Glossary

What are AV1 Golden Frames?

A golden frame is a special kind of reference frame in the AV1 video codec standard. It acts as a long-term reference in the encoding process and is different from a standard reference frames that are used to define a GOP or Group of Pictures. Golden frames are more important in AV1’s multi-layer reference frame hierarchy and this priority lets them remain in particularly for an extended time, occupying targeted reference frame buffers.

The choice of golden frames is an important one that the encoder must make. A lot of things such as complexity of scene, distance in time of frames, and rate distortion optimization are taken into account in order to choose the golden frames.

Once a frame is designated as a Golden Frame, they are signaled in the bit-stream as follows –

  • whether Golden Frames are used or not, and
  • what are their parameters.

Adding all of this information into the bitstream means that during the reconstruction phase of the decoding process, the decoder will be able to properly use these frames and more importantly, retain them in the decoder buffers appropriately.

Why are Golden Frames useful?

Golden frames are useful in boosting compression efficiency because not only are high quality and are long-term references, but they significantly improve the coding efficiency of the subsequent frames specially in scenes with low motion or repetitive elements, resulting in lower bit-rates without reducing video quality.

Furthermore, when looking at the references, golden frames hav superior prediction capabilities for frames that are far apart in time, reducing the dependence on the frequency of using intra-frame coding (we’ve talked about I-frames in the past).

Not only does AV1 encoding use Golden Frames to save bits, but because they are error-resilient, they are used to significantly enhance AV1’s error resilience performance. In lossy or error-prone networked video scenarios, golden frames serve as stable synchronization and recovery reference points for multiple application variants, greatly improving the robustness of the entire system.

AV1 is capable of adaptive golden frame refresh mechanism which varies the rate of golden frame insertion based on the characteristics of the content and encoding parameters. This adaptive technique optimizes the trade-off of compression efficiency and computational complexity, allowing encoders to adapt their practices to specific content types and delivery targets.

Encoder and Decoder Implications of Golden Frames

Because golden frames increase compression efficiency, they also add more computational complexity to both the encoder and decoder. Anyone writing a codec has to use golden frames carefully to achieve a proper trade-off between quality, bitrate efficiency, and corresponding level of computational resource utilization.

Especially, at the decoder, it’s crucial to handle golden frame buffers optimally. In order to fully benefit from golden frames, AV1 decoders have to properly store, recall, and employ these frames throughout the decoding process.

Visionular focuses on the use of AI in video compression (for H.264/AVC, HEVC, and AV1) that enables our video codecs to deliver the highest possible compression efficiency at the best encoding speeds!

Our tech is used by the largest content providers globally to deliver the best user experience and to reduce their OPEX by 25 – 30% at least!

To learn more,