Refactors itx into separate 10, 12 bit functions to prevent conditional jumps. inv_txfm_add_4x4_adst_adst_0_12bpc_c: 370.9 inv_txfm_add_4x4_adst_adst_0_12bpc_avx2: 68.6 inv_txfm_add_4x4_adst_adst_1_12bpc_c: 371.0 inv_txfm_add_4x4_adst_adst_1_12bpc_avx2: 68.7 inv_txfm_add_4x4_adst_dct_0_12bpc_c: 413.1 inv_txfm_add_4x4_adst_dct_0_12bpc_avx2: 69.2 inv_txfm_add_4x4_adst_dct_1_12bpc_c: 412.7 inv_txfm_add_4x4_adst_dct_1_12bpc_avx2: 68.8 inv_txfm_add_4x4_adst_flipadst_0_12bpc_c: 378.5 inv_txfm_add_4x4_adst_flipadst_0_12bpc_avx2: 74.9 inv_txfm_add_4x4_adst_flipadst_1_12bpc_c: 378.1 inv_txfm_add_4x4_adst_flipadst_1_12bpc_avx2: 74.6 inv_txfm_add_4x4_adst_identity_0_12bpc_c: 347.8 inv_txfm_add_4x4_adst_identity_0_12bpc_avx2: 48.8 inv_txfm_add_4x4_adst_identity_1_12bpc_c: 342.7 inv_txfm_add_4x4_adst_identity_1_12bpc_avx2: 49.0 inv_txfm_add_4x4_dct_adst_0_12bpc_c: 399.2 inv_txfm_add_4x4_dct_adst_0_12bpc_avx2: 73.1 inv_txfm_add_4x4_dct_adst_1_12bpc_c: 398.7 inv_txfm_add_4x4_dct_adst_1_12bpc_avx2: 72.2 inv_txfm_add_4x4_dct_dct_0_12bpc_c: 69.6 inv_txfm_add_4x4_dct_dct_0_12bpc_avx2: 32.9 inv_txfm_add_4x4_dct_dct_1_12bpc_c: 420.5 inv_txfm_add_4x4_dct_dct_1_12bpc_avx2: 72.2 inv_txfm_add_4x4_dct_flipadst_0_12bpc_c: 405.5 inv_txfm_add_4x4_dct_flipadst_0_12bpc_avx2: 75.9 inv_txfm_add_4x4_dct_flipadst_1_12bpc_c: 404.2 inv_txfm_add_4x4_dct_flipadst_1_12bpc_avx2: 75.6 inv_txfm_add_4x4_dct_identity_0_12bpc_c: 374.1 inv_txfm_add_4x4_dct_identity_0_12bpc_avx2: 51.6 inv_txfm_add_4x4_dct_identity_1_12bpc_c: 368.0 inv_txfm_add_4x4_dct_identity_1_12bpc_avx2: 51.8 inv_txfm_add_4x4_flipadst_adst_0_12bpc_c: 368.0 inv_txfm_add_4x4_flipadst_adst_0_12bpc_avx2: 69.2 inv_txfm_add_4x4_flipadst_adst_1_12bpc_c: 370.7 inv_txfm_add_4x4_flipadst_adst_1_12bpc_avx2: 70.4 inv_txfm_add_4x4_flipadst_dct_0_12bpc_c: 393.7 inv_txfm_add_4x4_flipadst_dct_0_12bpc_avx2: 70.1 inv_txfm_add_4x4_flipadst_dct_1_12bpc_c: 392.9 inv_txfm_add_4x4_flipadst_dct_1_12bpc_avx2: 69.6 inv_txfm_add_4x4_flipadst_flipadst_0_12bpc_c: 382.2 inv_txfm_add_4x4_flipadst_flipadst_0_12bpc_avx2: 74.6 inv_txfm_add_4x4_flipadst_flipadst_1_12bpc_c: 381.3 inv_txfm_add_4x4_flipadst_flipadst_1_12bpc_avx2: 74.9 inv_txfm_add_4x4_flipadst_identity_0_12bpc_c: 346.7 inv_txfm_add_4x4_flipadst_identity_0_12bpc_avx2: 48.2 inv_txfm_add_4x4_flipadst_identity_1_12bpc_c: 347.9 inv_txfm_add_4x4_flipadst_identity_1_12bpc_avx2: 48.7 inv_txfm_add_4x4_identity_adst_0_12bpc_c: 344.7 inv_txfm_add_4x4_identity_adst_0_12bpc_avx2: 59.8 inv_txfm_add_4x4_identity_adst_1_12bpc_c: 340.5 inv_txfm_add_4x4_identity_adst_1_12bpc_avx2: 59.2 inv_txfm_add_4x4_identity_dct_0_12bpc_c: 369.8 inv_txfm_add_4x4_identity_dct_0_12bpc_avx2: 59.3 inv_txfm_add_4x4_identity_dct_1_12bpc_c: 369.5 inv_txfm_add_4x4_identity_dct_1_12bpc_avx2: 59.2 inv_txfm_add_4x4_identity_flipadst_0_12bpc_c: 353.4 inv_txfm_add_4x4_identity_flipadst_0_12bpc_avx2: 65.6 inv_txfm_add_4x4_identity_flipadst_1_12bpc_c: 350.9 inv_txfm_add_4x4_identity_flipadst_1_12bpc_avx2: 65.9 inv_txfm_add_4x4_identity_identity_0_12bpc_c: 326.1 inv_txfm_add_4x4_identity_identity_0_12bpc_avx2: 39.5 inv_txfm_add_4x4_identity_identity_1_12bpc_c: 321.6 inv_txfm_add_4x4_identity_identity_1_12bpc_avx2: 39.5
|2 weeks ago|
|doc||1 year ago|
|examples||2 weeks ago|
|include||3 weeks ago|
|package||1 year ago|
|src||4 days ago|
|tests||5 days ago|
|tools||1 month ago|
|.gitignore||9 months ago|
|.gitlab-ci.yml||5 days ago|
|CONTRIBUTING.md||1 year ago|
|COPYING||3 years ago|
|NEWS||2 months ago|
|README.md||2 months ago|
|THANKS.md||11 months ago|
|gcovr.cfg||1 year ago|
|meson.build||5 days ago|
|meson_options.txt||5 days ago|
dav1d is a new AV1 cross-platform decoder, open-source, and focused on speed and correctness.
The canonical repository URL for this repo is https://code.videolan.org/videolan/dav1d
This project is partially funded by the Alliance for Open Media/AOM.
Goal and Features
The goal of this project is to provide a decoder for most platforms, and achieve the highest speed possible to overcome the temporary lack of AV1 hardware decoder.
It supports all features from AV1, including all subsampling and bit-depth parameters.
In the future, this project will host simple tools or simple wrappings (like, for example, an MFT transform).
dav1d is released under a very liberal license, a contrario from the other VideoLAN projects, so that it can be embedded anywhere, including non-open-source software; or even drivers, to allow the creation of hybrid decoders.
The reasoning behind this decision is the same as for libvorbis, see RMS on vorbis.
The plan is the following:
- Complete C implementation of the decoder,
- Provide a usable API,
- Port to most platforms,
- Make it fast on desktop, by writing asm for AVX2 chips.
- Make it fast on mobile, by writing asm for ARMv8 chips,
- Make it fast on older desktop, by writing asm for SSSE3+ chips,
- Make high bit-depth fast on mobile, by writing asm for ARMv8 chips.
- Make it fast on older mobile, by writing asm for ARMv7 chips,
- Make high bit-depth fast on older mobile, by writing asm for ARMv7 chips,
- Make high bit-depth fast on desktop, by writing asm for AVX2 chips,
- Make high bit-depth fast on older desktop, by writing asm for SSSE3+ chips,
- Improve C code base with various tweaks,
- Accelerate for less common architectures, like PPC, SSE2 or AVX-512.
- Improve threading.
- Use more GPU decoding, when possible.
Currently, we are looking for help from:
- C developers,
- asm developers,
- platform-specific developers,
- GPGPU developers,
Our contributions guidelines are quite strict. We want to build a coherent codebase to simplify maintenance and achieve the highest possible speed.
Notably, the codebase is in pure C and asm.
See the contributions document.
There is no CLA.
People will keep their copyright and their authorship rights, while adhering to the BSD 2-clause license.
VideoLAN will only have the collective work rights.
The VideoLAN Code of Conduct applies to this project.
- Install Meson (0.47 or higher), Ninja, and, for x86* targets, nasm (2.14 or higher)
mkdir build && cd buildto create a build directory and enter it
meson ..to configure meson, add
--default-library=staticif static linking is desired
Cross-Compilation for 32- or 64-bit Windows, 32-bit Linux
If you're on a linux build machine trying to compile .exe for a Windows target/host machine, run
meson build --cross-file=package/crossfiles/x86_64-w64-mingw32.meson
or, for 32-bit:
meson build --cross-file=package/crossfiles/i686-w64-mingw32.meson
mingw-w64 is a pre-requisite and should be installed on your linux machine via your preferred method or package manager. Note the binary name formats may differ between distributions. Verify the names, and use
alias if certain binaries cannot be found.
For 32-bit linux, run
meson build --cross-file=package/crossfiles/i686-linux32.meson
- In the root directory, run
git clone https://code.videolan.org/videolan/dav1d-test-data.git tests/dav1d-test-datato fetch the test data repository
- During meson configuration, specify
meson test -vafter compiling
This project is partially funded by the Alliance for Open Media/AOM and is supported by TwoOrioles and VideoLabs.
These companies can provide support and integration help, should you need it.
Why do you not improve libaom rather than starting a new project?
- We believe that libaom is a very good library. It was however developed for research purposes during AV1 design. We think that an implementation written from scratch can achieve faster decoding, in the same way that ffvp9 was faster than libvpx.
Is dav1d a recursive acronym?
Can I help?
- Yes. See the contributions document.
I am not a developer. Can I help?
- Yes. We need testers, bug reporters and documentation writers.
What about the AV1 patent license?
- This project is an implementation of a decoder. It gives you no special rights on the AV1 patents.
Please read the AV1 patent license that applies to the AV1 specification and codec.
Will you care about <my_arch>? <my_os>?
- We do, but we don't have either the time or the knowledge. Therefore, patches and contributions welcome.
Where can I find documentation?
- The current library documentation, built from master, can be found here.