Codecs and DVB – setting standards, not adoption

DVB has played a long-standing role in defining implementation guidelines for codecs. In 2015, the scope for this was extended from broadcast to broadband OTT delivery. Typical standards from MPEG range over more than 600 pages and are designed for multiple applications. It is not conceivable that TV set manufacturers implement them in full. Either they make a ‘bet’ on what will be used in the future and design their hardware chip accordingly, or they follow the DVB specifications. DVB has therefore been a trendsetter, guaranteeing codec interoperability and also serving as a basis for other regional specifications.

Driving deployment?

But how far does this trendsetter’s role go towards driving deployment? Having a profile is certainly a strong catalyst for a codec’s market introduction, but DVB does not set roadmaps as such. For example, while the VVC profile was released in 2022, Germany is currently undergoing a partial migration from MPEG-2 to MPEG-4 AVC, a DVB specification published in 2005.

There are several factors affecting the speed of adoption: does the new codec enable something disruptive that can provide a return on investment? Are there regulatory measures encouraging adoption? What is the licensing policy? What is the implementation complexity?

Otherwise, based purely on efficiency, introducing a new codec is in general only a long-term investment, because legacy codecs must be supported until switch-off. On broadcast, this costs additional transmission and encoding. On broadband, it costs additional encoding, although dynamic encoding approaches can mitigate this1. Still, most services, including the newer FAST channels, continue to bet on MPEG-4 as a ‘one fits all’ codec. In 2019, 91% of all streaming services were using MPEG- AVC2; and we estimate that had only fallen to 70% by 2024.

An increasing number of OTT service providers are migrating towards HEVC or AV1, while some terrestrial services (e.g., Germany and Netherlands) have fully switched to HEVC. Additionally, UHD services use HEVC. But the migration away from AVC on OTT will take a few more years, especially given the unknowns around licensing.

For VVC, while there are limited deployments in China and India, the most significant launch to date will be the roll-out in Brazil this year of SBTVD 3.0, with a profile based on DVB’s VVC specifications. This will be a big step, as several TV chipsets, which are often designed for the global market, will be compatible with the DVB VVC specifications.

What’s next?

For countries not anticipating a hard switch to a next-generation video codec, DVB is contemplating a backwards-compatible solution that can continue to serve both legacy and newer devices with lower total transmission costs.

While the principle of such a solution is clear, having a scalable system raises new questions on security and hardware requirements of end devices. The discussions are complex, including with respect to evaluating picture quality improvements.

With its role as a reference well established, DVB will continue to follow the newest codecs standardized, subject – as always – to strong backing from its Members. Still, the ‘automatic’ market adoption of newer, more efficient codecs has become less likely with no tangible new application – UHD is already possible and 8K has not, so far, been a success.

Newer applications go beyond 2D screens, but what is DVB’s role? We can leverage expertise among Members, develop operational recommendations that may be delivered as a “market representation partner” to other organizations (or accept their requirements in return), and seek commonalities between mobiles, headsets and TVs through cooperation with other organizations to enable economies of scale. Could DVB’s future focus on stronger cooperation for harmonization beyond DVB’s ecosystem, and more practical community efforts – a kind of “meta DVB”?