Video Coding Standards: A Comparison

Our previous stop was the progression of video coding standards through AVC, HEVC, and VVC (Read: https://www.dhirubhai.net/pulse/vvc-origin-evolution-sraas/). Without any further ado, let’s move on to the detailed differences among AVC, HEVC, and VVC.?

With the target of getting twice-better coding efficiency than the previous coding standard (AVC/H.264), HEVC/H.265 adopted the hierarchical structure with a coding unit (CU), supporting a separate prediction unit, and a transform unit partitioning. Although HEVC/H.265 introduced several new coding features in addition to AVC/H.264, such as a quad-tree based splitting, most of the algorithms in HEVC/H.265 can be traced to AVC/H.264. The AVC/H.264 algorithms were improved in HEVC/H.265 by increasing precision or adapting them to take advantage of increased computational resources without significant variations in common coding tools.

In the case of VVC/H.266, it has up to 50% better compression rate for the same perceptual quality, with support for lossless and subjectively lossless compression. It is projected to cut the bitrate roughly in half compared to HEVC. The expected encoding complexity is 10 times that of HEVC and the decoding complexity is about twice that of HEVC. Further, work on high-bit depth support (12 and 16 bits per component) is ongoing.

The following table compares the technical differences in the three standards:

A major streaming standard on track to outperform HEVC and giving a tough time to VVC is AV1 (AOMedia Video 1) developed by Alliance for Open Media. It has the advantage of being an open-source royalty-free video codec. Further, AV1 provides a 30% bit-rate reduction at the same visual quality compared to H.265/HEVC. It uses lossy compression, which means that the encoded version is not the same as the original version such that losses are imperceivable to the human eye. For example, instead of sending the full frame of video 20 times a second for a 20 fps (frames per second) video, only the changes will be sent. Thus, the video encoder will only track changes and differences thereby achieving the desired compression. In addition, the file size will be less, so the encoding time will also be less. Another significant advantage of AV1 is that many companies have already adopted AV1. Also, AV1 hardware decoders are being rapidly embedded in new processors.?

The following table illustrates the key points of comparison between AV1 and VVC.

Stay tuned for our next article next week. Do leave your comments and suggestions below.