Battle of the Video Codecs: Coding-Efficient VVC vs. Royalty-Free AV1

The explosion of Internet video inspires new software tools to deliver your favorite shows

3 min read
A person's hands holds up a transparent screen with a play button in the middle.
Image: iStockphoto

Video is taking over the world. It’s projected to account for 82 percent of Internet traffic by 2022. And what started as an analog electronic medium for moving visuals has transformed into a digital format viewed on social media platforms, video sharing websites, and streaming services.

As video evolves, so too does the video encoding process, which applies compression algorithms to raw video so the files take up less space, making them easier to transmit and reducing the bandwidth required. Part of this evolution involves developing new codecs—encoders to compress videos plus decoders to decompress them for playback—to support higher resolutions, modern formats, and new applications such as 360-degree videos and virtual reality.

Today’s dominant standard, HEVC (High Efficiency Video Coding), was finalized in 2013 as a joint effort between the Moving Picture Experts Group (MPEG) and the Video Coding Experts Group (VCEG). HEVC was designed to have better coding efficiency over the existing Advanced Video Coding (AVC) standard, with tests showing an average of 53 percent lower bit rate than AVC while still achieving the same subjective video quality. (Fun fact: HEVC was recognized with an Engineering Emmy Award in 2017 for enabling “efficient delivery in Ultra High Definition (UHD) content over multiple distribution channels,” while AVC garnered the same award in 2008.)

HEVC may be the incumbent, but two emerging options—VVC and AV1—could upend it.

Envisioned as HEVC’s successor, Versatile Video Coding (VVC) is another collaboration between MPEG and VCEG set to be finalized by mid-2020. VVC aims for a 30 to 50 percent bit-rate reduction for the same perceptual quality as HEVC, but with an estimated 10 times or more encoding complexity compared to its predecessor.

“Video content for HD, UHD, or even higher resolutions can take up a significant amount of data, so cutting that amount of data by half is a big advantage,” says Benjamin Bross, project manager at the video coding and analytics department of the Fraunhofer Institute for Telecommunications in Berlin and one of the specification editors for VVC.

As its name suggests, VVC is meant to be versatile, supporting applications such as gaming and adaptive streaming, a technique that adapts bit rate and video resolution to network conditions. “Typically, various applications are supported by different profiles of a standard, and this has been the case for HEVC,” Bross says. “However, in most devices, only the Main profile of HEVC has been implemented, leading to a lack of support for wider applications. VVC is expected to tackle a broader range of application scenarios in one profile, which is an advantage over HEVC.”

Yet similar to HEVC, VVC is a royalty-bearing video codec. This means that individuals and companies implementing VVC for their products will need to pay patent licensing fees.

That’s where AV1 comes in. Launched in 2018, AV1 is the first project to come out of the Alliance for Open Media, a consortium of tech companies—including Apple, Facebook, Google, Microsoft, Mozilla, and Netflix—that offers “open, royalty-free, and interoperable solutions for the next generation of media delivery.”

“Overall, I believe that having multiple competing codecs is a good thing”

Unlike HEVC and VVC, AV1 is a royalty-free video codec using Google’s VP9 open video codec as a base. According to Mozilla, it’s meant to “replace AVC as the predominant video format for the web and to compete with the HEVC codec, so high-quality video can be shared freely and efficiently on the open web platform.” In terms of performance, tests have shown between 30 to 40 percent better compression than AVC and HEVC, but lower encoding speed.

So how do these two video codecs measure up against each other? The BBC’s tests found that VVC offers more bit rate savings than AV1, but at the expense of processing time. This is in line with a recent study published in IEEE Access comparing VVC and AV1’s performance in healthcare applications, which showed that VVC provided the best coding efficiency, thereby outperforming AV1, but with both video codecs still subject to long compression times.

Based on these tests, it seems like VVC comes out at the top—but it’s not a clear-cut victory. “Selecting which video compression standard to use is not an easy task and is often application-specific,” says Andreas Panayides, a senior research fellow at the Electronic Health Lab of the University of Cyprus and co-author of the healthcare applications study. “Incurred royalties are likely to play their own role in deciding which codec better fits a video application deployment and rollout.”

For Kedar Tatwawadi, a Ph.D. student researching data compression at Stanford University under the guidance of Tsachy Weissman, who helped develop the compression algorithm depicted in the HBO TV series Silicon Valley, more extensive tests are needed before a winner emerges in the battle between VVC and AV1. “Personally, I support codecs which allow the benefits of improved compression to reach the masses at a faster rate,” Tatwawadi says. “Overall, I believe that having multiple competing codecs is a good thing, as it drives researchers toward making significant strides in the field of video compression.”

The Conversation (0)

The Future of Deep Learning Is Photonic

Computing with light could slash the energy needs of neural networks

10 min read

This computer rendering depicts the pattern on a photonic chip that the author and his colleagues have devised for performing neural-network calculations using light.

Alexander Sludds
DarkBlue1

Think of the many tasks to which computers are being applied that in the not-so-distant past required human intuition. Computers routinely identify objects in images, transcribe speech, translate between languages, diagnose medical conditions, play complex games, and drive cars.

The technique that has empowered these stunning developments is called deep learning, a term that refers to mathematical models known as artificial neural networks. Deep learning is a subfield of machine learning, a branch of computer science based on fitting complex models to data.

Keep Reading ↓ Show less