4K Monitors...

[Bill Anderson]

Thanks Bruce. I understand the ability of the larger space to comfortably encompass and represent the smaller space, but I thought the BITS were too spread out (wasted) on a larger than necessary space, resulted in potential for banding and such?

 

[Marden Blake]

Sony now talking about their x65 4K camera and 4k theater and 4k projector at CES.

OOOH 4K TV

 

[Bruce Allen]

Thanks Bruce. I understand the ability of the larger space to comfortably encompass and represent the smaller space, but I thought the BITS were too spread out (wasted) on a larger than necessary space, resulted in potential for banding and such?

SHV color standard is wide gamut but 12-bit so that is not a problem

Bruce Allen
www.boacinema.com

 

[Blair S. Paulsen]

4K is how far away?

4K is how far away?

Whatever timetable for the widespread adoption of 4K you predict, that's really just a question of time. Yes, I understand that impacts the viability of the technology in myriad ways, but several of the issues Bruce cites need to be addressed sooner than later. Sure would be nice to see an provision for including a 3D LUT in the file header that can be imported into the display device, perhaps as a basis for automated calibration routines... At its most basic level it would provide a protocol that could read color space flags like rec709 and implement the correct adjustments for that particular panel.

Disclaimer: I would really like to develop a technology for creating personalized LUTs mapped to the visual response curve of the viewer and this would make implementation much easier ;-)

Creating a standard that strikes an elegant balance between quality and the mechanics of creation/delivery in a rapidly evolving technological time is a serious challenge. In just the last few pages of this thread there were references to bit rates and delivery systems in the context of a rather wide variety of usages - each with its own market dynamics. There was general acknowledgment of the greater viability of open standards (or very low cost licensing) combined with mass market economic drivers. At the very least I/O standards need to be in place that are not silly expensive to implement and have massive headroom for higher speed variants down the road.

As an example, the first consumer targeted 4K projector I know of is the Sony ES-1000 and it is fitted with a single HDMI 1.4a port as the only way to feed it 4K. AFAIK this limits it to 4,096 by 2,160 at 24fps (23.976) and 3,840 by 2,160 (QHD) at 30fps (29.97). I understand the desire for a single cable solution and doubt the clock rate of the projector could do much more anyway, but how would you feel if you paid $25K for a projector that could never even do full res 4K S3D at 24 fps? In the past, some manufacturers of next generation hardware have used outboard processing boxes to feed display devices over proprietary connections as a stop gap until common interconnect formats catch up.

I call for a connection standard capable of playing back 4K S3D at 60fps. That should hold us a good long time. Quad link 3G SDI? HDMI multi-line snake with a hydra head?

FWIW I am not convinced that 8K is really going to catch on due to the physical characteristics of the wetware (human vision/size). I believe that even when prices drop, very few people would want a home screen with a diagonal of greater than 120" (slightly more than 3 meters). Even in a dedicated home theater, if you mounted it right up to the ceiling in most houses the bottom of the screen would be too low for comfortable viewing. So unless you fill an entire wall and sit closer than 5' the difference between 4K and 8K would be hard to discern. This is an analog to the argument about 4K vs 1080 on a 50" panel from 8' away. The big difference is, that panels will go beyond 50" and display devices live in a lot of places these days besides a traditional living room seating arrangement. In the theater, due to the limited number of arc degrees of high visual acuity we possess, you can only get so close before you can't keep the edges of the frame in your visual field. I wouldn't argue the point that a well crafted viewing environment with next generation display tech could show the difference between 4K and 8K, what I am saying is that in the real world issues unrelated to technology per se, but simply the reality of our physical bodies and the physical spaces we consume media in, will make that differential too small to justify much in the way of greater expense. Bottom line:

1) Theatrical: with the buzz about 3D already fading, 4K will be touted as a reason to go to the theater vs staying home so the trend of features being DIed at 4K will take off and, down the road, 8K for theatrical might serve a similar purpose.

2) Home: as display tech evolves, the difference between 1080 and 4K will be significant enough to impress the mass market and as soon as the price point hits the tipping point it will catch on.

As Mike noted some standards live very long lives (for a variety of sometimes unexpected reasons). Let's look at choke points on the delivery side that might motivate compromising best quality for expediency. What number of desirable (demographically speaking) households will be able to receive OTT (IP based distro, not MSO/"broadcast") minimum download speeds of 25mbs within 5 years? If that number is too low, then we run the risk that in a rush to monetize a technology too soon, it is compromised so much it misses the wheelhouse of its potential.

I would hate to see us repeat the sorry state of most digital music distribution these days where bit starved implementations of codecs never designed for those rates yield proxy level quality that is sold as "pristine, digital, ...".

Cheers - #19

 

[Petr Dvorak]

Sony now talking about their x65 4K camera and 4k theater and 4k projector at CES.

OOOH 4K TV

vaporware :biggrin:

... and they even announced 4K upscaling blu-ray player

 

[Tom Lowe]

I remember when the folks at cinematography.com told us that 4K screens for the home were at least 10 years away... and were "unnecessary."

 

[Subhadip Sen]

I have lost count on the number of 1080p/1200p 10 inch tablets and QHD/4K monitors I have read about at CES 2012. Between these we even have near-3K ultrabook laptops. Panasonic is flaunting a 20.4" 3840x2160 IPS-Pro panel monitor that is apparently just 3.5mm thick. Looking forward to a 3840x2160, 24 inch, sub-$1000 IPS display sometime in 2012. Although considering my viewing distance 3200x1800 will be adequate.

 

[Fergus Meiklejohn]

There are so many 4K monitors (TVs) being released that we can hardly keep track. I wonder what that means to shows shot on 1080P?

Jim

I think that the challenge for 4K believers is persuading someone with room for no more than a 42" TV that the picture quality is dramatically better than 1080p. So while the producers of big budget shows will have an eye on future revenues and will doubtless start to prefer to master in 4K, I don't yet see that if a consumer only has a 42" screen (4K or 1080p), that an old 1080p TV show won't scale and look just fine. It's certainly not going to be a deal breaker is it?

 

[Marc Wielage]

Either have two separate encodes... or have a 709 encode... and some compressed "wide-gamut" data per pixel that modifies the values if wide gamut is turned on. Why not just use NHK's SHV color space - it matches laser primaries well - and then all of those 4K masters and 4K RedRays will look good on those future 8K TVs too.

Interestingly, the Sony F65 has an available color gamut that appears to be identical to SHV color space:

I have no idea how easy or how hard it'll be to reproduce this color space in the real world, but it is interesting that Sony built this in to the new camera.

That's the kicker. Youtube is 4K now -- it sucks, but it's 4K.

Yeah... but it's incredibly compressed 4K. 4K that stutters, has bad motion problems, and looks real blocky is not good 4K. Throughput is a real killer, and let's also not forget about bandwidth caps.

I remember when the folks at cinematography.com told us that 4K screens for the home were at least 10 years away... and were "unnecessary."

I have never seen the DPs and members of the CML agree on anything or make a unified statement. One or two people might have said that, but not the group as a whole. If anything, they disagree more often than they agree, especially on stuff like 1080 vs. 4K.

The issues really boil down to moving this much data back and forth. I believe the post articles on Girl with the Dragon Tattoo indicate that the final 2-1/2 hour movie boiled down to 55TB of data. You know how long it takes just to copy 55TB of data and back it up, even on very, very fast drives? Call up your cable provider and ask them what they would do if you tried to download 55TB of data in a few days. Even if you could compress this 10:1 to 5TB of data, even that takes hours and hours to do. Last time I checked, I think it takes about 6 hours to copy 1TB of data on eSATA -- somewhat faster via FiberChannel and a few other connections.

All of this stuff is gonna happen someday, but me, I'd like to see HD perfected on consumer monitors before we move on to 4K.

 

[Hrvoje Simic]

Panasonic's 3840 x 2160 20" LCD is not designed for multimedia production, but for specialized purposes, such as medical and industrial.
Which means priced accordingly.

 

[Daniel Kunz]

Yeah... but it's incredibly compressed 4K. 4K that stutters, has bad motion problems, and looks real blocky is not good 4K. Throughput is a real killer, and let's also not forget about bandwidth caps.

I'ts not youtube's 4k that stutters or has bad motion problems, it's your computer choking on the flash-player.
If you download the MP4 stream (which is mostly 2.5K and not 4K) you can play it smoothly in another player like VLC or such.

The thing i have discerned the most is the spacing between keyframes that sucks bad.
from what i have been able to calculate the bitrate is between 5x and 6x higher than the 1080p stream which is a good thing since the difference in resolution is 4.5x more pixels.

For Google it would be a flick of a switch to make 4K Bitrate on youtube 10x or 20x. More buffering and waiting, but amazing quality for free. no proprietary hardware, no new cable-company, same ol', same ol'

 

[Wayne Morellini]

Well I need an advice!

We are establishing the first grading room in Israel that will have a projector.

With all the developments and debates over 2K vs 4k, and bearing in mind our provincial budgets and the very trendy film industry around.

Which projector would you choose.

The projection room is going to be 6 meter wide and 9 meter long.

Thanks.

Red Projector if it turns out good enough Ido. Equator speakers. Keep in mind the wide color gamut issues with the Red projector. You might be able to make great masters ti show but they won't show on normal projectors.

It really is too early to tell unless you want to spend lots of money. But post a thread to a matching forum, where people are more concerned with the issue, and can give you a better professional opinion than mine,

 

[Wayne Morellini]

Increased resolution is inevitable, this was obvious (or I would think it was) way before HD even got off the ground. I'm not sure I understand the attitudes and mindset that resist. All the current arguments against 4K are the same tired and crap arguments we saw against HD in the 90s.

4K is still just a stepping stone, but it's a monumental one. It's one where we are matching, and in many cases surpassing, the level of detail presented by 35mm cinematic film. I'm sure we'll have the same arguments all over again when 8K delivery arrives or 12K or whatever the next major jump happens to be. Resistance to 4K is relatively light compared to what we faced with HD and I'm betting to move beyond 4K will involve a bigger fight... Just a hunch.

With all that, 4K or QHD displays are EVERYWHERE at CES. Looks like it's the theme for this year's show and none of the technical rags have figured that out yet. I'm not at CES this year, but my ninja spy network is feeding me quite a bit already, even before the exhibits open.

One thing to watch out for with the first crop of "4K" or "QuadHD" displays this year and it's a real gotcha... Not all of them accept 4K or QHD input! A good portion of them are only using all the extra pixels for passive 3D use. Many (or most) of the 4K displays do extra internal processing to up-scale and further "enhance" the image. So if you don't think your 1080p is over-processed and over-sharpened enough, then these guys have a solution for you!

Seems like LG, Toshiba and Sharp are the front-runners of the 4K pack. Lots of talk over the Toshiba 55" glasses-free TV, but I'm just not finding a way to be interested in that one. I just want a 54" to 60" with good blacks and good color range that I can drive at QHD or 4K via a RED Rocket or upcoming RED RAY player.

+1 .

I have lost count on the number of 1080p/1200p 10 inch tablets and QHD/4K monitors I have read about at CES 2012. Between these we even have near-3K ultrabook laptops. Panasonic is flaunting a 20.4" 3840x2160 IPS-Pro panel monitor that is apparently just 3.5mm thick. Looking forward to a 3840x2160, 24 inch, sub-$1000 IPS display sometime in 2012. Although considering my viewing distance 3200x1800 will be adequate.

Hmm, at last, but what about 30 inch +?

Man the editing windows on red user are playing up tonight.

Lost some posts. But I think Jeff was posting about GPU's. I was going to say you are looking at it the wrong way. You would have the solution locked down on the system software and driver side. License Open cl 1.1 and dx 11 level imagination technologies or other GPUs, or even Amd, intel or Nvidia part lots. Forget driver updates, embedded Linux etc, locked down. If parts are no longer available, or you run out in two years, get new parts reverify the correct working, making adjustments as needed, and hold on for another few years. The fact now is that you guys probably are well along in designs, and conversion is a problem. If you are meaning on general hardware out there that is another problem. The reality is that you maybe able to make a verifiable system on Open CL, despite the problems you mentioned. What game companies used to do is make Code versions for each major platform difference, and falkback compatbility code. That is too much to do, but there should be a simpler way to do it. You have known points of variation, such as rounding, and performance points. With the variations you write test routines to test the accuracy, results and performance of various basic functions of your system, and accommodate as you are able, verifying enough performance left over. The performance points are just conventional, just isolate sections of code performance targets and verify that they can be reached, and time it's use so not to overload the system etc etc etc you know the usual.

Sorry if this is not a very good response, very late and I'm not all with it.

 

[Paul Sage]

The issues really boil down to moving this much data back and forth. I believe the post articles on Girl with the Dragon Tattoo indicate that the final 2-1/2 hour movie boiled down to 55TB of data. You know how long it takes just to copy 55TB of data and back it up, even on very, very fast drives? Call up your cable provider and ask them what they would do if you tried to download 55TB of data in a few days. Even if you could compress this 10:1 to 5TB of data, even that takes hours and hours to do. Last time I checked, I think it takes about 6 hours to copy 1TB of data on eSATA -- somewhat faster via FiberChannel and a few other connections.

Michael stated that the DI was 55TB, not the deliverables (DCP, HD Video, etc...).

 

[Patrick Grossien]

JVC 4K projector for $12K
http://pro.jvc.com/prof/attributes/list_price.jsp?model_id=MDL102123&feature_id=00

things are rolling out...

 

[Patrick Grossien]

The revolution now has hit the mainstream!

 

[Jarek Zabczynski]

JVC 4K projector for $12K
http://pro.jvc.com/prof/attributes/list_price.jsp?model_id=MDL102123&feature_id=00

things are rolling out...

Eh...looks like a fake 4k via magic pixel shifing....says it has three 1080p chips with "e-shift"

 

[Stuart English]

have never seen the DPs and members of the CML agree on anything or make a unified statement. One or two people might have said that, but not the group as a whole. If anything, they disagree more often than they agree, especially on stuff like 1080 vs. 4K.

The issues really boil down to moving this much data back and forth. I believe the post articles on Girl with the Dragon Tattoo indicate that the final 2-1/2 hour movie boiled down to 55TB of data. You know how long it takes just to copy 55TB of data and back it up, even on very, very fast drives? Call up your cable provider and ask them what they would do if you tried to download 55TB of data in a few days. Even if you could compress this 10:1 to 5TB of data, even that takes hours and hours to do. Last time I checked, I think it takes about 6 hours to copy 1TB of data on eSATA -- somewhat faster via FiberChannel and a few other connections.

All of this stuff is gonna happen someday, but me, I'd like to see HD perfected on consumer monitors before we move on to 4K.

If your above comment is meant to encompass distribution, then we have already demonstrated that such 4K DPX master data can be compressed to approximately 20 Mbit/sec.

 

[Bruce Allen]

Interestingly, the Sony F65 has an available color gamut that appears to be identical to SHV color space:

I have no idea how easy or how hard it'll be to reproduce this color space in the real world, but it is interesting that Sony built this in to the new camera.

You can reproduce something extremely close to SHV color space right now, with a laser-based Mitsubishi DiamondVue

In fact, they were doing it in 2008!
http://hdguru.com/mitsubishi-laservue-l65-a90-first-tech-review-hd-guru-exclusive/310/

Look at the gamut triangle lower down in the article...

OR here are the measured numbers - pretty close:
Brilliant Red: x.7159 y.2833
Brilliant Green: x.1725 y.7943
Brilliant Blue: x.1622 y.0129

I imagine that other laser-based projectors would be similar.

And OLED could probably come close too.

Bruce Allen
www.boacinema.com

 

[Jeff Kilgroe]

Eh...looks like a fake 4k via magic pixel shifing....says it has three 1080p chips with "e-shift"

*ALL* ofl the "4K" or "QHD" projectors based on LCOS technology are using pixel-shifting.

*MOST* of the DLP-based "4K" or "QHD" projectors are using "wobulation" where the DMD block is wobbled side to side while scanning the image to effect more pixels than it actually has.

There is a 3840x2160 DMD from TI, however it costs nearly 5X as much as the wobulated version. I'm not seeing any reports of a full-rez (not wobulated) DMD/DLP projector for under $35K. In fact, they may be more expensive than that.

A couple 4K laser projectors are being announced. I don't believe anyone is actually showing functional product on the show floor, have not been able to verify.