This project prepares a minimalist Docker image with FFmpeg. It compiles FFmpeg from sources following instructions from the Compilation Guide.
You can install the latest build of this image by running docker pull jrottenberg/ffmpeg.
This image can be used as a base for an encoding farm.
You can use jrottenberg/ffmpeg or jrottenberg/ffmpeg:3.3 to get the latest build based on ubuntu.
Note : I've made ubuntu the default after 3.1
You'll find centos based image using ffmpeg:X.Y-centos or ffmpeg:centos to get the latest.
alpine images ffmpeg:X.Y-alpine to get the latest.
scratch images ffmpeg:X.Y-scratch to get the latest. (Scratch is an experimental image containing only FFmpeg and libraries)
Format is ffmpeg:MAJOR.MINOR-VARIANT with MAJOR.MINOR in :
- 2.8
- 3.0
- 3.1
- 3.2
- 3.3
- 3.4
- 4.0
- 4.1
- snapshot
and VARIANT in :
- alpine
- centos
- nvidia
- scratch
- ubuntu
- vaapi
Recent images:
snapshot-vaapi 74mb
snapshot-ubuntu 86mb
snapshot-scratch 20mb
snapshot-nvidia 640mb
snapshot-centos 97mb
snapshot-alpine 35mb
4.1-vaapi 73mb
4.1-ubuntu 85mb
4.1-scratch 20mb
4.1-nvidia 640mb
4.1-centos 97mb
4.1-alpine 34mb
4.0-vaapi 73mb
4.0-ubuntu 83mb
4.0-scratch 20mb
4.0-nvidia 639mb
4.0-centos 97mb
4.0-alpine 34mb
3.4-vaapi 71mb
3.4-ubuntu 83mb
3.4-scratch 18mb
3.4-nvidia 637mb
3.4-centos 97mb
3.4-alpine 32mb
3.4 83mb
3.3-vaapi 71mb
3.3-ubuntu 83mb
3.3-scratch 18mb
3.3-nvidia 637mb
3.3-centos 96mb
3.3-alpine 31mb
3.3 82mb
3.2-vaapi 83mb
3.2-ubuntu 83mb
3.2-scratch 18mb
3.2-nvidia 623mb
3.2-centos 96mb
3.2-alpine 32mb
3.1-vaapi 83mb
3.1-ubuntu 82mb
3.1-scratch 17mb
3.1-nvidia 623mb
3.1-centos 96mb
3.1-alpine 32mb
3.1 81mb
3.0-ubuntu 82mb
3.0-scratch 17mb
3.0-nvidia 623mb
3.0-centos 96mb
3.0-alpine 31mb
2.8-vaapi 82mb
2.8-ubuntu 81mb
2.8-scratch 17mb
2.8-nvidia 622mb
2.8-centos 95mb
2.8-alpine 30mb
$ curl --silent https://hub.docker.com/v2/repositories/jrottenberg/ffmpeg/tags/?page_size=500 | jq -cr ".results|sort_by(.name)|reverse[]|.sz=(.full_size/1048576|floor|tostring+\"mb\")|[.name,( (20-(.name|length))*\" \" ),.sz,( (8-(.sz|length))*\" \"),.last_updated[:10]]|@text|gsub(\"[,\\\"\\\]\\\[]\";null)"
# If you want to compare the one you have locally
$ docker images | grep ffmpeg | sort | awk '{print $1 ":" $2 "\t" $7 $8}'
Please use Github issues to report any bug or missing feature.
ffmpeg version 3.3.4 Copyright (c) 2000-2017 the FFmpeg developers
built with gcc 5.4.0 (Ubuntu 5.4.0-6ubuntu1~16.04.4) 20160609
configuration: --disable-debug --disable-doc --disable-ffplay --enable-shared --enable-avresample --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-gpl --enable-libass --enable-libfreetype --enable-libvidstab --enable-libmp3lame --enable-libopenjpeg --enable-libopus --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx265 --enable-libxvid --enable-libx264 --enable-nonfree --enable-openssl --enable-libfdk_aac --enable-postproc --enable-small --enable-version3 --extra-cflags=-I/opt/ffmpeg/include --extra-ldflags=-L/opt/ffmpeg/lib --extra-libs=-ldl --prefix=/opt/ffmpeg
libavutil 55. 58.100 / 55. 58.100
libavcodec 57. 89.100 / 57. 89.100
libavformat 57. 71.100 / 57. 71.100
libavdevice 57. 6.100 / 57. 6.100
libavfilter 6. 82.100 / 6. 82.100
libavresample 3. 5. 0 / 3. 5. 0
libswscale 4. 6.100 / 4. 6.100
libswresample 2. 7.100 / 2. 7.100
libpostproc 54. 5.100 / 54. 5.100
configuration:
--disable-debug
--disable-doc
--disable-ffplay
--enable-shared
--enable-avresample
--enable-libopencore-amrnb
--enable-libopencore-amrwb
--enable-gpl
--enable-libass
--enable-libfreetype
--enable-libvidstab
--enable-libmp3lame
--enable-libopenjpeg
--enable-libopus
--enable-libtheora
--enable-libvorbis
--enable-libvpx
--enable-libwebp
--enable-libx265
--enable-libxvid
--enable-libx264
--enable-nonfree
--enable-openssl
--enable-libfdk_aac
--enable-postproc
--enable-small
--enable-version3
--extra-cflags=-I/opt/ffmpeg/include
--extra-ldflags=-L/opt/ffmpeg/lib
--extra-libs=-ldl
--prefix=/opt/ffmpeg
Capture output from the container to the host running the command
docker run jrottenberg/ffmpeg \
-i https://url/to/media.mp4 \
-stats \
$ffmpeg_options - > out.mp4
docker run jrottenberg/ffmpeg -stats \
-i https://archive.org/download/thethreeagesbusterkeaton/Buster.Keaton.The.Three.Ages.ogv \
-loop 0 \
-final_delay 500 -c:v gif -f gif -ss 00:49:42 -t 5 - > trow_ball.gif
docker run -v $(pwd):$(pwd) -w $(pwd) jrottenberg/ffmpeg:3.4-scratch \
-stats \
-i https://www.jell.yfish.us/media/jellyfish-20-mbps-hd-hevc-10bit.mkv \
-c:v libx265 -pix_fmt yuv420p10 \
-t 5 -f mp4 test.mp4
The image has been compiled with X265 Multilib. Use the pixel format switch to change the number of bits per pixel by suffixing it with 10 for 10bits or 12 for 12bits.
Let's assume original.gif is located in the current directory :
docker run -v $(pwd):$(pwd) -w $(pwd)\
jrottenberg/ffmpeg:3.2-scratch -stats \
-i original.gif \
original-converted.mp4
Thanks to qmfrederik for the vaapi ubuntu based variant
jrottenberg/ffmpeg:vaapi or jrottenberg/ffmpeg:${VERSION}-vaapi
- Run the container with the device attached /dev/dri from your host into the container :
docker run --device /dev/dri:/dev/dri -v $(pwd):$(pwd) -w $(pwd) jrottenberg/ffmpeg:vaapi [...]
- Have the Intel drivers up and running on your host. You can run
vainfo(part of vainfo package on Ubuntu) to determine whether your graphics card has been recognized correctly. - Run ffmpeg with the correct parameters, this is the same as when running ffmpeg natively.
Thanks to ShaulMyplay for the nvidia based variant
Supports nvenc only on all ffmpeg versions, and hardware decoding and scaling on ffmpeg >= 4.0
- Install nvidia latest drivers on host machine.
- Install nvidia-docker on host machine.
- Run container using "--runtime=nvidia" flag and use supported ffmpeg hwaccel options
Hardware encoding only example:
docker run --runtime=nvidia jrottenberg/ffmpeg:2.8-nvidia -i INPUT -c:v nvenc_h264 -preset hq OUTPUT
Full hardware acceleration example:
docker run --runtime=nvidia jrottenberg/ffmpeg:4.1-nvidia -hwaccel cuvid -c:v h264_cuvid -i INPUT -vf scale_npp=-1:720 -c:v h264_nvenc -preset slow OUTPUT
docker run -it --entrypoint='bash' jrottenberg/ffmpeg
for i in ogg amr vorbis theora mp3lame opus vpx xvid fdk x264 x265;do echo $i; find /usr/local/ -name *$i*;done
See Dockerfile-env to update a version
- FFMPEG_VERSION: GNU Lesser General Public License (LGPL) version 2.1
- OGG_VERSION: BSD-style license
- OPENCOREAMR_VERSION: Apache License
- VORBIS_VERSION: BSD-style license
- THEORA_VERSION: BSD-style license
- LAME_VERSION: GNU Lesser General Public License (LGPL) version 2.1
- OPUS_VERSION: BSD-style license
- VPX_VERSION: BSD-style license
- WEBP_VERSION: BSD-style license
- XVID_VERSION: GNU General Public Licence (GPL) version 2
- FDKAAC_VERSION: Liberal but not a license of patented technologies
- FREETYPE_VERSION: GNU General Public License (GPL) version 2
- LIBVIDSTAB_VERSION: GNU General Public License (GPL) version 2
- LIBFRIDIBI_VERSION: GNU General Public License (GPL) version 2
- X264_VERSION: GNU General Public License (GPL) version 2
- X265_VERSION:GNU General Public License (GPL) version 2
# Add / fix stuff
${EDITOR} templates/
# Generates the Dockerfile for all variants
./update.py
# Test a specific variant
docker build -t my-build docker-images/VERSION/
# Make sure all variants pass before CI
find ffmpeg/ -name Dockerfile | xargs dirname | parallel --no-notice -j 4 --results logs docker build -t {} {}
Commit the templates files THEN all the generated Dockerfile for a merge request. So it's easier to review the template change.