CE vendors are starting to push Ultra HD Premium television receivers, but what does it mean for the consumer, what can they watch in UHD and what are the implications upstream for content distributors and creators?
ITU recommendation BT.2020 standardised UHD television parameters, opening the door to the delivery of 4K content to the viewer. 2020 defines the resolutions, 2160P and 7680P, and frame rates up 120fps. The recommendation included a new colorimetry: a wider colour space with new colour primaries. It also recommended the uses of an alternative electro-optical transfer function (EOTF) to the existing gamma of cathode ray tubes (CRT) that would allow for an increased dynamic range. Surprisingly, gamma was not formally defined until 2011 in BT.1886, a bit late for the CRT.
Further work was needed by the standards body to arrive at a full specification for the new EOTF that would support a higher dynamic range, and this was released as BT.2100.
The 2020 colour primaries are as yet not fully achievable with current consumer technology, and are more of an aspiration. The DCI P3 is being used as a step on the way.
With 2020 and 2100, all the pieces are in place for the roll-out from camera to consumer of a new improved viewing experience. The key components for the initial phase:
- Ultra HD resolution
- Phase 1, 3840 x 2160, progressive
- High frame rate (HFR)
- Phase 1, 50/59.94 progressive frame rate
- Wide colour gamut (WGC)
- An aspiration to reach the new primaries
- High Dynamic Range (HDR)
- Perceptual quantisation (PQ) and hybrid log gamma
Later developments will extend operations to ultimately support 8K (7680P) at 120 fps.
Where will the content come from?
The first two sources of content are movies and live sports. Many movies are shot and edited at 4K resolution, so distribution as UHD television content is as simple as resizing from 4096 × 2160 to 3840 × 2160. Movies are still largely shot at the old mechanical film rate of 24fps, and colour space is likely to be DCI P3 unless there is a trim pass to grade to Rec.709.
Sport, well those guys are champing at the bit to use the full gamut of the new standards. 2160P scanning at 50 or 60 fps, along with HDR/WGC.
Why the divergence? Why are the movie guys sticking with the film look, and not, except in a few case, going for a higher frame rate?
The Film Look
There is an element of chicken and egg about the film look. Cinema developed within the constraints of the technology that was possible at the time, yet now the film look has become intimately associated with what was possible with twentieth century technology.
The frame rate was fixed at 24fps when a standard was needed for projection and audio reproduction. Much of the grammar of motion reproduction stemmed from the constraints of 24fps and a 180-degree shutter. Audiences learnt to accept the aliasing of vehicle wheels and motion blur as “the movies”.
DPs and directors learnt to keep the camera motion smooth, through tripods cranes, tracks and the Steadicam. They accepted that panning speeds should be limited aside from effects like whip pans.
However, now that digital cinematography is not restricted to 24fps, there is a body of aficionados who insist that this low frame rate is integral to cinema as an artform. The argument runs that cinema is not reality, but an artistic medium for story-telling. The film look is part of the suspension of reality as we immerse ourselves in the movie.
In contrast, television, especially sports, seeks reality. The viewer wants to feel that they are right in there with the action.
Directors are concerned with the look of their productions, the artistic intent. But there is another group who have a different set of motives, CE manufacturers are selling technology. At first look, both parties look to improve the viewing experience, but underneath both must turn out new product for the continuation of their businesses in very different ways.
Director want better pixels to enhance the look of their productions, the CE guys want better pixels to sell us new TVs. We can enjoy a movie made with 1930s technology or within its constraints, but the manufacturers’ need technological obsolescence to sell new products.
Film and Video
Photochemical film had constraints on frame rate, video had constraints on dynamic range, colour volume, and until recently, frame rate. As film and video converge, the question “what do we want it to look like” has come to the fore. As discussions around The Hobbit and latterly Billy Lynn's Long Halftime Walk demonstrate, high frame rate can be an emotive issue.
There are several issues at play here. One is around increasing static definition. As reproduction moves to 4K resolution, motion artefacts become more visible unless we move to higher temporal resolution, high frame rate.
Can the cinema experience, the detachment from reality, happen at high frame rate? Are the older generation just used to 24fps, whereas the video game generation have grown up with a different set of expectations?
The projection of film prints had many artefacts, film weave, a resolution of about 1K equivalent, low brightness, and an unnatural motion cadence stemming from the triple flash projection used to avoid flicker. Now we have digital projection, these artefacts cannot be used to mask poor temporal resolution with what was effectively noise.
One could say that an imaging system that exploits the advantages of the 2020 and 2100 standards gives directors a toolkit that they can use as they deem fit. They can stick with legacy standards or increase frame rate, colour volume and dynamic range.
How they use the tools will be different for a sports director compared to a movie director. The old constraints of the film and video technologies have been lifted, the scales fall from our eyes.
It seems to be a given that we will finally throw off the coarse compression of a century ago that was interlace, along with the associated artefacts of jaggies and poor vertical resolution. HFR will become the gold standard for sports production.
Modern digital cinema cameras can capture wide colour gamut at HDR and high frame rate. Seen here the Canon C700 (L) and Sony F65.
HDR and BT.2100
The game changer is going to be HDR. Ironically, not part of 2020, it has been found in viewing tests to be the improvement over legacy video that viewers most appreciate. HDR is specified in 2100, approved four years after 2020 UHD spec. One could argue it’s not high dynamic range, but regular dynamic range. What we had before was poor dynamic range, restricted by tube technology in both cameras and displays. Film negative has had a wide dynamic range for decades. The colorist has always had space to work, to rescue highlights, to pull out detail in the shadows, to move around the colour space.
I can go out and buy an HDR display today, but what can I watch? Much of our content distribution network remains eight-bit resolution, when ten or ideally twelve bits are needed to distribute HDR content.
Over-the-air awaits the rollout of new standards like ATSC 3.0. The DVB has updated their coding specification to include UHD HEVC coding with HDR. For broadcasters the rollout is going to resurrect all the challenges of the digital and HD switchovers, with issues around bandwidths and whether there is a need to simulcast (where would the free spectrum come from?).
Cable and satellite operators will also need to reallocate bandwidths as appropriate. The big problem for them is the tens of millions of legacy set-top boxes that are only 8-bit.
Over the top can already deliver HDR using the HEVC codec with the Main 10 profile. The UHD Blu-ray format has been approved and that supports HDR. However, much of the content is only 24fps, so avoids the restrictions on bandwidth over older HDMI interconnects.
The HDMI Snag
For viewers, plugging it all together will have its own set of problems. For the carriage of UHD HDR video and static metadata the HDMI 2.0a is needed (2.0b if using hybrid log gamma), but only the very latest consumer products have this interface. The yet to be released HDMI 2.1 will support dynamic metadata, which allows HDR parameters to be changed scene-by-scene, or even to the frame level. Many early 4K products have the older HDMI 1.4 connections, their owners are going to be disappointed, but those are the risks of being an early adopter.
So how are the CE guys getting around the current lack of content? Uprezzing, colour space expansion, all sorts of sometimes dubious tricks to give the consumer an HDR, WCG, HFR picture and to hell with artistic intent. I guess we are all familiar with the ‘vivid’ setting on TVs, it gives that punch in the retailer’s showroom. Once consumers tire of these gimmicks, and want real UHD, then they are going to need a complete swap-out of all their A/V equipment to really enjoy 2020/2100 as it was intended.
Related Editorial Content
Europe’s TV standards body DVB has taken a further step towards Ultra HD TV (UHD) commercial deployment by approving UHD-1 Phase 2 at a recent meeting of its Steering Board. This second phase of the UHD commercial requirements adds support…
Better pixels are coming, but there is much detail to be resolved. Will consumers be confused? Will a “4K” receiver also support high dynamic range (HDR), high frame rate (HFR) and wide colour gamut (WGC)? And where will the content…
The EBU (European Broadcasting Union) has included two enhanced versions of HD among four alternative profiles for broadcasters to adopt on the road way towards future immersive services that incorporate all aspects of full Ultra HDTV (UHDTV). These are detailed…