared to the software colorspace filter,
> +and at the current time only supports color range conversion between jpeg/full
> +and mpeg/limited range.
> +
> +The filter accepts the following options:
> +
> +@table @option
> +@item range
> +Specify output color range.
> +
> +The accepted values are:
> +@table @samp
> +@item tv
> +TV (restricted) range
> +
> +@item mpeg
> +MPEG (restricted) range
> +
> +@item pc
> +PC (full) range
> +
> +@item jpeg
> +JPEG (full) range
> +
> +@end table
> +
> +@end table
> +
> +@anchor{overlay_cuda}
> +@subsection overlay_cuda
> +
> +Overlay one video on top of another.
> +
> +This is the CUDA variant of the @ref{overlay} filter.
> +It only accepts CUDA frames. The underlying input pixel formats have to match.
> +
> +It takes two inputs and has one output. The first input is the "main"
> +video on which the second input is overlaid.
> +
> +It accepts the following parameters:
> +
> +@table @option
> +@item x
> +@item y
> +Set expressions for the x and y coordinates of the overlaid video
> +on the main video.
> +
> +They can contain the following parameters:
> +
> +@table @option
> +
> +@item main_w, W
> +@item main_h, H
> +The main input width and height.
> +
> +@item overlay_w, w
> +@item overlay_h, h
> +The overlay input width and height.
> +
> +@item x
> +@item y
> +The computed values for @var{x} and @var{y}. They are evaluated for
> +each new frame.
> +
> +@item n
> +The ordinal index of the main input frame, starting from 0.
> +
> +@item pos
> +The byte offset position in the file of the main input frame, NAN if unknown.
> +Deprecated, do not use.
> +
> +@item t
> +The timestamp of the main input frame, expressed in seconds, NAN if unknown.
> +
> +@end table
> +
> +Default value is "0" for both expressions.
> +
> +@item eval
> +Set when the expressions for @option{x} and @option{y} are evaluated.
> +
> +It accepts the following values:
> +@table @option
> +@item init
> +Evaluate expressions once during filter initialization or
> +when a command is processed.
> +
> +@item frame
> +Evaluate expressions for each incoming frame
> +@end table
> +
> +Default value is @option{frame}.
> +
> +@item eof_action
> +See @ref{framesync}.
> +
> +@item shortest
> +See @ref{framesync}.
> +
> +@item repeatlast
> +See @ref{framesync}.
> +
> +@end table
> +
> +This filter also supports the @ref{framesync} options.
> +
> +@anchor{scale_cuda}
> +@subsection scale_cuda
> +
> +Scale (resize) and convert (pixel format) the input video, using accelerated CUDA kernels.
> +Setting the output width and height works in the same way as for the @ref{scale} filter.
> +
> +The filter accepts the following options:
> +@table @option
> +@item w
> +@item h
> +Set the output video dimension expression. Default value is the input dimension.
> +
> +Allows for the same expressions as the @ref{scale} filter.
> +
> +@item interp_algo
> +Sets the algorithm used for scaling:
> +
> +@table @var
> +@item nearest
> +Nearest neighbour
> +
> +Used by default if input parameters match the desired output.
> +
> +@item bilinear
> +Bilinear
> +
> +@item bicubic
> +Bicubic
> +
> +This is the default.
> +
> +@item lanczos
> +Lanczos
> +
> +@end table
> +
> +@item format
> +Controls the output pixel format. By default, or if none is specified, the input
> +pixel format is used.
> +
> +The filter does not support converting between YUV and RGB pixel formats.
> +
> +@item passthrough
> +If set to 0, every frame is processed, even if no conversion is necessary.
> +This mode can be useful to use the filter as a buffer for a downstream
> +frame-consumer that exhausts the limited decoder frame pool.
> +
> +If set to 1, frames are passed through as-is if they match the desired output
> +parameters. This is the default behaviour.
> +
> +@item param
> +Algorithm-Specific parameter.
> +
> +Affects the curves of the bicubic algorithm.
> +
> +@item force_original_aspect_ratio
> +@item force_divisible_by
> +Work the same as the identical @ref{scale} filter options.
> +
> +@item reset_sar
> +Works the same as the identical @ref{scale} filter option.
> +
> +@end table
> +
> +@subsubsection Examples
> +
> +@itemize
> +@item
> +Scale input to 720p, keeping aspect ratio and ensuring the output is yuv420p.
> +@example
> +scale_cuda=-2:720:format=yuv420p
> +@end example
> +
> +@item
> +Upscale to 4K using nearest neighbour algorithm.
> +@example
> +scale_cuda=4096:2160:interp_algo=nearest
> +@end example
> +
> +@item
> +Don't do any conversion or scaling, but copy all input frames into newly allocated ones.
> +This can be useful to deal with a filter and encode chain that otherwise exhausts the
> +decoders frame pool.
> +@example
> +scale_cuda=passthrough=0
> +@end example
> +@end itemize
> +
> +@subsection yadif_cuda
> +
> +Deinterlace the input video using the @ref{yadif} algorithm, but implemented
> +in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec
> +and/or nvenc.
> +
> +It accepts the following parameters:
> +
> +
> +@table @option
> +
> +@item mode
> +The interlacing mode to adopt. It accepts one of the following values:
> +
> +@table @option
> +@item 0, send_frame
> +Output one frame for each frame.
> +@item 1, send_field
> +Output one frame for each field.
> +@item 2, send_frame_nospatial
> +Like @code{send_frame}, but it skips the spatial interlacing check.
> +@item 3, send_field_nospatial
> +Like @code{send_field}, but it skips the spatial interlacing check.
> +@end table
> +
> +The default value is @code{send_frame}.
> +
> +@item parity
> +The picture field parity assumed for the input interlaced video. It accepts one
> +of the following values:
> +
> +@table @option
> +@item 0, tff
> +Assume the top field is first.
> +@item 1, bff
> +Assume the bottom field is first.
> +@item -1, auto
> +Enable automatic detection of field parity.
> +@end table
> +
> +The default value is @code{auto}.
> +If the interlacing is unknown or the decoder does not export this information,
> +top field first will be assumed.
> +
> +@item deint
> +Specify which frames to deinterlace. Accepts one of the following
> +values:
> +
> +@table @option
> +@item 0, all
> +Deinterlace all frames.
> +@item 1, interlaced
> +Only deinterlace frames marked as interlaced.
> +@end table
> +
> +The default value is @code{all}.
> +@end table
> +
> +@anchor{CUDA NPP}
> +@section CUDA NPP
> +Below is a description of the currently available NVIDIA Performance Primitives (libnpp) video filters.
> +
> +Prerequisites:
> +@itemize
> +@item Install Nvidia CUDA Toolkit
> +@item Install libnpp
> +@end itemize
> +
> +To enable CUDA NPP filters:
> +
> +@itemize
> +@item Configure FFmpeg with @code{--enable-nonfree --enable-libnpp}.
> +@end itemize
> +
> +
> +@anchor{scale_npp}
> +@subsection scale_npp
> +
> +Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
> +format conversion on CUDA video frames. Setting the output width and height
> +works in the same way as for the @var{scale} filter.
> +
> +The following additional options are accepted:
> +@table @option
> +@item format
> +The pixel format of the output CUDA frames. If set to the string "same" (the
> +default), the input format will be kept. Note that automatic format negotiation
> +and conversion is not yet supported for hardware frames
> +
> +@item interp_algo
> +The interpolation algorithm used for resizing. One of the following:
> +@table @option
> +@item nn
> +Nearest neighbour.
> +
> +@item linear
> +@item cubic
> +@item cubic2p_bspline
> +2-parameter cubic (B=1, C=0)
> +
> +@item cubic2p_catmullrom
> +2-parameter cubic (B=0, C=1/2)
> +
> +@item cubic2p_b05c03
> +2-parameter cubic (B=1/2, C=3/10)
> +
> +@item super
> +Supersampling
> +
> +@item lanczos
> +@end table
> +
> +@item force_original_aspect_ratio
> +Enable decreasing or increasing output video width or height if necessary to
> +keep the original aspect ratio. Possible values:
> +
> +@table @samp
> +@item disable
> +Scale the video as specified and disable this feature.
> +
> +@item decrease
> +The output video dimensions will automatically be decreased if needed.
> +
> +@item increase
> +The output video dimensions will automatically be increased if needed.
> +
> +@end table
> +
> +One useful instance of this option is that when you know a specific device's
> +maximum allowed resolution, you can use this to limit the output video to
> +that, while retaining the aspect ratio. For example, device A allows
> +1280x720 playback, and your video is 1920x800. Using this option (set it to
> +decrease) and specifying 1280x720 to the command line makes the output
> +1280x533.
> +
> +Please note that this is a different thing than specifying -1 for @option{w}
> +or @option{h}, you still need to specify the output resolution for this option
> +to work.
> +
> +@item force_divisible_by
> +Ensures that both the output dimensions, width and height, are divisible by the
> +given integer when used together with @option{force_original_aspect_ratio}. This
> +works similar to using @code{-n} in the @option{w} and @option{h} options.
> +
> +This option respects the value set for @option{force_original_aspect_ratio},
> +increasing or decreasing the resolution accordingly. The video's aspect ratio
> +may be slightly modified.
> +
> +This option can be handy if you need to have a video fit within or exceed
> +a defined resolution using @option{force_original_aspect_ratio} but also have
> +encoder restrictions on width or height divisibility.
> +
> +@item reset_sar
> +Works the same as the identical @ref{scale} filter option.
> +
> +@item eval
> +Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
> +
> +@table @samp
> +@item init
> +Only evaluate expressions once during the filter initialization or when a command is processed.
> +
> +@item frame
> +Evaluate expressions for each incoming frame.
> +
> +@end table
> +
> +@end table
> +
> +The values of the @option{w} and @option{h} options are expressions
> +containing the following constants:
> +
> +@table @var
> +@item in_w
> +@item in_h
> +The input width and height
> +
> +@item iw
> +@item ih
> +These are the same as @var{in_w} and @var{in_h}.
> +
> +@item out_w
> +@item out_h
> +The output (scaled) width and height
> +
> +@item ow
> +@item oh
> +These are the same as @var{out_w} and @var{out_h}
> +
> +@item a
> +The same as @var{iw} / @var{ih}
> +
> +@item sar
> +input sample aspect ratio
> +
> +@item dar
> +The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
> +
> +@item n
> +The (sequential) number of the input frame, starting from 0.
> +Only available with @code{eval=frame}.
> +
> +@item t
> +The presentation timestamp of the input frame, expressed as a number of
> +seconds. Only available with @code{eval=frame}.
> +
> +@item pos
> +The position (byte offset) of the frame in the input stream, or NaN if
> +this information is unavailable and/or meaningless (for example in case of synthetic video).
> +Only available with @code{eval=frame}.
> +Deprecated, do not use.
> +@end table
> +
> +@subsection scale2ref_npp
> +
> +Use the NVIDIA Performance Primitives (libnpp) to scale (resize) the input
> +video, based on a reference video.
> +
> +See the @ref{scale_npp} filter for available options, scale2ref_npp supports the same
> +but uses the reference video instead of the main input as basis. scale2ref_npp
> +also supports the following additional constants for the @option{w} and
> +@option{h} options:
> +
> +@table @var
> +@item main_w
> +@item main_h
> +The main input video's width and height
> +
> +@item main_a
> +The same as @var{main_w} / @var{main_h}
> +
> +@item main_sar
> +The main input video's sample aspect ratio
> +
> +@item main_dar, mdar
> +The main input video's display aspect ratio. Calculated from
> +@code{(main_w / main_h) * main_sar}.
> +
> +@item main_n
> +The (sequential) number of the main input frame, starting from 0.
> +Only available with @code{eval=frame}.
> +
> +@item main_t
> +The presentation timestamp of the main input frame, expressed as a number of
> +seconds. Only available with @code{eval=frame}.
> +
> +@item main_pos
> +The position (byte offset) of the frame in the main input stream, or NaN if
> +this information is unavailable and/or meaningless (for example in case of synthetic video).
> +Only available with @code{eval=frame}.
> +@end table
> +
> +@subsubsection Examples
> +
> +@itemize
> +@item
> +Scale a subtitle stream (b) to match the main video (a) in size before overlaying
> +@example
> +'scale2ref_npp[b][a];[a][b]overlay_cuda'
> +@end example
> +
> +@item
> +Scale a logo to 1/10th the height of a video, while preserving its display aspect ratio.
> +@example
> +[logo-in][video-in]scale2ref_npp=w=oh*mdar:h=ih/10[logo-out][video-out]
> +@end example
> +@end itemize
> +
> +@subsection sharpen_npp
> +Use the NVIDIA Performance Primitives (libnpp) to perform image sharpening with
> +border control.
> +
> +The following additional options are accepted:
> +@table @option
> +
> +@item border_type
> +Type of sampling to be used ad frame borders. One of the following:
> +@table @option
> +
> +@item replicate
> +Replicate pixel values.
> +
> +@end table
> +@end table
> +
> +@subsection transpose_npp
> +
> +Transpose rows with columns in the input video and optionally flip it.
> +For more in depth examples see the @ref{transpose} video filter, which shares mostly the same options.
> +
> +It accepts the following parameters:
> +
> +@table @option
> +
> +@item dir
> +Specify the transposition direction.
> +
> +Can assume the following values:
> +@table @samp
> +@item cclock_flip
> +Rotate by 90 degrees counterclockwise and vertically flip. (default)
> +
> +@item clock
> +Rotate by 90 degrees clockwise.
> +
> +@item cclock
> +Rotate by 90 degrees counterclockwise.
> +
> +@item clock_flip
> +Rotate by 90 degrees clockwise and vertically flip.
> +@end table
> +
> +@item passthrough
> +Do not apply the transposition if the input geometry matches the one
> +specified by the specified value. It accepts the following values:
> +@table @samp
> +@item none
> +Always apply transposition. (default)
> +@item portrait
> +Preserve portrait geometry (when @var{height} >= @var{width}).
> +@item landscape
> +Preserve landscape geometry (when @var{width} >= @var{height}).
> +@end table
> +
> +@end table
> +
> +@c man end CUDA Video Filters
> +
>   @chapter OpenCL Video Filters
>   @c man begin OPENCL VIDEO FILTERS
>   

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".