#encoder #cross-platform #av1 #quality #h264 #av1an

av1an-core

Cross-platform command-line AV1 / VP9 / HEVC / H264 encoding framework with per scene quality encoding [Core library]

7 unstable releases

0.4.3 Aug 30, 2024
0.4.1 Mar 13, 2023
0.4.0 Dec 19, 2022
0.3.1 Jan 17, 2022
0.1.0 Aug 30, 2021

#445 in Video

Download history 4/week @ 2024-07-07 1/week @ 2024-07-14 10/week @ 2024-07-21 9/week @ 2024-07-28 13/week @ 2024-08-04 11/week @ 2024-08-11 1/week @ 2024-08-18 149/week @ 2024-08-25 35/week @ 2024-09-01 25/week @ 2024-09-08 32/week @ 2024-09-15 53/week @ 2024-09-22 35/week @ 2024-09-29 14/week @ 2024-10-06 14/week @ 2024-10-13 12/week @ 2024-10-20

79 downloads per month
Used in 2 crates

GPL-3.0 license

210KB
6K SLoC

Av1an

av1an fully utilizing a 96-core CPU for video encoding

Discord server CI tests

Buy Me A Coffee

Av1an is a video encoding framework. It can increase your encoding speed and improve cpu utilization by running multiple encoder processes in parallel. Target quality, VMAF plotting, and more, available to take advantage for video encoding.

For help with av1an, please reach out to us on Discord or file a GitHub issue.

Av1an Book

Features

  • Hyper-scalable video encoding
  • Target Quality mode, using VMAF control encoders rate control to achieve the desired video quality
  • VapourSynth script support
  • Cancel and resume encoding without loss of progress
  • Minimal and clean CLI
  • Docker images available
  • Cross-platform application written in Rust

Usage

For complete reference, refer to CLI or run av1an --help

Encode a video file with the default parameters:

av1an -i input.mkv

Or use a VapourSynth script and custom parameters:

av1an -i input.vpy -v "--cpu-used=3 --end-usage=q --cq-level=30 --threads=8" -w 10 --target-quality 95 -a "-c:a libopus -ac 2 -b:a 192k" -l my_log -o output.mkv

Supported encoders

At least one encoder is required to use Av1an. The following encoders are supported:

Note that Av1an requires the executable encoder. If you use a package manager to install encoders, check that the installation includes an executable encoder (e.g. vpxenc, SvtAv1EncApp) from the list above. Just installing the library (e.g. libvpx, libSvtAv1Enc) is not enough.

Installation

av1an can be installed from package managers, cargo.io, or compliled manually. There are also pre-built Docker images which include all dependencies and are frequently updated.

For Windows users, prebuilt binaries are also included in every release, and a nightly build of the current master branch is also available.

Package managers

Arch Linux & Manjaro: pacman -S av1an Cargo: cargo install av1an

Manual installation

Prerequisites:

Optional:

  • L-SMASH VapourSynth plugin for better chunking (recommended)
  • DGDecNV Vapoursynth plugin for very fast and accurate chunking, dgindexnv executable needs to be present in system path and an NVIDIA GPU with CUVID
  • ffms2 VapourSynth plugin for better chunking
  • bestsource Vapoursynth plugin for slow but accurate chunking
  • mkvmerge to use mkvmerge instead of FFmpeg for file concatenation
  • VMAF to calculate VMAF scores and to use target quality mode

VapourSynth plugins on Windows

If you want to install the L-SMASH or ffms2 plugins and are on Windows, then you have two installation options. The easiest way is using the included plugin script:

  1. Open your VapourSynth installation directory
  2. Open a command prompt or PowerShell window via Shift + Right click
  3. Run python3 vsrepo.py install lsmas ffms2

Dependencies

~27–43MB
~804K SLoC