Why Was libdav1d Created

This article explores the primary motivations behind the creation of libdav1d, the popular open-source AV1 video decoder. It details how the performance limitations of the initial reference decoder prompted VideoLAN and the Alliance for Open Media to develop a highly optimized, fast, and lightweight software alternative to drive the global adoption of the AV1 video codec.

The Need for a Fast Software Decoder

The AV1 video coding format was developed by the Alliance for Open Media (AOMedia) as an open, royalty-free video buffering standard designed to replace older codecs. While AV1 offered superior compression efficiency, it was computationally expensive to decode.

When AV1 was first launched, the official reference decoder, libaom, was primarily designed for research and testing rather than real-world consumer playback. It was slow, resource-heavy, and incapable of providing smooth, real-time software decoding on most consumer hardware, such as budget laptops, older computers, and mobile devices. Because dedicated AV1 hardware decoding chips would take years to reach the market, the adoption of AV1 stalled.

The Core Motivation Behind libdav1d

The main motivation for creating libdav1d was to solve this performance bottleneck by building a highly optimized software decoder from scratch. Initiated by VideoLAN (the creators of VLC) and the FFmpeg community, and funded by AOMedia, the project aimed to make AV1 playback viable on virtually any device using only the CPU.

To achieve this, the developers of libdav1d focused on several key engineering goals:

By successfully delivering on these goals, libdav1d bypassed the hardware limitation barrier. It was quickly integrated into major web browsers like Mozilla Firefox and Google Chrome, as well as media players like VLC, effectively enabling billions of users to stream AV1 video smoothly years before AV1 hardware acceleration became mainstream.