When it comes to the user experience for next-generation big screen video, 4K resolution is increasingly viewed as just the starting point, with other technologies, most notably HDR (high dynamic range) but also higher frame rates and wider colour gamut, providing the added spice that will convince consumers to buy new television sets and seek out upgraded streaming or broadcast services. Now there is the intriguing possibility that HDR could actually lead to a more diverse and potentially fragmented market for next-generation video.
First, we could see a choice of 4K/UHD television sets that support HDR or do not support it. Second, there is growing support for the idea of a stepping-stone broadcast and streaming format that is noticeably better than 1080i HDTV and nothing like as bandwidth hungry as the â€˜premium UHDâ€™ that many broadcasters will decide to hold out for (which combines the yet-to-be agreed and standardized combination of 4K resolution, HDR, wider colour gamut and higher frame rates) and which is ready to deploy much sooner.
â€œSome people believe it is better to have high-definition TV with HDR and high frame rate like p60 than to have UHD without HDR and with lower frame rates,â€ says Thierry Fautier, VP of Video Strategy at Harmonic Inc, which provides encoders, transcoders and media storage among other headend and playout solutions. What he is talking about could be referred to as super-HD or HD-plus. It includes at least two of the next-generation technologies that until fairly recently were assumed to be the preserve of UHD television.
High dynamic range improves the contrast range for video, with a better spread between dark and light that is visible to consumers. It results in noticeably better detail. Importantly, it provides this effect on HD video as well as 4K/UHD. But as Fautier points out, there are competing specifications for supporting HDR today, as witnessed by the previews of new television sets at CES in January.
Even once it becomes clear which specification(s) will win out, there is likely to be a market for television sets that support HDR and sets that do not, even if major set makers prefer HDR as a way to sell more expensive displays. Fautier believes the set market will then be determined by the amount of HDR content that is available. If broadcasters provide UHD channels with HDR support then the market for non-HDR ultra high-definition will quickly disappear, he reckons. â€œAnd it makes sense that they should push hard with HDR and high frame rate [for UHD],â€ he argues.
Nevertheless, high-definition with HDR and high frame rate could be an option on digital terrestrial TV and for OTT services, Fautier believes. Both have relatively limited bandwidth compared to satellite or cable. A terrestrial broadcaster might choose this instead of UHD because they will not have to sacrifice other channels in their multiplex to make room for super-HD, at least while they await regulatory decisions about over-the-air spectrum allocation. But with more bandwidth available to them, cable and satellite operators will prefer to direct their efforts, including marketing efforts, towards UHD, Fautier argues.
For Fautier, super-HD has a lot of technical merit but will only be a gap-filler before operators deploy UHD. Then, once one broadcaster delivers UHD, others will have to follow suit. We will be reporting soon about how Ericsson, another leading provider of compression solutions, views the potential role for HD with HDR and if anything they are more bullish about its prospects.
Robert Comas, a technology consultant specializing in product development who was Product Development and Planning Manager at Sharp Electronics Europe for five years until December (and previously Product Development Manager and also Hardware Development Manager at Sharp Electronics Spain) believes that additional manufacturing and licensing costs will lead to â€˜standardâ€™ UHD products and â€˜specialâ€™ products featuring UHD with HDR. â€œBut who knows what will happen when Samsung or some other big player promotes HDR heavily,â€ he adds.
Though there will be additional silicon costs of up to US$ 2-3 per television set, depending on the version of HDR used, this will be short-term. It is the optical components of the television sets that will determine any long-term premium for HDR sets, he points out.
HDR technologies can be divided into two groups, according to Comas. The first uses a conventional standard dynamic range (SDR) signal that is upscaled to HDR and then rendered. There are no additional SoC (system on chip) costs associated with this approach. The second takes an HDR signal and renders it. This latter approach is then divided into two groups: a mixed HDR/SDR signal that uses dynamic metadata, such as Dolby Vision, and an exclusive HDR signal that uses static metadata, such as the BD alliance format. The metadata includes maximum luminance, maximum value for the average of each frame, CIE value for RG and B, and white point, etc.
An exclusive HDR signal that uses static metadata does not create any additional costs in terms of the SoC, Comas says. â€œAll you need is simple HEVC/VP9 decoding capability, without any polling for metadata.â€
The HDR/SDR mixed signal approach with dynamic metadata is more complex. He says it requires dual HEVC real-time decoding and requires that the television set â€˜pollsâ€™ dynamic metadata and adjusts to the most appropriate colour mapping. This currently requires additional circuitry, maybe as FPGA or a companion chip, at the cost of around $2-3. But by 2016 these functions should be supported by a single SoC. â€œTherefore we do not need to think about additional cost or additional circuitry beyond the near future,â€ he predicts.
The more important cost considerations, therefore, are the optical components of the television set such as the backlight (LED) and the panel in order to achieve high luminance. â€œPractically speaking, the edge-light type LCD module does not meet the requirements so a â€˜direct typeâ€™ LCD backlight system is needed and that is very costly,â€ Comas reports. He adds that if the chosen HDR approach has to support high colour gamut as well as dynamic range, that could mean adding Quantum Dot onto the LCD module, which is expensive.