There was a lot of skepticism and debate in the digital signage community when 4K commercial displays started coming on the market, with industry observers openly wondering if visual messaging applications needed that high level of resolution.
Several years later, 4K is perhaps not common, but certainly being used in many projects, and both accepted and supported.
And as 4K bedded in, the industry started seeing some of the bigger display manufacturers showing 8K displays at trade shows, and the debate about the demand and the challenges for super high resolution displays started all over again.
One of the ways an industry builds awareness, acceptance, support and standards for a new technology is to have working groups or organizations of stakeholders. There's an 8K Association now, and the companies that got it going asked display industry veteran Chis Chinnock to step in and run it. An industry observer, writer, analyst and consultant, Chris has been around displays forever and seen the evolution. He understands what the engineers are going on about, but has the skills to explain it in terms mere mortals can understand.
He explained to me where 8K is at on the adoption curve (it's still early) and we get into the implications of 8K on infrastructure. He also explain who will want and use 8K, and why.
Hey, Chris. Thanks for joining me. Can you tell me what the 8K Association is all about?
Chris Chinnock: Sure. Thanks for having me on your podcast, appreciate that.
The 8K Association was formed at CES in 2019, so about a little over two years ago, and it was formed with panel makers and TV makers primarily and that's when 8K TVs were starting to come into the market and we had some initial goals which was to promote those 8K TVs, to develop a certification program for those 8K TVs, to begin gathering 8K content for our member use, and to begin education of the professional community, because we're going to need a full 8K ecosystem from content creation, through distribution and display for this to become mainstream.
Is this the sort of thing that wouldn’t necessarily just happen organically?
Chris Chinnock: It would happen organically. We just wanted to form the organization to try and help facilitate communication and maybe move the ecosystem a little bit faster than it would've done organically. That's all.
And is this something that manufacturers do? I think of some of the other certifications and reference designs out there?
Chris Chinnock: Oh, yeah and there's tons of these organizations out there with different marketing goals or ecosystem development goals, so we're not reading any new ground here. We're, this is a tried and true kind of approach.
For example, in the 4K transition, the UHD Alliance came up to do a promotion and development of UHD or 4K content, mostly aimed at consumers and then the UHD Forum was originated not long afterwards, which was focused more on trying to educate and develop the 4K ecosystem and the professional community and they developed a bunch of guidelines and whatnot to help broadcasters and filmmakers implement 4K workflows. In many senses, we're following that model and learning from what they did and trying to leverage that going forward.
Yeah. It's an interesting thing. You have lots of people saying, “8K: is that something we're ever really going to need?” “There's no content for our…” blah, blah, blah, blah, blah. These are all the things that were said about 4K and lots of questions as to whether 4K whatever take route at all and it certainly has, is it the same argument or is this a little more nuanced because 8K is like super duper high resolution?
Chris Chinnock: It is the same argument. We had naysayers six, seven years ago for 4K. We've got naysayers for 8K now. Absolutely, it's a different environment now, but there's also a lot of things that are similar to what was happening in 4K six, seven, eight years for sure.
With 8K, you're talking about super high resolution. In the context of digital signage, where would 8K be particularly useful and applicable?
Chris Chinnock: From what I'm gathering, we've actually been poking around in the proAV space, trying to understand what the needs are for 8K, to tell you the truth and what we're learning is the big need is really in distribution and transport. So the canvases are clearly getting bigger and larger in digital signage. An 8K digital signage is not uncommon, I don't think nowadays. But it's not necessarily in a standard 16:9 format. They come in all kinds of aspects, ratios and configurations.
But what we're starting to see is, these big canvases, you want to start with a higher resolution source of master file, so that you can have a piece of that 8K master going out to various parts of this display. So if you letterbox it or clip a PC in there, you want to start with a high resolution piece and not have to do upscaling at the display itself, if you can avoid that, because there are some issues with that.
So the main argument is interesting, with the 8K Association and the website, you even have on the navigation bar, just straight out, “Why 8K” and I go through things and some of those objections, so to speak are: you can't see the resolution, that the human eye can't even raise, can't even resolve 8K now.
I don't think that's quite accurate, is it?
Chris Chinnock: No, it's not and people make that argument based on simple acuity, that is the Snellen chart and it's literally based on geometry and that is a big part of vision. There's no doubt about that. But human vision's far more complex. There are higher order things going on, they call it hyper acuity, so that allows you to see, for example, the differences between parallel lines and slightly unparalleled lines. It allows you to see stars in the night sky that simple acuity says you can't see and perhaps more importantly, we form images in the brain, the retina and the eye is part of it, but the brain puts those signals together to create an image. So we sometimes and often do fill in details in our brain to create that image of the world.
So if you have an 8K image versus a 4K image, it has less artifacts, it has more texture and detail. So it creates a higher sense of realism. It's subtle and the hyper acuity may say you can't see that difference, but all these other factors reinforce that it's more real, it's more present.
Do you have to be within a certain kind of viewing cone or proximity in order to appreciate that difference between 4K and 8K?
Chris Chinnock: Yeah, certainly the closer you sit, you will see more detail and sharpness and texture, and that's for sure, because that's part of the simple acuity part. But also remember, we're talking HDR signals for the most part with 8K content now. So it's high dynamic range, it's white color gamut. All these things make a big difference.
Yeah, if you're using high dynamic range, then you can see an incredible amount of detail that isn't otherwise revealed.
Chris Chinnock: Exactly.
From your point of view, are there certain kinds of applications for signage that make more sense than others?
Like from my point of view if it's at a museum or something where you're going to get people who are going to be walking up close, that's when it starts to make some sense.
I just don't know that anybody's ever going to need a 8K digital menu board in a QSR.
Chris Chinnock: I agree with that, absolutely and museums are one perfect example. I know I've seen in some trade show demonstrations, they'll have an artifact that can be either a video capture, a 3D video capture of an artifact, or it could be a very high resolution computer generated image.
But now you can go up to the screen, you can really look at this artifact. You can zoom in on it. You can rotate it and you don't see any discrepancies in that image. You've got lots and lots of resolution to play with here, so it's much closer to lifelike
In terms of math, what's the difference? And I hope I'm not putting you on the spot here, but just generally the difference between a full HD file and an 8K file, in terms of size and what are the implications in terms of the equipment, graphics cards and everything you need to play it out.
Chris Chinnock: File size is going to depend on the compression that you use. Maybe a better way to look at it is what's the bandwidth you would need, the uncompressed bandwidth you need for various files.
So I think full HD is somewhere around three gigabits as I recall. But now if you want to do 8K, I think the highest reasonable level that you want to do is 8K at 60 frames per second, 10 bit and now the difference comes with color subsampling. If it's video, you're going to do four to zero color subsampling, that's about 30 gigabits per second. So ten times full HD, right? If you want to do broadcast quality, that's four to two color sampling. That's 40 gigabits. You want to do four-four,-four for high resolution graphics in proAV, 60 gigabits per second.
Chris Chinnock: Yeah, so here's the problem: it's a distribution challenge, right? So there are solutions out there. If you want to do proAV, you can use HDMI 2.1. You may have to use two connectors if you want to do four-four,-four. That's a real challenge, just to sync those and it's going to be short distance, right? So your player's gotta be right by, probably a standalone 8K display of some sort.
The other side of the coin is IP distribution, right? That's a huge trend in the whole proAV space. So there's a lot of solutions that are out there to do that now. A lot of them are focused on one gigabit networks and that's just not gonna cut it for 8K, right? So we're starting to see, there’s at least two organizations that I know about that are trying to develop some standards in this. One is the Software Defined Video Over Ethernet (SDVOE) and that's focused on using a 10 gigabit network to support it.
10 gigabits is okay if you use some kind of a mezzanine codec, like JPEG XS. That's supposed to be a lossless codec that you can compress up to fifteen to one. So you can get all those signals onto a 10 gigabit network easily with JPEG XS and then the other organizations I'm aware of is the AIMS Alliance organization, and they're developing what's called the IPMX standard. What they're doing is borrowing video over IP standard from the SMT organization from broadcast and that standard is called ST2110 and it has all kinds of high-end features in it for broadcast, including redundant distribution. So you have two Ethernet channels So if one fails the other one's always there. It's got high-end timing and grand master clocks.
We don't need that for proAV for the most part. So the AIMS Alliance is specifically trying to take that broadcast standard, strip out what you don’t need, add in some things that you do need for proAV and develop a new standard.
So if I'm somebody who runs a facility operations for a Fortune 500 company and at their main headquarters office campus, the CEO has bought an 8K TV for his home and says, “I love 8K. I want my whole digital signage network converted over to 8Kx8K displays. It can replace the 4K's or the 2Ks that are hanging up on mounts and all that.”
What are the infrastructure implications if you want to be moving files around on the wide area network and everything else? I suspect you're thinking about even the cabling, certainly have a lot of the hardware that's moving data around everything else.
Chris Chinnock: Absolutely and that's why you have to have a network that can support this. If your corporate network is based on a one gig ethernet structure, you're going to have to upgrade that. I know some of the new Intel motherboards support both 1 gig and now 2.5 gig ethernet outputs over, I think that's probably over Cat 5 cables and that may be sufficient if you're using a video and can compress that enough to get on a 2.5 gig network. It's pushing it a little bit but it's possible. It, again, depends on the client here, if they're really sensitive to having pixel accurate images or if it just has to look pretty good, I should say really good, but...
If you want to go to a 10 gig network or a 5 gig network, these are all much better because you can use less compression. But they come at a cost: all your network switches and maybe the cabling have to be upgraded to support this as well.
Yeah.So like a regular Cat 5 may not support it and then you're pulling hundreds of meters of new cabling?
Chris Chinnock: Potentially. Yeah, absolutely. Or, maybe just go to a fiber network to be future-proofing who knows?
Wow and so I would imagine some new builds are future-proof like crazy, but there would be probably 90% of the built environment out there would probably need to be tweaked in some way, right?
Chris Chinnock: Yeah, I think so.
One of the other arguments about 8K is that there's no content available. Is that true?
Chris Chinnock: It is true. There is very limited content out there. The interesting part of that is that actually a lot of content is shot on 8K cameras and there are now 12K cameras out there. Black magic design is a 12K camera. So it's being captured in 8K or higher, but it's not being finished in 8K or distributed in 8K yet.
Is that because there's no market for it?
Chris Chinnock: Yeah, partly. You gotta have a certain critical mass of 8K TVs out there before you start streaming to it and I think streaming is going to be the first way that we see 8K content coming to consumers and you need good codecs out there to distribute it too.
NHK has been broadcasting 8K content for over two years now but they're broadcasting at 80 to a 100 megabits per second with high compression ratios, and that's just too high. Netflix is 15 to maybe 25 megabits per second, that's where most of the streamers are coming in right now for 4K content. That's where you need to get, maybe you could start at 40 or 50 for a premium 8K streaming service, but you quickly got to get down to that 25 area, I think in my opinion.
Is there a likelihood that there will be more content produced that is in the right format? What changes that?
Chris Chinnock: Yeah, I think again, you need that critical mass of TVs out there. You need a cost-effective distribution system and when that arrives, especially with these new codecs that are coming like VDC, I think you'll see major streamers jumping in with an 8K service.
Another argument is that production costs are also high. Is that primarily the costs of cameras that are like Black Magic Design and RED cameras that can shoot in that? Or, are there a whole bunch of things?
Chris Chinnock: Yeah, there's a bunch of things there and those were exactly the same arguments for 4K adoption, six, seven years ago. It's more memory, it's more bandwidth. The camera costs are a little bit higher. The storage costs are a little bit higher. That's all true but I think we're also in a very different era now. So with this pandemic, we've seen a big acceleration of production workflows to the cloud. There's no doubt about that, with all the remote production that had to go on. We're also, I think, going to see an evolution of that. So more and more production will go to the cloud and I think that actually favors 8K production as well, because what we're seeing and a company, FrameIO just demonstrating this, they have a camera to cloud service now. So you can be on set shooting with 8K cameras and as soon as you finish that take, it goes right up to the cloud. The original camera files are in the cloud and then from the cloud, you can do proxy editing proxy color grading. You can do everything and have dailies right back on the set the next morning.
This is going to revolutionize, I think, the way movies and TV shows are produced,
When you’re talking about compression. I think in terms of compression somewhat clobbering the file, does it have any noticeable impact on quality as opposed to the native file?
Chris Chinnock: Sure and that just depends on the compression ratio.
JPEG XS up to 15:1, that's supposed to be a lossless thing. So visually lossless, if you're at that kind of compression ratio, but if you get into a distribution, that's called a Mezzanine Kodak. If you get into a distribution Kodak, one goes to consumers, Amazon or Netflix is using, HEVC and is going to be hundreds to one in compression and so you can potentially see our artifacts that way.
Especially now, when you put us on a very large screen, that's tens of meters wide, you'll definitely see things on that size screen that you wouldn't see on an 80-inch screen.
Is 8K best suited for flat panel displays as opposed to LED?
Chris Chinnock: Not necessarily. With LED again, because it's a bigger screen, it's less forgiving, because any artifact is just more visible.
What about the timelines on all this? You mentioned how six, seven years ago that the fuss that was out there was around 4K and nobody's ever going to use it or anything else… What do you see are the timelines to a time when 8K is a shoulder shrug?
Chris Chinnock: Well, there's a graphic that we one of the market research companies put together that showed that the resolution transitions and now we're talking about displays here. So when a display of a new resolution was introduced to the time it was selling at 50% of retail sales.
So SD to Full HD, Full HD to 4K, and now 4K to 8K, that cycle is seven years and consistently seven years. So you could argue that we're a year, maybe two years into the 8K cycle at this point. The pandemic probably added a year to that. So if in seven years 50% of sales are now 8K TVs, we saw how fast 4K TVs were adopted and how fast 4K streaming came onboard.
Will history repeat? Probably.
Is this primarily a consumer driven product or do you envision a lot of commercial adoption of 8K displays?
Chris Chinnock: I think the push today and certainly the focus of the 8K Association has been on consumer entertainment, production and the entertainment production value chain. But as we have already discussed, there are clear needs in the proAV space here as well, particularly for all these larger canvases for rental and staging, for corporate environments, for pop-up events. We talked about museums, there's medical imaging that's an important area as well, that's coming on board.
How about 8K VR? That's starting to happen as well. So there are a range of different applications here, including broadcast as well. It's happening in broadcast too. So yeah, it's happening in a lot of different areas, even security cameras. There's 8K security cameras now.
Is some of that just a function of end users and integrators and everybody else wanting to have the latest and greatest and say, “We're doing 8K, we're doing AI, we're doing machine learning.”? Or, they're just jumping on the latest?
Chris Chinnock: Yeah, and I think, that's what companies do, and they always have to make the next product a little bit better than the previous one. So 8K is a natural next step and I'm glad you brought up this idea of AI as well, because that's also very different from where we were during the 4K transition.
Upscaling in the 8K TV, AI based or machine learning based and neural network based now is a completely different technology from when we had upscaling and 4K TVs. AI is being used in the encoding space as well to help reduce those bit rates and do seen optimized encoding now. So we're just at the very beginning of this AI capability and the cloud capability. So the combination of these I think is going to be very powerful.
If you're using display technology that has AI based upscaling and it's that good, do you really even need to produce at native 8K or are you in a lot of situations and are going to be just fine with 4K upscaled?
Chris Chinnock: Yeah, today that's a very acceptable solution. In fact, it has to be the solution because that's what we have out there.
But one of the scenarios that we're trying to standardize in the 8K Association is, we don't really have a good name for this yet, let's call it Smart Streaming just for purposes. But the idea here is that you encourage those 8K camera original files to now be mastered in 8K, so you create a naked finished movie or piece of content.
You now smartly downscale that to 4K with some metadata about how you did that. You now use conventional encoding techniques, HEVC, or AV1 to distribute that content the way streamers are doing that now and if you have an 8K TV, you can now read this metadata smartly, upscale that back to 8K and theoretically, that gets pretty close to a native 8K distribution scheme in terms of image quality.
If we're talking about four to five to six years before 8K is really settled in, companies that are thinking right now about a refresh of their display technology and the supporting infrastructure for the digital signage network they're running across whatever environment it may be in, do they need to be future-proof now?
Or they're fine with what they have at the moment and they can just have in their heads that the next refresh cycle, we'll be looking at 8K?
Chris Chinnock: I think it depends on the number of pixels they want to put up there. If it's a big canvas with a fairly tight pixel pitch, which means it's a lot of pixels then absolutely, I would be thinking of a higher network structure for that. If it's a smaller display with a bigger pixel pitch, then maybe you don't necessarily need it.
And we're talking 8K, but I have seen the stories for 12K and I believe even 16K. Are those things that will exist beyond labs?
Chris Chinnock: I would not underestimate that capability, yes. I believe that will probably happen and we'll go through this whole cycle again. “No one could see it. We don't need it. It's too expensive.”
I read something saying 16K is pretty much perfection, that's like 20/20 equivalent, perfect eyesight. You're seeing everything!
Chris Chinnock: That depends on the screen size.
Ah, okay. This is all over my head.
Chris, how did you get involved with the association?
Chris Chinnock: I was actually asked by the TV manufacturers to help form it. So I said, yes.
And is this a full-time gig or among the many things you do?
Chris Chinnock: No, it's one of several things that I do, yeah.
Just for the listeners, what kind of work do you do?
Chris Chinnock: I've been in the display industry for close to a thousand years now. (Laughter)
I've done all kinds of things. We published newsletters for quite a while. We did market research reports, consulting services. I've run a bunch of events on this. I do white papers for clients, et cetera, et cetera. So my focus has always been on the cutting edge of display related technology. So 8K is one of them.
I'm also very active in the light field and holographic displays, AR and VR are key areas for me and micro-LED and mini-LED is also an exciting technology for me as well.
Yeah, and the fun thing is everything you're looking at is ever evolving and always interesting.
Chris Chinnock: That's why this job is fun. I learn something every day, right?
Yeah. That's the same with me. People ask me, I'm in my early sixties now and people are asking me, “Are you thinking of winding down?” And I say, not really because the stuff I work on is interesting and I follow the stuff that's emerging and that's always fun.
Chris Chinnock: Exactly. I'm with you.
All right, Chris, I appreciate you taking some time with me.
Chris Chinnock: I appreciate the time. Thanks, Dave.