We would be rich if we had been given a buck for every scrap of “evidence” that the era of Ultra HD (UHD) is coming and the last week has been especially generous. On the face of it opinions from an august body such as the Society of Motion Picture & Television Engineers (SMPTE) should be accorded more weight than most, but in this case the headline is only a caricature of the underlying story.
In its latest Newswatch e-newsletter, the SMPTE argues that a critical milestone on the road to UHD will soon be passed when manufacture of standard “full HD” TVs supporting 1080p resolution ceases. The newsletter points out that supposedly UHD-ready 50-inch panels are now available for under $900 and the price gap with full HD TVs is closing fast as manufacturing costs converge.
It has been argued many times that UHD does not achieve a significant quality improvement compared with 1080p on screens of that size anyway, but the wider debate, which the SMPTE touches on later in the article, is over the definition of what UHD ready means. What the article barely mentions are the other aspects of the UHD ecosystem, including encoding, security and that little matter of content availability, which all need to be in place before we can say the UHD era has truly arrived.
Of course if all premium TV sets are defined as UHD ready there is nothing to be lost buying them, but the point is that these same panels almost certainly won’t be up to the job by the time the dust has settled and services defined perhaps as “full UHD” in a few years’ time are widely available.
The SMPTE seems to contradict itself here, on the one hand suggesting that conversion by manufacturers to UHD-only TVs will be complete by late 2016 or early 2017, but in the same breath agreeing that this transition is about more than just “getting more pixels onto the screen.”
The SMPTE itself points out that UHD as currently understood is just a subset of what should be called “next generation” or “immersive” television. This includes at least three other substantial improvements in display, including 10 bit rather than 8 bit color representation to allow more shades, high dynamic range to accentuate dark and light areas so that they are visible under normal viewing conditions, and higher frame rates, so that the increased viewing angle made possible by UHD resolutions does not introduce perceptible judder or motion blur.
There is also the question of display technology, given that OLED (Organic Light Emitting Diode) displays at 1080p resolution have achieved superior feedback than LCD UHD TVs at 2160 x 3840 in some consumer tests. Admittedly the perceptual difference between full HD and UHD is not that great anyway without the corresponding improvements such as HDR and color depth, but the point is that display technology is still a significant part of the equation.
That is why LG is persisting with OLED technology for large screens despite the manufacturing yield problems that have caused all other major TV makers outside China, including Samsung, to back off. This is because the quality is so stunning, which is why we think OLED will prevail in the end, given also that LG does appear to have turned the corner and is making OLED panels not just for itself but also some Chinese makers such as TCL and Haier, although the former at least is building its own OLED TV factory. The Chinese are also believers in OLED for big screens.
But whatever the fate of OLED, UHD will come because on any type of display the higher resolution, color, HDR and frame rate together will yield significant quality improvements for at least some content. Fast moving sports action will always look better on a big screen at frame rates of 60 fps or more. Close range nature or art shots will always look better with 10 bit color. Nocturnal scenes will be immeasurably improved by HDR.
The problem then is that TVs described as UHD ready will not support these other so called next generation TV improvements, or not all of them anyway. It is not just the TV hardware but the HDMI connection itself that will stand in the way. Currently, as the SMPTE pointed out, HDMI 1.3 or 1.4 are the prevailing TV interfaces and have a maximum bit rate of 10.2 Gbps. That is enough to pass UHD content to the TV but only at a maximum frame rate of 30 per sec and only with 8 bit color.
The next version HDMI 2.0 supports up to 18 Gbps, enough for 60 frames per sec, but even that would be bust by a step up to 10 bit color as well. And there is growing conviction that even higher frame rates up to 120 per sec and 12 bit color will be desirable, which implies we would need an HDMI 3.0.
This also derails the bandwidth argument that you can deliver UHD in 15 Mbps, which is the bit rate recommended by Netflix using HEVC. With further optimization of the codec it looks like a UHD stream could be compressed down to perhaps 11 Mbps, but that is at 30 frames per second and 8 bit color. At 60 fps and 10 bit color the requirement would go up to around 25 Mbps, which would rule out all but about 4% of the world’s population at present.
The issue of bandwidth has stoked further work on video compression and in a separate story we cover one company, V-Nova, which has developed a new algorithm involving inter-frame correlation in its system called Perseus with the claim that it can deliver a UHD stream at existing frame rates in as little as 6 Mbps. What does seem certain is that HEVC is not the last word in video compression and also that bandwidth is unlikely to be the factor that holds back UHD, even in its full blown form with higher frames rates, color depth and HDR.