HDR, Part 3: High Dynamic Range Options

Sony BVM-X300
Sony BVM-X300

What are the options?

By Andrew Devis (doddleNEWS)

One of the keys for many players in the broadcast industry is ‘backwards compatibility.’ There are millions of SDR (Standard Dynamic Range) TVs out there so we need to make sure that when we start broadcasting HDR that those with the older TVs aren’t left out.

So when we talk about the HDR production options, the problem is that today there are lots of different ways we consume media. Once it was just over the airways, then video, then cable, satellite, disc, internet … and all of these systems have different investment requirements for their infrastructure and while upgrading systems is probably acceptable, putting in whole new broadcasting systems won’t work for many. Especially for terrestrial broadcasters (at least not in the short to medium term) – and this along with its associated cost is going to have a big effect on what system the broadcaster is going to choose when moving to HDR broadcast.

Components of HDR technology

Keepixo Components of HDR (Keepixo White Paper – ‘High Dynamic Range Video – The Future of TV Viewing Exerience’)


And because HDR is such a big deal, there are a lot of very powerful companies working very hard to try and become the dominant player with all the rewards that will bring. Organizations such as Dolby, Technicolor, Panasonic, BBC, NHK, Consumer Technology Association, etc., are rushing to get out standards, which they hope will make them the market leader – so the market is presently very turbulent to say the least.

Backwards Compatibility implementations in HDR delivery chain

Keepixo – Backwards Compatible Delivery Systems

There are also those who for various reasons aren’t that interested in backwards compatibility. For example, the makers of the new HDR disc machines don’t need to think backwards as their products will only every be played on HDR TVs. So the Consumer Technology Association has come up with a standard (HDR10) that can’t be ported backwards. But the rest need to think about backwards compatibility, and the leader at the moment would seem to be Dolby with a system that would need upgrades to present broadcast systems and probably different grades for those with HDR TVs and those with SDR TVs and (importantly for some) requires a licence fee – but is already seeing good market penetration.

The Dolby approach – PQ (Perceptual Quantization – which is based on the way our eyes actually see) is already being taken up by organisations such as Netflix and Amazon Video. The thing about the PQ approach is that it uses metadata about the broadcast monitor that the program was actually graded on to help the receiving TV match as closely as possible the render intent of the original. It also supports a wide luminance range up to 10 000 nits – which is way beyond anything the current generation (and probably the next couple of generations) can show – so it has lots of room for growth and development.

Content Production Workflow using Reference Display Metadata

Keepixo – Content Production Workflow

Another approach gaining some traction recently (especially in the UK and Japan for obvious reasons as it’s being developed by the UK’s BBC and Japan’s NHK) is called HLG (Hybrid Log-Gamma) and as far as I can tell it is a system that’s designed mainly for terrestrial broadcasters and tries to deal with the issue of having to produce 2 versions of a programme from 1 HDR original – the HDR and the SDR versions. They do this by trying to do things ‘automatically’ in a similar way to how things are done with the present TV systems. While this seems to be a good idea, from what I have seen so far the end results favours HDR but doesn’t look as good as it should on SDR sets – but it is early days for HLG.

As mentioned above, there are also other systems in development and which one will become the dominant system is very much up for grabs. In the authors view, TVs can be made now that are capable of being able to work off different standards – for example, quite a few manufacturers support both HDR10 and Dolby PQ – so I can see the next generation of TVs supporting HDR10, PQ and HLG + A N Other – assuming that consumers are happy to pay the extra licence fee to use PQ (HLG is royalty free).

While PQ is probably the most advanced solution at the moment with some implementation and support from TV manufacturers such as LG, the downside is that there is a licence fee needed to be paid to Dolby and you really need to produce both an HDR and an SDR grade which costs more (although no-doubt the tools will get better for conversion quite quickly). The end result for HDR content produced with Dolby PQ is that it is going to be a much more faithful reproduction of the render intent of the producer especially with the use of metadata from the producer to ‘calibrate’ the users TVs to show the end product looking as near to the original as possible.

As for HLG (as far as I understand it) it doesn’t use metadata from grading monitor to TV, but uses metadata from the camera to adjust the content for the screen. Also, not quite as big a range of brightness and ‘automatic(ish)’ down conversion of HDR content to SDR (there’s a white paper on it from the BBC that looks great but it has lots of maths in it so I didn’t get as far into it as I probably should have – see link at the end of this blog). Big up-side for HLG – royalty free and the backing of big names in broadcast – even if that is quite geographically limited at the moment. And, technically, HDR could be argued as being anything greater than SDR – so while it may not try to be as wide-ranging as PQ (PQ specs up to 10 000 nits while HLG goes to 5 000 nits) – the end results are sure to still look pretty impressive and anyway there aren’t any 5000 nit TVs as yet, let along 10 000 nit TVs!

Let me finish part 3 with one more important thought.

In many ways the whole HDR thing is still in its ‘Wild West’ phase – nothing much is really settled yet. For example, the HDR/WCG colour space that has been talked about for several years (rec. 2020) is now being overtaken by rec. 2100 – same colour space but support for HLG and interestingly support for up to 8K UHD and fascinatingly enough support for HDR in HDTV (1920 x 1080) all with a wide range of frame rates – this makes the market even wider with openings for HDR/WCG/HFR (High Frame Rate) on smaller TVs without the need to go to UHD!!

HDMI standards are changing – 2.0 going to 2.1 and with it support for dynamic metadata and who knows how that will affect HDR production, broadcast and decoding at the users TVs let alone how broadcasters can or will try to implement that?

Even colour spaces are up in the air with Y’C’bC’r looking like it may not be suitable for HDR broadcast going foward and so new spaces such as ICtCp are being proposed, as in this paper from Dolby.

So, because of all this, there are risks. The leading standard of today may be the backwater of tomorrow and who knows what that will mean for your investment, production choices and workflows? But as you probably can image, with risk can sometimes come great rewards!

Another good read is the BBC HLG White Paper.

Check out part 1 and part 2 of my HDR primer.

Be the first to comment

Leave a Reply

Your email address will not be published.