QSVEnc

by rigaya

日本語版はこちら＞＞

This software is meant to investigate performance and image quality of HW encoder (QSV) of Intel. There are 2 types of software developed, one is command line version that runs independently, and the nother is a output plug-in of Aviutl.

• QSVEncC.exe ... Command line version supporting transcoding.
• QSVEnc.auo ... Output plugin for Aviutl.

Downloads & update history

rigayaの日記兼メモ帳＞＞
github releases

Install

Install instructions for Windows and Linux

Build

Build instructions for Windows and Linux

System Requirements

Windows

Windows 10/11 (x86/x64)
Aviutl 1.00 or later (QSVEnc.auo)

Linux

Debian/Ubuntu (QSVEncC)
Fedora (QSVEncC)
Requires Broadwell CPU or later.
It may be possible to run on other distributions (not tested).

Precautions for using QSVEnc

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND.

Usage and options of QSVEncC

Option list and details of QSVEncC

Major Features

Common to QSVEnc.auo / QSVEncC

• Encoding using QuickSyncVideo
  • H.264/AVC
  • H.265/HEVC (8bit/10bit)
  • MPEG2
  • VP9 (8bit/10bit)
  • AV1 (8bit/10bit)
• Encode mode of QuickSyncVideo
  • CQP (Fixed Quantization)
  • CBR (Constant bitrate)
  • VBR (Variable bitrate)
  • AVBR (Adaptive Variable bitrate)
  • QVBR (Quality based Variable bitrate)
  • LA (Lookahead mode)
  • LA-HRD (HRD compatible Lookahead mode)
  • ICQ (Constant Quality)
  • LA-ICQ (Constant Quality with Lookahead)
  • VCM (Video Conference mode)
• Quality preset of 7 steps
• Interlaced encoding (by PAFF)
• Supports setting of codec profile & level, SAR, colormatrix, maxbitrate, GOP len, etc...
• Calculation of ssim/psnr of the encode
• Supports various vpp(video pre-processing) filters
  • MFX (Media Fixed function) filters
    • resize
    • deinterlace (normal / bob / it)
    • detail enhancement
    • denoise
    • image stablizier
  • OpenCL filters
    • deinterlacer
      • afs (Automatic field shift)
      • nnedi
    • decimate
    • mpdecimate
    • colorspace conversion
      • hdr2sdr
      • tonemap (libplacebo)
    • delogo
    • subburn
    • resize
      • bilinear
      • spline16, spline36, spline64
      • lanczos2, lanczos3, lanczos4
      • libplacebo
    • transpose / rotate / flip
    • padding
    • deband
      • deband
      • libplacebo
    • noise reduction
      • knn (K-nearest neighbor)
      • pmd (modified pmd method)
    • edge / detail enhancement
      • unsharp
      • edgelevel (edge ​​level adjustment)
      • warpsharp

QSVEncC

• Supports QSV(hw) decoding
  • MPEG2
  • H.264/AVC
  • HEVC
  • VP8
  • VP9
  • AV1
• Supports various formats such as avs, vpy, y4m, and raw
• Supports demux/muxing using libavformat
• Supports decode using libavcodec

QSVEnc.auo (Aviutl plugin)

• Audio encoding
• Mux audio and chapter
• afs (Automatic field shift) support

Supported Features

This depends on the version of QSVEnc, the generation of the GPU, and also the GPU driver.

CPU Gen	GPU Gen	Windows	Linux
SandyBridge	Gen6	i5 2410M
IvyBridge	Gen7
Haswell	Gen7.5	i3 4170 i7 4610Y
Broadwell	Gen8	i7 5500U	i7 5500U
SkyLake	Gen9
KabyLake	Gen9.5	i5 7500	i7 7700K
CoffeeLake	Gen9.5
CommetLake	Gen9.5
Cannonlake	Gen10
IceLake	Gen11	i5 1035G7	i5 1035G7
TigerLake	Gen12
RocketLake	Gen12	i7 11700K	i7 11700K
AlderLake	Gen12	i9 12900K i3 N305	i9 12900K
	DG2	Arc A380	Arc A380

Supported HW Encoder (@ Windows OS)

CPU Gen	GPU Gen	MPEG2	H.264 PG	H.264 FF	HEVC PG	HEVC FF	VP9 FF	AV1 FF
SandyBridge	Gen6	8bit	8bit
IvyBridge	Gen7	8bit	8bit
Haswell	Gen7.5	8bit	8bit
Broadwell	Gen8	8bit	8bit
SkyLake	Gen9	8bit	8bit		8bit
KabyLake	Gen9.5	8bit	8bit	8bit	10bit
CoffeeLake	Gen9.5	8bit	8bit	8bit	10bit
CommetLake	Gen9.5	8bit	8bit	8bit	10bit
CannonLake	Gen10	-	-	-	-	-	-	-
IceLake	Gen11	8bit	8bit	8bit	10bit	10bit	10bit
TigerLake	Gen12	8bit	8bit	8bit	10bit	10bit	10bit
RocketLake	Gen12	8bit	8bit	8bit	10bit	10bit	10bit
AlderLake	Gen12	8bit	8bit	8bit	10bit	10bit	10bit
	DG2			8bit		10bit	10bit	10bit

Supported HW Decoder

CPU Gen	GPU Gen	MPEG2	H.264	HEVC	VP8	VP9	AV1
SandyBridge	Gen6	8bit	8bit
IvyBridge	Gen7	8bit	8bit
Haswell	Gen7.5	8bit	8bit
Broadwell	Gen8	8bit	8bit		8bit
SkyLake	Gen9	8bit	8bit	8bit	8bit
KabyLake	Gen9.5	8bit	8bit	10bit	8bit	8bit
CoffeeLake	Gen9.5	8bit	8bit	10bit	8bit	8bit
CommetLake	Gen9.5	8bit	8bit	10bit	8bit	8bit
CannonLake	Gen10	-	-	-	-	-	-
IceLake	Gen11	8bit	8bit	10bit	8bit	10bit
TigerLake	Gen12	8bit	8bit	12bit		12bit	10bit
RocketLake	Gen12	8bit	8bit	12bit		12bit	10bit
AlderLake	Gen12	8bit	8bit	12bit		12bit	10bit
	DG2	8bit	8bit	12bit		12bit	12bit

Auto GPU selection in multi GPU envinronment (Windows Only)

QSVEncC will automatically select a GPU depending on the options used, when there are multiple GPUs available which support QSV. --device option can be used to specify on which GPU to run manually.

1. Select GPU which supports...
   Items below will be checked whether the GPU supports it or not

• Codec, Profile, Level
• Additionally, below items will be checked if specified
  • 10bit depth encoding
  • HW decode

2. If there are multiple GPUs which suports all the items checked in 1. and 2., GPU below will be prefered.

• GPU which has low Video Engine(VE) utilization
• GPU which has low GPU core utilization

The purpose of selecting GPU with lower VE/GPU ulitization is to assign tasks to mulitple GPUs and improve the throughput of the tasks.

Please note that VE and GPU ulitization are check at the initialization pahse of the app, and there are delays in values taken. Therefore, it is likely that the multiple tasks started at the same time to run on the same GPU, and divided into multiple GPUs, even if the options are supported in every GPUs.

QSVEnc source code

• MIT license.
• This software depends on oneVPL, ffmpeg, libplacebo, libhdr10plus, libdovi, libass, tinyxml2, dtl, clRNG, ttmath & Caption2Ass. For these licenses, please see the header part of the corresponding source and license.txt.

About source code

Windows ... VC build
Linux ... makefile + gcc/clang build

Character code: UTF-8-BOM
Line feed: CRLF
Indent: blank x4