There is considerable talk at the moment – both from consumer electronics manufacturers and broadcast equipment vendors – about higher resolution production and delivery. Currently, the buzz is around 4K, Ultra HD as the consumer industry calls it; or UHDTV1 as SMPTE calls it. 4K in the video context is generally regarded as twice HD in each direction.
Some broadcast vendors have 4K products available for sale, not least a number of established and new entrants to the camera market. These tend to use large sensors to achieve an acceptable signal to noise ratio. This means a shallow depth of field, which makes them more suited to drama rather than to live television.
4K signals are inherently difficult to handle in a broadcast infrastructure. Most current solutions use four parallel HD-SDI connections, but there are no established standards.
In the file-based world, cameras often have their own unique resolutions – the Blackmagic 4.5K or the Red 5K, for example – and of course their own file formats which need to be processed before integrating into broadcast-style workflows.
At the same time that some are advocating higher spatial resolutions, engineers are pushing the causes of high dynamic range (HDR), which increases the available colour gamut as well as allowing for brighter highlights and deeper blacks; and high frame rates (HFR) that increase the perceived picture sharpness, especially in scenes with a lot of movement, without the need to add to the pixel count.
The biggest interest at present is given to higher numbers of pixels and lines. So-called 4K Ultra HD is 3840 x 2160, and 8k is 7680 x 4320. SMPTE in its work towards standardisation and recommendations is currently calling these two spatial resolutions UHDTV1 and UHDTV2 respectively. In considering whether to move from today’s HD to 4K, it is worth thinking about what the eye can actually see.
The definition of 20/20 vision is slightly arcane, but it equates to a pixel spacing of a little under an arc minute. So, if a screen is 1920 pixels wide, it needs to subtend an angle of roughly 30˚ to the eye for the individual to be able to see all the detail in a smooth image without obvious pixellation.
If you work through the geometry, at a viewing distance of two metres, 1920 x 1080 HD is a perfect match for the eye on a 44” diagonal screen. Many households will be comfortable with a 42” television, and this is almost perfect for watching high definition TV. To appreciate twice as many pixels in each direction, as in 4K, the horizontal field of view needs to be almost 60˚. With the same viewing distance of two metres, you are now up to a 104” diagonal screen.
So to appreciate 4K at home, you either need to watch from further away or change viewing habits. The richness of a 4K picture could be explored, which implies a certain sort of content. Big dramas and nature documentaries, for example, would benefit from it, but for fast moving sports and events it would be a distraction.
That said, a number of broadcasters and production companies are originating some material in 4K today. Most is being finished in HD, but some productions – such as Netflix’s ‘House of Cards’ – are going all the way to an Ultra HD delivery format.
There is another significant reason for shooting in 4K today; it brings higher resolution into the edit suite, giving more flexibility in post. Shots can be reframed without losing quality and the extra pixels can mask artefacts in other post processes like extreme colour correction.
A simple increase in spatial resolution is not the only proposal being put forward for Ultra HD and beyond. One of the most persuasive is the use of higher frame rates.
When the content of the programme involves rapid movement, like sport, then increasing the frame rate yields a remarkable improvement in perceived sharpness. This really has to be seen to be believed: without changing the number of pictures in an image, but having more images a second, it appears much sharper and clearer.
High frame rate video implies moving on a further step, to 100 or 120 progressive frames a second. For most viewers, 100fps HD is a better experience than 25fps (or 50i) 4K, at least on material where an increase in sharpness is appropriate.
As with increased spatial resolution, the benefits of HFR will only be appreciated on some material. When Peter Jackson chose to make his ‘Hobbit’ trilogy in what was, for the cinema, a high frame rate (48fps), there was an outcry from many critics because “it did not look like a movie”.
DYNAMIC RANGE AND COLOUR
HD, as with SD before it, has a relatively limited colour gamut, defined by the capabilities of the electronics at the time. The gamut was also limited by the capabilities of the TV display, which had a relatively low maximum output. The dynamic range, and colour gamut, had to be limited by the ‘white’ level the screen could display, and therefore the levels of shading available below that.
New display technologies mean we can now get a wider dynamic range on screen, increasing the subtlety of the colour shading. Finally, increasing the white point in each of the colour channels means you push the colour gamut out in all three directions, allowing the system to show colours which were previously impossible in video.
HDR could be seen to be an easy fix, as moving from eight bits to 12 per colour is only an increase of 50% in raw data, whereas going from 1080 to 4K is 400%. Increasing the frame rate from 50i to 100p is also a 400% growth in data.
However, pushing up the screen brightness makes the flicker inherent in video pictures more obvious: the flicker fusion threshold is in part dependent on absolute brightness. So HDR will probably need to be combined with HFR to produce a satisfying solution.
What is clear is that the different forms of extending the resolution will be appropriate for different types of content. 4K pictures might be very attractive for big nature documentaries; high frame rate for sport; and extended dynamic range and colour for atmospheric dramas. But a period drama would be negatively affected with the application of HFR, for example.
The conclusion is that any future successful Ultra HD system has to not only support high spatial resolutions, HDR and HFR, but allow each of them to vary by each individual piece of content. It makes sense, therefore, to choose technology platforms which are software-defined.
Just as we include screen resolution as part of the metadata in a file wrapper, so in future we will be able to define frame rates and bits per colour. Most wrappers already support this in theory, so this is not a huge leap.
Software-defined platforms will work with today’s content and, provided the hardware has sufficient processing power, with any element of Ultra HD or beyond in future.
Individual product vendors would decide what functionality they would develop, depending on what their systems are required to deliver. Some may choose to licence different functionality in different ways: maybe you could pay per hour when it is being used for HDR, for example. That is a commercial decision: the key issue is that the technology must be flexible.
4k ROLL OUT
It is notoriously difficult to predict consumer trends with any accuracy. But there are some points we can consider for Ultra HD. First, the consumer electronics industry is keen to promote 4K resolution. The computer industry has driven up screen resolutions so the raw glass is widely available for 4K resolution, so television sets are relatively easy to manufacture.
On the other hand, just because the sets are in the stores does not mean they necessarily change consumer behaviour. 3D television received massive promotion but was largely ignored by consumers. It may be that we have the situation where the next generation of television sets are 4K-ready – and indeed probably still 3D-ready – but few will want to connect to 4K services.
Will broadcasters see a business case in re-equipping playout and content management infrastructure for Ultra HD, and pay for the extra bandwidth to deliver the high bitrate signals? That will depend upon consumer interest. If consumers are clamouring for content the business case may look stronger.
On the other hand, will there be a market for on-demand downloads of 4K content, if it takes significantly longer than HD? In the age of instant gratification, on demand generally means just that: now, not in a couple of hours when the movie has downloaded. Will immediacy trump quality?
If Ultra HD downloads are seen as worth waiting for then that will be an extremely disruptive trend. Those downloads will be served by new players like Netflix and Amazon Prime, not the traditional broadcast providers of leading content. That could be a big shift in power in the industry.
Finally, is 4K enough? What if audiences do not find the incremental change enough of a wow factor? Those who know that 8k will be coming soon may well want to wait. Japan is proposing to broadcast Super Hi-Vision from the 2020 Olympics. Some other broadcasters, such as the BBC, appear to be lukewarm towards 4K because 8k is coming.
The significance of this is not just that it future-proofs any system. It also means that broadcasters, production companies and service providers can experiment with 4K delivery, either as a proof of concept in-house or on a trial channel.
4K production, and Ultra HD delivery, is one example of a broader challenge for the industry. Across all the areas of technology, change is accelerating. New resolutions, new ways of working, new means of recording, sharing and transporting content, new platforms for consumption and more are appearing ever more frequently.
This barrage of constant change leads to uncertainty. Which of these new ideas and capabilities will find a resonance with audiences? In what areas, therefore, should broadcasters, production companies and facilities make investments, and will those investments see a return through new revenues?
The solution is to move, as far as possible, towards a software-defined architecture, in conjunction with vendors who are dedicated to the industry and have a proven record of continuing innovation.
The broadcaster that plans for a software-defined future can at least be confident of agility: you can use the technology you have today to experiment with, or roll out new features and capabilities, as your business develops and responds in the future.
James Gilbert is Co-Founder and CEO of Pixel Power, a supplier of broadcast playout solutions with integrated graphics to media companies globally.