Prospects are looking up on two fronts for the third generation ATSC 3.0 digital terrestrial standard in the US a year ahead of deployment in 2020. Firstly, there is growing conviction that the major networks will collaborate to provide a coherent nationwide service exploiting new capabilities, notably ability to centralize transmissions by banding multiple TV stations in a single market using SFN (Single Frequency Network) technology. This follows the announcement at NAB Show New York in October 2018 that Fox, NBC, Univision, Pearl TV and Nexstar Media Group would collaborate over introduction of ATSC 3.0.
Secondly, the features of ATSC 3.0 have received surprisingly strong endorsement from consumers in a recent survey, suggesting that it can help the networks compete more strongly than they perhaps expected with pure play OTT alternatives. The research was conducted in consumer lab testing by research consultancy Magid in conjunction with Pearl TV and the Phoenix Model Market project, which is a test bed for ATSC 3.0 deployment. The main finding was that the combination of enhanced video, especially 4K with HDR (High Dynamic Range), and immersive audio was very compelling and likely to bring many viewers back to broadcast TV if widely deployed. Overall 91% of consumers in the lab said they were either “interested” or “very interested” in using a service with the features enabled by ATSC 3.0. Surprisingly perhaps immersive audio had strongest appeal, with ability for users to boost specific portions of the sound using a dialogue enhancement feature and also customize individual audio tracks scoring highly.
Also encouraging for the networks was that 80% of consumers said they were either “interested” or “very interested” in purchasing an ATSC 3.0 TV or appropriate add-on device. This is significant because there was fear that ATSC 3.0’s lack of backwards compatibility therefore requiring new TVs or add-on devices would restrict its appeal initially. That is why deployment has been held back till next year.
More generally, customization was identified by survey participants as a primary attraction of ATSC 3.0, helping them find what they want and interact with content in new ways. Three other key themes emerged, one being the importance of over the air signal access enabled by ATSC 3.0, giving easier access to live content. Then naturally, interactivity is crucial since that is new for over the air TV in the US, while the final theme cited was inclusivity, although that seems no different from customization since it involves personalizing content for family and friends.
The findings are interesting in the context of ATSC 3.0’s timing, coming up to a decade after the first deployments of the second-generation DVB-T2 standard in Europe and a few Asia Pacific countries. Although nominally a generation ahead ATSC 3.0 is often compared with DVB-T2 and shares a lot of underlying technology, including support for SFN operation which is critical to its success as we will see.
It is true that having been developed later, ATSC 3.0 does incorporate some technical advances, most notably integration between over the air and broadband services, having been designed from the ground up for hybrid operation. By contrast DVB-T2 needs HbbTV to support full hybrid broadband/broadcast services.
That apart, there are differences in technical parameters, of which three areas stand out. One is that ATSC 3.0 supports two higher QAM modulations than DVB-T2, 1024 QAM and 4096 QAM, which enables it to carry more data per 6 MHz channel, up to 57 Mbps instead of 38 Mbps, which makes it better able to support 4K with HDR. Secondly it supports bonding of two RF channels, which again increases potential bit rate. Thirdly it supports LDM (Layer Division Multiplexing), which optimizes hybrid services that combine fixed delivery to traditional roof top aerials with mobile services in a single RF channel, despite both having different requirements for error correction. Essentially it multiplexes the two signals at different levels of resilience so that the mobile layer is encoded with a powerful low-rate forward error correction (FEC) code which means receivers can decode even at very low SNR (Signal To Noise) below 0 dB. This is a significant advance because until now DTT networks worldwide have been optimized solely for fixed rooftop reception.
Despite these advances though the features popular with US consumers illuminated by that survey could all have been delivered over DVB-T2 with the help of HbbTV. We described last week how France Télévisions has been testing substitution of a commercial with a targeted spot on DTT households equipped with HbbTV-compliant set tops, highlighting interactive features of DVB-T2.
For networks the main point about ATSC 3.0 is its ability to underpin a national service combining high levels of quality with incorporation of broadband and interactivity. This is where SFN comes in. In fact, the networks have indicated they will emulate the European model of SFN operation as they migrate towards a single centralized transmission infrastructure. This could well be built and managed by a single entity, just like Europe where Arqiva in the UK and TDF in France for example have managed the DTT infrastructure on behalf of all broadcasters. This makes more efficient use of spectrum, expanding total overall coverage while reducing operational costs, to the extent US networks should overcome their previous anathema for such arrangements.
This all boils down to SFN and the advantages it brings. Before SFN came along with DVB-T2, terrestrial signals were transmitted via MFN (Multi Frequency Network). Neighboring transmitters used different frequencies to avoid self-interference at receivers since signals would arrive slightly later from more distant transmitters that were still within range. This made inefficient use of spectrum because a given channel would require multiple frequencies within a given signal area to avoid the self-interference.
SFNs were then devised, based on an ingenious yet very simple concept embodied in OFDM (Orthogonal Frequency Division Multiplexing) invented at Bell Labs in 1966. It brings multiple transmitters located at different distances from any receiver into effective synchronization by dividing signals into large numbers of symbols each carrying small numbers of digital bits so that each can be sent at a low bit rate. This means each symbol takes longer to arrive at receivers, which means that the effect of the time difference between their reception from different transmitters is reduced. It means that symbols can be protected against interference resulting from the variable transmission time by inserting smaller buffers, known as guard intervals, between them. Because each symbol takes longer to arrive, the offset resulting from variable time delays from different transmitters is smaller in proportion so that a shorter buffer will suffice. That way each symbol has been fully received by every destination client from every transmitter within range before the first bit of the next one is decoded, avoiding any interference.
It remains to be seen whether SFN within ATSC 3.0 will be the catalyst that binds together hundreds of US TV stations into a coherent centralized infrastructure. Until recently such an initiative would have faced strong resistance as well as logistical challenges. But the competitive landscape has changed drastically and the need to reduce costs while opening up new revenue streams, for example through advanced advertising, is becoming overwhelming. This is generating consolidation so that the number of stations is being reduced, making it easier to build and test ATSC 3.0 networks.
There is then the question of whether the US will follow the European multiplex operator model around a single entity running the DTT infrastructure, but the odds against that are shortening.