README.md 4.23 KB
Newer Older
1 2
# dav1d

Jean-Baptiste Kempf's avatar
Jean-Baptiste Kempf committed
3
**dav1d** is a new **AV1** cross-platform **d**ecoder, open-source, and focused on speed and correctness.
4 5 6

The canonical repository URL for this repo is https://code.videolan.org/videolan/dav1d

Jean-Baptiste Kempf's avatar
Jean-Baptiste Kempf committed
7 8
This project is partially funded by the *Alliance for Open Media*/**AOM**.

9 10
## Goal and Features

11
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.
12

13
It supports all features from AV1, including all subsampling and bit-depth parameters.
14 15 16 17 18

In the future, this project will host simple tools or simple wrappings *(like, for example, an MFT transform)*.

## License

19
**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.
20

21
The reasoning behind this decision is the same as for libvorbis, see [RMS on vorbis](https://lwn.net/2001/0301/a/rms-ov-license.php3).
22 23 24 25 26

# Roadmap

The plan is the folllowing:

27
### Reached
28 29 30
1. Complete C implementation of the decoder,
2. Provide a usable API,
3. Port to most platforms,
31
4. Make it fast on desktop, by writing asm for AVX-2 chips.
32

33 34 35 36 37
### On-going
5. Make it fast on mobile, by writing asm for ARMv8 chips,
6. Make it fast on older desktop, by writing asm for SSE chips.

### After
38
7. Improve C code base with [various tweaks](https://code.videolan.org/videolan/dav1d/wikis/task-list),
39 40
8. Accelerate for less common architectures,
9. Use more GPU, when possible.
41 42 43 44 45 46

# Contribute

Currently, we are looking for help from:
- C developers,
- asm developers,
Jean-Baptiste Kempf's avatar
Jean-Baptiste Kempf committed
47
- platform-specific developers,
48
- GPGPU developers,
Jean-Baptiste Kempf's avatar
Jean-Baptiste Kempf committed
49
- testers.
50

51
Our contributions guidelines are quite strict. We want to build a coherent codebase to simplify maintenance and achieve the highest possible speed.
52 53 54 55 56 57 58 59 60 61 62

Notably, the codebase is in pure C and asm.

We are on IRC, on the **#dav1d** channel on *Freenode*.

See the [contributions document](CONTRIBUTING.md).

## CLA

There is no CLA.

63
People will keep their copyright and their authorship rights, while adhering to the BSD 2-clause license.
64 65 66 67 68 69 70 71 72

VideoLAN will only have the collective work rights.

## CoC

The [VideoLAN Code of Conduct](https://wiki.videolan.org/CoC) applies to this project.

# Compile

73
1. Install [Meson](https://mesonbuild.com/) (0.47 or higher), [Ninja](https://ninja-build.org/), and, for x86* targets, [nasm](https://nasm.us/) (2.13.02 or higher)
74
2. Run `meson build --buildtype release`
75
3. Build with `ninja -C build`
76

77 78 79 80 81 82 83 84 85
# Run tests

1. During initial build dir setup or `meson configure` specify `-Dbuild_tests=true`
2. In the build directory run `meson test` optionally with `-v` for more verbose output, especially useful
   for checkasm

# Run testdata based tests

1. Checkout the test data repository
86

87 88 89 90 91 92
   ```
   git clone https://code.videolan.org/videolan/dav1d-test-data.git tests/dav1d-test-data
   ```
2. During initial build dir setup or `meson configure` specify `-Dbuild_tests=true` and `-Dtestdata_tests=true`
3. In the build directory run `meson test` optionally with `-v` for more verbose output

93 94 95 96
# Support

This project is partially funded by the *Alliance for Open Media*/**AOM** and is supported by TwoOrioles and VideoLabs.

Jean-Baptiste Kempf's avatar
Jean-Baptiste Kempf committed
97
These companies can provide support and integration help, should you need it.
98 99 100 101 102 103


# FAQ

## Why do you not improve libaom rather than starting a new project?

104
- We believe that libaom is a very good library. It was however developed for research purposes during AV1 design.
105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
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?

- Yes.

## Can I help?

- Yes. See the [contributions document](CONTRIBUTING.md).

## I am not a developer. Can I help?

- Yes. We need testers, bug reporters, and documentation writers.

## What about the AV1 patent license?

121
- This project is an implementation of a decoder. It gives you no special rights on the AV1 patents.
122 123

Please read the [AV1 patent license](doc/PATENTS) that applies to the AV1 specification and codec.
124 125 126 127 128

## 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.