Changes between Version 52 and Version 53 of HWAccelIntro


Ignore:
Timestamp:
Jul 25, 2017, 8:13:37 AM (9 years ago)
Author:
slhck
Comment:

minor nits

Legend:

Unmodified
Added
Removed
Modified

  • HWAccelIntro

    v52 v53  
    11[[PageOutline(2, Contents)]]
    22
    3 Many platforms offer access to dedicated hardware to perform a range of video-related tasks.  Using such hardware allows some operations like decoding, encoding or filtering to be completed faster or using less of other resources (particularly CPU), but may give different or inferior results, or impose additional restrictions which are not present when using software only.  On PC-like platforms, video hardware is typically integrated into a GPU (from AMD, Intel or Nvidia), while on mobile SoC-type platforms it is generally an independent IP core (many different vendors).
     3Many platforms offer access to dedicated hardware to perform a range of video-related tasks. Using such hardware allows some operations like decoding, encoding or filtering to be completed faster or using less of other resources (particularly CPU), but may give different or inferior results, or impose additional restrictions which are not present when using software only. On PC-like platforms, video hardware is typically integrated into a GPU (from AMD, Intel or NVIDIA), while on mobile SoC-type platforms it is generally an independent IP core (many different vendors).
    44
    5 Hardware decoders will generate equivalent output to software decoders, but may use less power and CPU to do so.  Feature support varies - for more complex codecs with many different profiles, hardware decoders rarely implement all of them (for example, hardware decoders tend not to implement anything beyond YUV 4:2:0 at 8-bit depth for H.264).  A common feature of many hardware decoders to be able to generate output in hardware surfaces suitable for use by other components (with discrete graphics cards, this means surfaces in the memory on the card rather than in system memory) - this is often useful for playback, as no further copying is required before rendering the output, and in some cases it can also be used with encoders supporting hardware surface input to avoid any copying at all in transcode cases.
     5Hardware decoders will generate equivalent output to software decoders, but may use less power and CPU to do so.  Feature support varies – for more complex codecs with many different profiles, hardware decoders rarely implement all of them (for example, hardware decoders tend not to implement anything beyond YUV 4:2:0 at 8-bit depth for H.264). A common feature of many hardware decoders to be able to generate output in hardware surfaces suitable for use by other components (with discrete graphics cards, this means surfaces in the memory on the card rather than in system memory) – this is often useful for playback, as no further copying is required before rendering the output, and in some cases it can also be used with encoders supporting hardware surface input to avoid any copying at all in transcode cases.
    66
    7 Hardware encoders typically generate output of significantly lower quality than good software encoders like x264, but are generally faster and do not use much CPU resource.  (That is, they require a higher bitrate to make output with the same perceptual quality, or they make output with a lower perceptual quality at the same bitrate.)
     7Hardware encoders typically generate output of significantly lower quality than good software encoders like x264, but are generally faster and do not use much CPU resource. (That is, they require a higher bitrate to make output with the same perceptual quality, or they make output with a lower perceptual quality at the same bitrate.)
    88
    9 Systems with decode and/or encode capability may also offer access to other related filtering features.  Things like scaling and deinterlacing are common, other postprocessing may be available depending on the system. Where hardware surfaces are usable, these filters will generally act on them rather than on normal frames in system memory.
     9Systems with decode and/or encode capability may also offer access to other related filtering features. Things like scaling and deinterlacing are common, other postprocessing may be available depending on the system. Where hardware surfaces are usable, these filters will generally act on them rather than on normal frames in system memory.
    1010
    11 There are a lot of different APIs of varying standardisation status available.  FFmpeg offers access to many of these, with varying support.
     11There are a lot of different APIs of varying standardisation status available. FFmpeg offers access to many of these, with varying support.
    1212
    1313
    … …  
    1515
    1616|| ||||||= Linux =||||||= Windows =||= Android =||||= Apple =||= Other =||
    17 || ||= AMD =||= Intel =||= Nvidia =||= AMD =||= Intel =||= Nvidia =||= =||= macOS =||= iOS =||= Raspberry Pi =||
     17|| ||= AMD =||= Intel =||= NVIDIA =||= AMD =||= Intel =||= NVIDIA =||= =||= macOS =||= iOS =||= Raspberry Pi =||
    1818|| CUDA / CUVID / NVENC || N || N || Y || N || N || Y || N || N || N || N ||
    1919|| Direct3D 11          || N || N || N || Y || Y || Y || N || N || N || N ||
    … …  
    6868Internal hwaccel decoders are enabled via the `-hwaccel` option.  The software decoder starts normally, but if it detects a stream which is decodable in hardware then it will attempt to delegate all significant processing to that hardware.  If the stream is not decodable in hardware (for example, it is an unsupported codec or profile) then it will still be decoded in software automatically.  If the hardware requires a particular device to function (or needs to distinguish between multiple devices, say if several graphics cards are available) then one can be selected using `-hwaccel_device`.
    6969
    70 External wrapper decoders are used by setting a specific decoder with the `-codec:v` option.  Typically they are named `codec_api` (for example: `h264_cuvid`).  These decoders require the codec to be known in advance, and do not support any fallback to software if the stream is not supported.
     70External wrapper decoders are used by setting a specific decoder with the `-codec:v` option.  Typically they are named `codec_api` (for example: `h264_cuvid`). These decoders require the codec to be known in advance, and do not support any fallback to software if the stream is not supported.
    7171
    72 Encoder wrappers are selected by `-codec:v`.  Encoders generally have lots of options - look at the documentation for the particular encoder for details.
     72Encoder wrappers are selected by `-codec:v`.  Encoders generally have lots of options – look at the documentation for the particular encoder for details.
    7373
    74 Hardware filters can be used in a filter graph like any other filter.  Note, however, that they may not support any formats in common with software filters - in such cases it may be necessary to make use of `hwupload` and `hwdownload` filter instances to move frame data between hardware surfaces and normal memory.
     74Hardware filters can be used in a filter graph like any other filter.  Note, however, that they may not support any formats in common with software filters – in such cases it may be necessary to make use of `hwupload` and `hwdownload` filter instances to move frame data between hardware surfaces and normal memory.
    7575
    7676
    7777== VDPAU ==
    78 [http://http.download.nvidia.com/XFree86/vdpau/doxygen/html/index.html Video Decode and Presentation API for Unix]. Developed by NVidia for UNIX/Linux systems. To enable this you typically need the libvdpau development package in your distribution, and a compatible graphic card.
     78
     79[http://http.download.nvidia.com/XFree86/vdpau/doxygen/html/index.html Video Decode and Presentation API for Unix]. Developed by NVIDIA for Unix/Linux systems. To enable this you typically need the `libvdpau` development package in your distribution, and a compatible graphics card.
    7980
    8081Note that VDPAU cannot be used to decode frames in memory, the compressed frames are sent by libavcodec to the GPU device supported by VDPAU and then the decoded image can be accessed using the VDPAU API. This is not done automatically by FFmpeg, but must be done at the application level (check for example the {{{ffmpeg_vdpau.c}}} file used by {{{ffmpeg.c}}}). Also, note that with this API it is not possible to move the decoded frame back to RAM, for example in case you need to encode again the decoded frame (e.g. when doing transcoding on a server).
    … …  
    8687Video Acceleration API (VAAPI) is a non-proprietary and royalty-free open source software library ("libva") and API specification, initially developed by Intel but can be used in combination with other devices.
    8788
    88 It can be used to access the Quick Sync hardware in Intel GPUs and the UVD/VCE hardware in AMD GPUs.  See [wiki:Hardware/VAAPI VAAPI].
     89It can be used to access the Quick Sync hardware in Intel GPUs and the UVD/VCE hardware in AMD GPUs. See [wiki:Hardware/VAAPI VAAPI].
    8990
    9091== DXVA2 ==
    … …  
    9596http://msdn.microsoft.com/en-us/library/windows/desktop/cc307941%28v=vs.85%29.aspx
    9697
    97 Several decoders are currently supported, in particular H.264, MPEG2, VC1 and WMV3.
     98Several decoders are currently supported, in particular H.264, MPEG-2, VC-1 and WMV 3.
    9899
    99100DXVA2 hardware acceleration only works on Windows. In order to build FFmpeg with DXVA2 support, you need to install the dxva2api.h header.
    100 For MinGW this can be done by downloading the header maintained by VLC:
     101For MinGW this can be done by [http://download.videolan.org/pub/contrib/dxva2api.h downloading the header maintained by VLC] and installing it in the include patch (for example in {{{/usr/include/}}}).
    101102
    102 http://download.videolan.org/pub/contrib/dxva2api.h
    103 
    104 and installing it in the include patch (for example in {{{/usr/include/}}}).
    105 
    106 For MinGW64, the dxva2api.h is provided by default.  One way to install mingw-w64 is through a {{{pacman}}} repository, and can be installed using one of the two following commands, depending on the architecture:
     103For MinGW64, `dxva2api.h` is provided by default. One way to install mingw-w64 is through a {{{pacman}}} repository, and can be installed using one of the two following commands, depending on the architecture:
    107104{{{
    108105pacman -S mingw-w64-i686-gcc
    … …  
    120117== VDA ==
    121118
    122 Video Decoding API, only supported on MAC. H.264 decoding is available in FFmpeg/libavcodec.
    123 
    124 Developers documentation:
    125 https://developer.apple.com/library/mac/technotes/tn2267/_index.html
     119[https://developer.apple.com/library/mac/technotes/tn2267/_index.html Video Decode Acceleration Framework], only supported on macOS. H.264 decoding is available in FFmpeg/libavcodec.
    126120
    127121== NVENC ==
    … …  
    133127* {{{ffmpeg}}} configured without {{{--disable-nvenc}}}
    134128
    135 Visit [https://developer.nvidia.com/nvidia-video-codec-sdk NVIDIA Video Codec SDK] to download the SDK and to read more about the supported GPUs and supported drivers.
     129Download the [https://developer.nvidia.com/nvidia-video-codec-sdk NVIDIA Video Codec SDK], and check the website for more info on the supported GPUs and drivers.
    136130
    137131Usage example:
    138 {{{ 
     132{{{
    139133ffmpeg -i input -c:v h264_nvenc -profile high444p -pixel_format yuv444p -preset default output.mp4
    140134}}}
    … …  
    149143== CUDA/CUVID/NvDecode ==
    150144
    151 CUVID, which is also called nvdec by Nvidia now, can be used for decoding on Windows and Linux.
     145CUVID, which is also called nvdec by NVIDIA now, can be used for decoding on Windows and Linux.
    152146In combination with nvenc it offers full hardware transcoding.
    153147
    154 CUVID offers decoders for H264, HEVC, MJPEG, mpeg1/2/4, vp8/9, vc1.
     148CUVID offers decoders for H.264, HEVC, MJPEG, MPEG-1/2/4, VP8/VP9, VC-1.
    155149Codec support varies by hardware. The full set of codecs being available only on Pascal hardware, which adds VP9 and 10 bit support.
    156150
    157 While decoding 10 bit video is supported, it is not possible to do full hardware transcoding currently (See the partial hardware example below).
     151While decoding 10 bit video is supported, it is not possible to do full hardware transcoding currently (see the partial hardware example below).
    158152
    159153Sample decode using CUVID, the cuvid decoder copies the frames to system memory in this case:
    … …  
    180174== libmfx ==
    181175
    182 libmfx is a proprietary library from Intel for use of Quick Sync hardware on both Linux and Windows.  On Windows it is the primary way to use more advanced functions beyond those accessible via DXVA2/D3D11VA, particularly encode. On Linux it has a very restricted feature set and is hard to use, but may be helpful for some use-cases desiring maximum throughput.
     176libmfx is a proprietary library from Intel for use of Quick Sync hardware on both Linux and Windows. On Windows it is the primary way to use more advanced functions beyond those accessible via DXVA2/D3D11VA, particularly encode. On Linux it has a very restricted feature set and is hard to use, but may be helpful for some use-cases desiring maximum throughput.
    183177
    184178See [wiki:Hardware/QuickSync QuickSync].
    … …  
    186180== OpenCL ==
    187181
    188 Official website:
    189 
    190 https://www.khronos.org/opencl/
    191 
    192 Currently only used in filtering (deshake and unsharp filters). In order to use OpenCL code you need to enable the build with {{{--enable-opencl}}}. An API to use OpenCL API from FFmpeg is provided in libavutil/opencl.h. No decoding/encoding is currently supported (yet).
     182[https://www.khronos.org/opencl/ OpenCL] is currently only used for filtering (deshake and unsharp filters). In order to use OpenCL code you need to enable the build with {{{--enable-opencl}}}. An API to use OpenCL API from FFmpeg is provided in `libavutil/opencl.h`. No decoding/encoding is currently supported (yet).
    193183
    194184For enable-opencl to work you need to basically install your local graphics cards drivers, as well as SDK, then use its .lib files and headers.
    … …  
    198188AMD UVD is usable for decode via VDPAU and VAAPI in Mesa on Linux.  VCE also has some initial support for encode via VAAPI, but should be considered experimental.
    199189
    200 For windows there have been port [https://ffmpeg.zeranoe.com/forum/viewtopic.php?t=4231 attempts] but nothing official yet.
     190For Windows there are port [https://ffmpeg.zeranoe.com/forum/viewtopic.php?t=4231 attempts], but nothing official has been released yet.
    201191
    202192== External resources ==