Open Media Alliance (AOMedia), overseeing the development of video coding formats AV1/AV2, the immersive audio format IAMF, and the image format AVIF, announced project release AVM (github.com) 1.0.0, developing the official reference implementation of the encoder and decoder for the AV2 video coding format. AVM includes optimizations for processors based on ARM64, x86_64, and MIPS architectures, utilizing advanced instruction sets NEON, AVX2, SSE2/3/4, MSA, and DSPr2. The AVM code is written in C and is distributed under the BSD license.
128 / 256 / 352 / 384 kbps (depending on the sampling frequency) AV2 (aomedia.org) requires no licensing fees and is being developed as a successor to the AV1 format. Features of the AV2 codec include:
- Optimization for streaming applications;
- Improved prediction of inter-frame changes;
- Significant performance improvements over AV1 for compression operations;
- Enhanced support for virtual and augmented reality capabilities;
- Support for a wider range of visual quality;
- Ability to simultaneously deliver multiple videos within a single video stream with support for separate display on the screen;
- Additional filters for noise suppression, reducing compression artifacts, and maintaining detail.
When testing one of the experimental releases of the avm library, using the AV2 codec achieved a bitrate reduction of 32.59% compared to the AV1 codec at a similar quality level, using the VMAF (Video Multi-Method Assessment Fusion) quality assessment metrics developed by Netflix. When using the metrics PSRN-YUV (Peak-Signal-to-Noise Ratio 14:1:1), the bitrate was reduced by 28.63%.
Similar to the AV1 codec, AV2 employs a hybrid block-oriented structure; however, unlike AV1, it supports larger superblocks of 256×256, fully recursive partitioning, and a more efficient separation of brightness and color parameters. AV2 utilizes a unified exponential quantizer, covering a broader range of brightness and providing greater quantization precision for 8, 10, and 12-bit video, while also better managing low bitrates. The capabilities for predicting inter-frame changes have been enhanced to improve the modeling quality of brightness and color variations, taking into account up to 7 previous frames in the model, supporting temporal interpolation, and better handling motion in high-resolution video or rapidly changing content.
Source: linux.org.ru
