IEEE University of Lahore


Q&A With Emmy Award–Winning Engineer

Gary Sullivan and his team were recognized for the HEVC standard, which improves picture quality and shortens video upload time

Photo: Courtesy of RWTH Aachen University

The standard that IEEE Fellow Gary Sullivan [above] and others developed can be found in many of the latest televisions, smartphones, and video game consoles, as well as on Netflix and other streaming services.

Sullivan led the group that developed the High-Efficiency Video Coding standard—an effort that involved nearly 100 companies and roughly 1,000 people. The HEVC standard, published in 2013, began showing up that year in Samsung phones and other products.

As co-chair of the Joint Collaborative Team on Video Coding, Sullivan and his group received a 2017 Primetime Emmy Engineering Award in October. This award is given for an outstanding achievement that has transformed the transmission, recording, or reception of television.

Sullivan began working on digital video in the late 1980s as a graduate student at the University of California, Los Angeles, when the technology was in its infancy. He started out developing some of the first video conferencing systems. Today he is a video and image technology research architect at Microsoft. In this interview with The Institute, he tells us more about HEVC technology.

Congratulations on your latest recognition. What does the award mean to you and your team?

Thank you. Most of what I’ve done has been in collaboration with other people [right], who deserve much of the credit. In fact, most of it.

It’s such an honor for our group to receive this Emmy—a validation that HEVC has had a major impact on broadcasting and represents a milestone in the history of television technology. I accepted the award at the event on behalf of our team with our other team co-chair, Jens-Rainer Ohm, director of the Institute for Communications Engineering at RWTH Aachen University, in Germany. We were joined by key members from our partner organizations, the International Organization for Standardization and International Electromechanical Commission’s Moving Picture Experts Group (MPEG) and the International Telecommunication Union’s Video Coding Experts Group (VCEG).

Tell us about your work in developing standards.

This is the third time that a collaborative team was formed to develop a new video coding standard. I served as chair of the previous effort for Advanced Video Coding, or AVC—currently the main format used for recording, compression, and distribution of video content. It was released in 2003. Before that was MPEG-2, a format for digital video used for television broadcasts and DVDs, released in 1994.

Developing HEVC was a huge effort over the course of three years. While AVC is still the format most commonly used for video today, the technology continued to advance and we made it better. Our team released HEVC in 2013, and we put into it everything we learned since the release of AVC a decade earlier.

For those who are unfamiliar, can you explain what HEVC does?

HEVC is a data-compression format for video that is about twice as good at compression as AVC. This means that with a good encoder it takes only half as much data to get the same picture quality. With HEVC, you can store twice as much video on a hard drive or upload or send twice as many videos in the same amount of time.

HEVC can also be used to achieve higher-quality video. Instead of storing or sending more footage, HEVC’s compression capability can result in higher-resolution videos for sharper pictures or in higher frame rates for more natural motion. The technology is especially useful in the latest streaming services and devices, which offer 4K ultra-high-definition pictures, high-dynamic-range color, 10-bit precision pictures, and high frame rates.

Tell us about your group’s accomplishments, beyond developing the standard.

We produced several important extensions of HEVC, which include format range extensions for alternative picture formats such as GIF or JPEG, multiview and 3D extensions for 3D applications, and screen content coding extensions for handling computer-generated content like desktop screen sharing.

Additionally, we defined metadata that can be sent to indicate what can be done with the video after decoding it, such as indicating how the video should properly be displayed. This is especially important for high dynamic range content that has a larger scale of brightness and color.

The MPEG and VCEG standards groups have now formed another new partnership called the Joint Video Experts Team. We’re planning to create the next-generation video coding standard by 2020. This international standard will have video compression capability well beyond what has been achieved by HEVC. Proposals are now being evaluated.