We sat down with a few members of NEP’s CTO Group for a round-table discussion of 4K and 1080p HDR. See their thoughts below on considerations, limitations, and benefits as these technologies develop in the marketplace.
Meet the Contributors:
Scott Rothenberg - SVP, Technology & Asset Management, NEP Group
Marc Segar - Director, Technology, NEP Australia
Rob Newton - Director of Engineering & Technology, NEP UK & Ireland
Dan Turk - Chief Engineer, NEP US Mobile Units
Where are you seeing clients decide to make a switch to 4K? What types of shows, clients and projects do you see using this new technology? And what seems to be the driving force behind it?
Rob: We are seeing some people make the switch, mostly in high-end sport. Occasionally we see something like a Royal Event for posterity purposes. The driving force appears to be prestigious events and wanting a premium project.
Dan: In the US, we have seen very little interest in 4K in the sports world, and I would say there have been few requests for 4K overall. The only segment 4K seems to be growing in is entertainment and corporate events - and most of the driving force behind it seems to be streaming television, not traditional broadcast.
Marc: I think the driving force is marketing. People all have 4K TVs, so it’s easy to market 4K content.
From a technology perspective, what are the major considerations when looking at 4K versus 1080p? What would you ask your client to consider - or what would you consider alongside them - if they told you “we want to start doing this in 4K”?
Rob: The major problems for live production are the number of signals. For example, if you have a 25-camera sporting event, the number of 1080 you are now handling goes up by a factor of 4/5. This means you need a much larger infrastructure. Now, as we move into IP, that infrastructure is a bit easier to scale.
I would ask our clients to consider what other signals need to be converted because not every specialist camera is available in 4K. I would ask them to consider archive formats and, of course, storage size because files are so much bigger.
Marc: Yeah, for us it’s mostly EVS channel count, bandwidth, and storage is a massive thing.
Dan: We see many of our clients move to 1080P rather than 4K, because they don’t want to give up any tools or toys, and the cost to scale up to go full 4K can be steep. With 4K, the specialty cameras can get very expensive – that is a consideration.
The next hurtle is size of production. We are very limited with switchers and routers when you get to 4K. This will all be solved in the next few years with the gen2 4K devices coming out, but we aren’t quite there yet.
Also, you have to consider the cost of converters for all the feeds for everybody else who isn’t doing 4K. Most stadiums and smaller productions cannot take 4K.
Scott: I agree with all of this completely. I think, though, that the most important question for the client is: “What are you trying to accomplish? What are your goals?”. So, on the 4K side, do you archive in 4K? In 1080p? Replays are already compressed – you are already compressing it, how much more are you going to need to compress it to fit your storage budget. In most cases, we aren’t talking about an uncompressed master of ‘Game of Thrones’, so you have to look at the content and what makes sense.
All of our clients have a scheme, a set of guidelines and best practices that span outside of a remote – or an entire season of remotes – so we have to make sure it fits their strategy.
They need to understand the cost – both time and money. Compressing takes time, storage takes money, etc.
Does an increasingly IP-driven environment change any of these considerations?
Scott: It does help with 4K and HDR, but IP has its own challenges. There are still bandwidth considerations, but they are significantly less of a concern. But there is a lot of complexity to IP and control overflows (signals) is still a work in progress. However, we are getting there.
Dan: Yes. One of the biggest advantages of IP is the size we can make systems. Once we can do 4K in a single 12G IP stream we will be able to make it simpler. But all connected devices will have to support this single stream 12G. It might be a few years down the road.
Marc: Yeah. At the HUB, we have a robust IP infrastructure, so the upgrade was definitely easier. But it depends on what you have, what you are building, what planning to build, whether you have a virtualized environment or physical stuff – there are a lot of levers to pull. So, is IP easier or better? It depends on what you have now.
What are (or, are there?) major bottlenecks in the production environment that throw up roadblocks to adoption of 4K? Specific workflows or gear just not quite there yet?
Rob: I would say the problem is having the platform on which to show it, both 4K and HDR – because unless you can, what is the point? It’s not the gear, it’s the end user – how can they watch it if we don’t have a platform on which to show it?
Dan: I would say, in the US, cost increases are a major bottle neck. Many broadcasters still cannot get 4K to the home. They are not willing to sacrifice anything or have a major cost increase for 4K that goes nowhere. I call this a light switch. Right now, it’s off, but once the distribution is figured out it could be 4K everywhere.
Scott: I agree – I think cost is a major factor. 4K adds a lot of cost - equipment, file size, storage costs etc. In addition to cost, there are some workflows that aren’t possible in 4K yet.
Bandwidth and infrastructure are huge issues. It just requires so much more. From an infrastructure perspective, how do you do a large show with 25 cameras in 4K with the switchers and routers we have available? For live/near live environments, we need to think about bandwidth. It is things like: how are you editing in 4K? Will you have enough time to edit things like highlight packages and shuffle things back and forth during breaks without compromises?
Also, a lot of the specialty stuff we have come to rely on is a consideration. For instance, Super Slo Mo is still VERY expensive. Can you do a football game without it? Sure. Would the viewer notice? Absolutely.
Outside of those, client deliverable is a factor too. If 4K is important to the client, that’s what they want, and that is obviously important.
Marc: Absolutely. If the client wants it, we do it.
Where are you seeing clients decide to make a switch to HDR? What types of shows/clients/projects? What seems to be the driving force?
Rob: This tends to be a pattern we see with one-off specials, for the most part – but there are of course some exceptions. Mostly we see it on specials or shows you expect to air again, or have viewers for again and again in the future. You may not have a platform currently on which to show it, but you may have in a few years’ time.
Dan: In the US, many people want to take advantage of the newer technology and have decided to go 1080P HDR. This also helps them compete with the 4K content being produced (mostly for streaming). This is a standard that is attainable without re-engineering and purchasing all new infrastructure, so it is more in reach.
Marc: I mean, I see the driving force as HDR looks better. It makes a difference. We went and shot a bunch of content in HDR. We showed it in HDR and 4K SDR on the same TVs you buy domestically, and everyone gravitated towards the HDR. If you do it properly, it looks awesome.
From a technology perspective, currently, what are the major considerations when looking at HDR versus SDR?
Rob: Most programs must be done in SDR and HDR, so monitoring and conversion are incredibly important.
Scott: Yes, I would say that the biggest consideration is what the viewer will see, how the cameras are shaded and the quality and type of conversions. Most of the world views in SDR, so it is critical that the SDR looks good; that takes work and knowledge. When we had the HD transition, the conversion to SD was fairly simple and, unless you used a low-quality converter, all SD looked better when it originated from HD. There were very few user settings.
However, in HDR, there are several settings to make SDR - and some of them are subjective and they vary with the amount of light and lighting. You have to be looking at it and have to be careful – if you aren’t looking, it could actually look worse than if you weren’t shooting in HDR at all.
Dan: Absolutely, SDR is still where 98% of what the audience is, so we have to be able to create perfect SDR out of the HDR. Onsite everything needs to be HDR and convert the outputs. It is very easy to intermix signals, this and that is not recoverable without an edit system.
Marc: I agree – it is as important to get the SDR part right – because that is what most people are viewing. We need to make sure that both SDR and HDR look good at the same time.
Like Dan said, if we are doing HDR, we are capturing everything in HDR. And if it is not native HDR we are converting it [to HDR] so everything is running through the ecosystem in HDR, so it comes out the other end in HDR. For us, a big piece is monitoring, because we are in a virtualized environment (in the HUB). Where we don’t have it, we will create it locally.
Where are we in terms of the tools that we use every day when it comes to HDR - can it span the entire ecosystem?
Rob: The largest problem with HDR is there is no flag in the SDI. You have to know what it looks like to know if you’re in the correct mode.
Marc: From our perspective, at the HUB, I think we have all of the tools we need. It’s a big game changer when you have proper monitoring and an HDR grading suite in a broadcast room. It’s more difficult in an OB truck, but doable. Rob is doing it – he has the same exact equipment that we do here, I know.
There are some pretty cool tools out there now for conversion from HDR to SDR – and I think we will see these tools mature over the next year or two. For now, you really can’t beat looking at both the SDR and HDR.
I do think graphics is an important area – they are on the screen a long time, and I don’t know that we are 100% there yet.
What are major bottlenecks in the production environment that throw up roadblocks to adoption of HDR?
Rob: Again, it is the end transmission system, here in the UK. Having somewhere to show your material in HDR. At the moment it’s only online, really.
Dan: In HDR you have to have a plan for all video that is part of the production to be converted to HDR. You cannot intermix because it’s all converted to SDR for transmission.
With 1080p to 4K, we are talking about greater resolution. More pixels. Same with SD to HD. There is a mathematical conversion. With SDR to HDR there is more of an aesthetic decision that needs to be made - which can change during the course of a production as light changes, etc. How does this affect a production technically and from a crewing perspective? Does lighting become more of an issue or concern (both natural light and stage light)?
Rob: The challenge really comes from having decent crew who are familiar with the working environment and requirements of HDR – in particular, vision engineers.
Dan: We have found that you do not have to change the HDR offsets because you follow it through iris. But like Rob said, it does take a video crew that understands how to initially set up the camera. There also is an artistic side of HDR: How do you want it to look? Some run cameras higher leaving less room for specular highlights and others run it darker leaving more room for specular highlights.
Marc: Listen, it’s all new technology. And technology always gets better. All of this is going to get there.
Luckily, we have the depth in what we are doing; Scott does too, Rob, Dan. We have more experience with HDR – and certainly loads more with 4K – than anyone else on the planet.