Your browser is not supported. Please update it.

13 January 2022

Brazilian goldrush for V-Nova as LCEVC etched into SBTVD 3.0

V-Nova’s mic-drop moment has arrived. The UK video compression specialist is about to go mainstream in Brazil, of all places – securing a single order that could be worth tens of $millions in licensing revenues over several years.

Faultline has been sitting impatiently on this major development for a number of weeks, after we initially caught wind of a game-changing email sent by Brazil’s digital terrestrial television body – the SBTVD Forum – together with the Brazilian government’s Ministry of Communications department. While the gory details aren’t strictly public, we understand it involves notifying the legal departments of every TV manufacturer, silicon maker, encoder supplier, and every other relevant video component active in Latin America that they must support LCEVC – V-Nova’s low complexity enhancement codec – to do business on the continent from 2023.

The cat is finally and officially out of the bag this week, as the SBTVD Forum published the official results of its Call for Proposals for SBTVD TV 3.0 – Brazil’s next-generation digital TV standard which is approaching the final Phase 2 testing and evaluation phases. In this publication, the Brazilian authorities have revealed the 30 candidate technologies selected to develop TV 3.0’s six system components – covering the OTA physical layer, transport layer, video coding, audio coding, captions, and application coding.

Focusing on the video coding layer, V-Nova has worked with Harmonic as the supplier supporting the MPEG-5 Part 2 video enhancement codec for TV 3.0, which has gained MPEG recognition for bringing 30% bandwidth reduction to any video codec while crunching compute resources of those notoriously thirsty next-gen codecs. This is significant because the main video coding layer of TV 3.0 is using the latest VVC (H.266) standard, based on proponents from Ateme, Interdigital, Fraunhofer and the Digital Broadcast Experts Group (DiBEG).

It’s clear that TV 3.0 wants to target start-of-the-art coding efficiency above and beyond what HEVC offers. HEVC brings 50% bitrate reduction over AVC, while VVC aims to bring 50% bitrate reduction over HEVC. We know LECVC works with both HEVC and VVC, so the decision for the SBTVD Forum to select LCEVC to support this rollout came as a no-brainer. The software is also backwards-compatible for 4K and even 8K broadcast, with moderate bandwidth increases over HD, where the video enhancement information can be sent via broadband with upscaling.

For anyone wondering whether SBTVD TV 3.0 is similar to the hybrid ATSC 3.0 standard rolling out across the US (and long established in South Korea), TV 3.0 takes many elements from ATSC 3.0 – including for the OTA physical layer, the transport layer, the application coding, HEVC encoding, HDR mapping codec, the audio coding layer, captions, and the emergency warning system manager. However, TV 3.0 is considered to be more sophisticated than ATSC 3.0 technologically, although the proof won’t be in the pudding until services go live.

For the vendors involved, we believe the big difference with SBTVD TV 3.0 is that when the specification completes, it is etched into law, unlike what the Western world is used to with recommendations and guidelines.

Our contact at the SBTVD Forum was unfortunately unable to retrieve a comment in time, but assured us it would be releasing an announcement with additional details in the coming days. We have also heard that TV 3.0 will expand to other countries across Latin America too, which would make this an even bigger deal for V-Nova and the other technology vendor candidates involved, although the SBTVD Forum couldn’t confirm this for us.

Being a technical publication on the Call for Proposals, many of the long-term business implications have not been released by the SBTVD Forum and therefore are not yet public knowledge. This is where Faultline comes in. What the deployment means is that, from 2023 onwards, the Brazilian government will start making it mandatory for all devices to be LCEVC-equipped by 2027 for all devices delivering SBTVD TV 3.0 signals.

For V-Nova, it effectively means the end of the argument about where the signs of LCEVC adoption would come from. We have known for some time about the technology’s proof in performance, which was then made a reality as a standard, while a crucial piece of the puzzle saw the licensing model slide into place last year. So, after years of work and painstaking promises, this Latin American governmental order is one of the last elusive pieces.

This single deal could be a lifeline for V-Nova, as we understand the company is currently fetching in the region of $1 million in annual revenues, which is nowhere close to offsetting its overheads. This unprofitable little vendor is now within touching distance of bragging rights – although work isn’t done yet as V-Nova won’t make a penny until 2023. Licensing agreements should then ramp up over the following four years.

V-Nova is about to become a crucial cog in TV 3.0 across Brazil, which will bring numerous improvements to the viewer experience, including better sound and image quality, more geographically segmented content according to the viewer profile, and more transparent integration between open TV and the internet.

For background on the Brazilian digital terrestrial television market, TV 1.0 began in the 1950s using the PAL-M format. It took half a century before TV 1.5 kicked off in 2007, bringing color TV, closed captions and stereo audio. TV 2.0 arrived with changes based on the ISDB-T standard out of Japan, with added support for AVC, sweeping the country with HD video, closed captioning and some elements of interactivity in the country’s first-generation DTT. This was followed by TV 2.5, which brought a constrained version of HDR, with immersive audio and the DTV Play interactive feature set.

With TV 3.0, backwards compatibility has been removed as a requirement, comprising on OTA physical layer operating on current DTT bands in Brazil and co-existing with adjacent ISDB-Tb channels. The IP-based transport layer in TV 3.0 brings a broadband interface connection to each receiver, requiring frame-accurate synchronization for video, audio and data across OTT and OTA.

Another technology being studied in TV 3.0 is 5G Standalone Broadcast, developed by silicon royalty Qualcomm. The project replaces the broadcast system currently used, which transmits the same signal to multiple homes, with another one, which sends specific content to users.

To sum up the significance of the video coding layer of SBTVD TV 3.0, anyone who wants to sell a TV set in Brazil will have to come talk to V-Nova at some point. However, we must remember that the device makers aren’t the ones paying royalties. V-Nova has set out its licensing structure for video services to pay for Usage Licenses determined by the type of video service and by the number of users it has, with prices ranging from less than one cent per user per year, up to 12 cents per user per year. The maximum fee is capped at $3.7 million a year and 50 million subscribers, while SVoD services below 100,000 are exempt.

On a rainy day later this year, we might even attempt to figure out how much money this deal could generate, once more concrete details are made available, if at all.

Wrapping this revelation up, what we have is following satisfactory completion of SBTVD TV 3.0 tests, as decided by the Deliberative Council of the SBTVD Forum together with the evaluations of the Technical, Market and Intellectual Property Modules, the Ministry of Communications has considered recommendations on a set of technologies and decided that LCEVC MPEG-5 Part 2 fits like hand in glove for the next-generation of digital television in Brazil.

V-Nova itself is choosing modesty, as the vendor is choosing not to publish its own PR around TV 3.0 or comment on the deal for the time being. The V-Nova team has put more than enough time in getting to this point – now it’s time for others to do the talking on its behalf.