"4K UHD", "True 4K", "Real 4K", and "Native 4K"

[darinp]

I originally posted this on the avsforum, but I got an infraction and looks like the thread was deleted, so I figured I would post it here. Here is basically what I posted there:

We don't have universal agreement about what some of these 4K terms mean and I see quite a bit of confusion out there (including on websites other than this one) about some of these terms and how some of these projectors work. Also, it seemed that it would be good to have one location to discuss these matters instead of spreading across multiple threads. So, I figured I would start a thread where we can discuss it.

Resolution is a complicated matter in our current world with non-standard approaches to getting resolution into images that humans see, so much of where we draw lines for what we call things is opinion and not hard fact. Much of the below and probably some of what others post on this will contain opinions. Some of this will be facts, but people could argue that the threshold for definitions could be drawn one way or another way.

On this website it seems that "True 4K" and "Real 4K" are allowed for projectors that fill the screen with 4 million pixels, then fill it again with another 4 million shifted pixels, which overlap with the first set, as the DLP XPR projectors with DMDs that are 1528p do. These chips have mirrors in a 2716x1528 configuration and an eShift element to shift the pixels between sub-frames. They do not leave gaps for the 2nd set of 4 million pixels. They let the pixels from each sub-frame overlap with multiple pixels from the other sub-frame.

This is similar to JVC and Epson eShift projectors, except that those start with chips/panels with a 1920x1080 layout of pixels.

The CTA started out defining the term "4K UHD" for displays as follows for the resolution requirement:

   "Has at least eight million active pixels, with at least 3840 horizontally and at least 2160 vertically."

Others have decided to use terms like "True 4K" or "True 4K UHD" in place of "4K UHD". The CTA may be fine with those, but I have not seen an opinion from them on those or variations other than the one they did define ("4K UHD"). I didn't get any clarification on that when I asked somebody with the CTA whether they were okay with terms other than "4K UHD".

After the CTA came up with their original requirement for display resolution to qualify for the term, "4K UHD" they added a second sentence:

   "Physical pixels shall be individually addressable such that the horizontal and vertical resolution above can be demonstrated
   over the full range of colors provided by the display."

I've been exchanging emails with somebody from the CTA and have suggested to them that it would be good to clarify whether the horizontal and vertical resolutions have to be achievable on a single image or whether manufacturers can use 2 different test patterns for measuring those (as it seems TI does). I suggested that they make a choice between a 3rd sentence somewhat like one of the following:

  1: The horizontal and vertical resolution requirements must be demonstrable on a single test image.
  2: The horizontal and vertical resolution requirements can be demonstrated with separate test images.

If I hear any more I may suggest that they also draw a line in the sand for MTF, such as 25% MTF to count a resolution, or explicitly say that MTF doesn't matter.
 
TI has announced projectors with 1080p chips and 4 shift positions, but at this point I haven't heard anything about their performance, so I'm not sure if all pixels (outside the edges of the screen) will overlap with 12 other pixels, as would seem somewhat likely.

I was thinking of doing a Q and A here with my answers, but my original comments would get much longer than they will already be. So, I am going to list some questions I think are relevant in the 2nd post and give other people a crack at providing their answers.

Before I end this, I'm going to add one thing. There is one poster here who is a huge advocate for DLP XPR projectors who in my opinion will say pretty much anything to make them seem better, no matter how illogical or even if he just has to make stuff up out of thin air. I'm talking about the kind of spin I expect from marketing departments and car salesmen. I can't recall him taking responsibility for misinformation he posted even once. My view is that he just leaves his misinformation and moves on to think of more things.

People can of course like whatever technology and image characteristics they want, but I hope he (and anybody else) will keep BS and nonsense out of this thread. There are clearly many threads here where people can post those things to their hearts content. Please don't post claims here that you could have realized were false if you had just thought for another nanosecond.

--Darin

[darinp]

Here are some questions I consider relevant:

Q: Are the XPR projectors "Native 4K"?

Q: If "No" to the above, and they put up 8.3 million pixels, why aren't they "Native 4K"?

Q: Are the 8.3 million pixels with XPR projectors independent?

Q: Does qualifying as "Native 4K" require that 8.3 million pixels be displayed simultaneously?

Q: Would it change things if there was no pixel overlap? In other words, what if TI used a low fill ratio so that there was room leftover for other
sub-frame to only fill in black parts of the image on screen?

Q: Is DLP projector model xxxx for well under $20k one that uses pixel shifting to qualify as 4K UHD, or does it have 8.3 million
mirrors on the DMD(s)?

Q: Are the 8.3 million pixels from the 1528p plus 2x eShift projectors "properly aligned", as one person posted?

Q: Is it irrelevant how the pixels are projected on the screen, as long as the correct number of pixels are projected, as one
person posted?

Q: Does projector qualifying as "4K UHD" mean it will have better image quality than a projector that doesn't qualify for that high
a resolution?

Q: There are reports that some Sony projectors with 8.3 million pixels on each chip can't do a single pixel checkerboard
correctly. Are these "Native 4K"?

Others can add questions if they want.

--Darin

[AVSCraig]

Why on earth did this get you an infraction ? Too much of a possibility for arguing ?

Q: Does qualifying as "Native 4K" require that 8.3 million pixels be displayed simultaneously?

A: in my opinion, yes. The standard up until now has been the number of pixels on the chip = the native resolution. Everything else could be considered " 4K compatible " .

[Dylan Seeger]

Darin, this is an interesting topic and I suspect most people will answer these questions based upon what projector they own or what display technology they prefer. I've seen that several people can't give objective answers on the matter.

XPR, just like e-shift on Epson and JVC, is NOT native 4K. The reason is very simple. You cannot display discreet information to all 4 portions of each "new" pixel through XPR. The only way this would be possible is if XPR flashed four or more subframes to try and fix this issue, but as it currently stands, that's not how XPR works.

I see some people say that XPR is "CTA certified to be native 4K" as if that means anything. I could tell you that my speakers are THX certified, but what does that have to do with being accurate to the source material?

[darinp]

Why on earth did this get you an infraction ?

It said I insulted another member. They then deleted the whole thread, including answers to many questions from one member.

Q: Does qualifying as "Native 4K" require that 8.3 million pixels be displayed simultaneously?

A: in my opinion, yes. The standard up until now has been the number of pixels on the chip = the native resolution. Everything else could be considered " 4K compatible " .

My view is that the pixels don't need to be displayed simultaneously, much like I would consider a CRT to be native 1080p if it really could do 1080p properly, even though it takes time to paint the whole screen.

My issue with the XPR projectors and whether they qualify as "Native 4K" isn't about when the pixels are put up. It is that they overlap, which they do not in the source. So, they are not displayed correctly from a spatial standpoint.

I've referred to the "composite images" before. That is what I think really matters. Just like with single chip DLP, where they can show colors to a human viewer that a high speed camera doesn't see. As long as people aren't seeing rainbow artifacts or detecting both sub-frames separately, what matters is what the person or a low speed lightmeter sees.

If TI changed their algorithm and chips so that they had less than 50% fill ratio and made it so that none of the 8.3 million pixels overlapped with each other then I would call them "Native 4K". However, TI would never do that because it would kill the light output ratings and make native on/off CR even worse.

--Darin

[darinp]

I see some people say that XPR is "CTA certified to be native 4K" as if that means anything. I could tell you that my speakers are THX certified, but what does that have to do with being accurate to the source material?

While I can see the argument with going by what a standard's group says (especially since I am a voter for one ), in this case I think it is kind of interesting that their specification seems to have been written with TVs in mind. Maybe flat panel computer monitors too, but I think the odds are high that not a single person who worked on that standard even knew that projectors would overlap pixels. This just isn't something that TVs do. I have tried to explain how the technology works to a person in the CTA group and why I think they should clarify their standard given what projectors are now doing.

What are the odds that the CTA would have certified a projector with 1080p chips that shifts them 4 times for a lot of overlap between pixels, and not much independence in the 4K space to human vision, as "4K UHD", back in 2014, if they had known something like this might be coming? My guess is that they wouldn't have. It seems like they left a big enough crack in the definition for somebody to interpret things as this qualifying as "4K UHD".

--Darin

[ericglo]

I could see how your thread/post might be interpreted as antagonistic towards Runt even though you didn't mention him by name. Based on his history, I am sure he hit the report button more than once. Also, I am sure he doesn't want clarification on the subject.

I noticed one CRT owner decided to go digital recently. He said he went with the Optoma because it is 4k. My guess is he received most of his info from the AVS thread. It is info like that that one could correct in the past so readers could make a more informed decision. It doesn't look like that is possible any more.

[Dylan Seeger]

I could see how your thread/post might be interpreted as antagonistic towards Runt even though you didn't mention him by name. Based on his history, I am sure he hit the report button more than once. Also, I am sure he doesn't want clarification on the subject.

I noticed one CRT owner decided to go digital recently. He said he went with the Optoma because it is 4k. My guess is he received most of his info from the AVS thread. It is info like that that one could correct in the past so readers could make a more informed decision. It doesn't look like that is possible any more.

It's interesting that this CRT owner went with the Optoma. I've had the Optoma UHD65 and it's not a terrible projector, it's just not great by today's standards. It would be pretty good for casual viewing of TV and sports. I also think the UHD65 is priced well given the performance. I don't think it would function well for movies as the APL of that content usually needs more contrast to look subjectively good and the Optoma's are not up to the task, especially if you're coming from a CRT projector. However, it was very sharp with classic DLP pop, motion and a clean image.

[stanger89]

Well, I had answers to all these before the other forum deleted an interesting thread

Here are some questions I consider relevant:

Q: Are the XPR projectors "Native 4K"?

Nope.

Q: If "No" to the above, and they put up 8.3 million pixels, why aren't they "Native 4K"?

Q: Are the 8.3 million pixels with XPR projectors independent?

No they are not, and that's why they are not "Native 4K".  While electronically, 8.3 million pixels may be individually addressable, optically/physically they are not.  Just like e-Shift, XPR produces an image from two subframes, it uses an optical actuator to shift the second subframe 1/2 pixel diagonally, which means each pixel of the second subframe overlaps 4 pixels from the first.  This means the two subframes interfere with each other.  This actually multiplies the resolution, meaning XPR actually produces 17.2 Million pixels, however they are not individually addressable.

Q: Does qualifying as "Native 4K" require that 8.3 million pixels be displayed simultaneously?

Not necessarily, consider the case of a scanning laser projector.  If it were to scan 2160 lines (without overlap), and were to modulate fast enough to produce 3840 distinct sections per line, then I think we could safely call that a "Native 4K" device.

Q: Would it change things if there was no pixel overlap? In other words, what if TI used a low fill ratio so that there was room leftover for other
sub-frame to only fill in black parts of the image on screen?

Sure, I think the criteria is if you can render 3840x2160 individual pixels on screen, independent of each other, then that should count as Native 4K.

Q: Is DLP projector model xxxx for well under $20k one that uses pixel shifting to qualify as 4K UHD, or does it have 8.3 million
mirrors on the DMD(s)?

I'm pretty sure any DLP under $100k isn't native 4K.

Q: Are the 8.3 million pixels from the 1528p plus 2x eShift projectors "properly aligned", as one person posted?

Not sure what this means.

Q: Is it irrelevant how the pixels are projected on the screen, as long as the correct number of pixels are projected, as one
person posted?

It's irrelevant how the pixels are projected, so long as on screen all 3840x2160 (or 4096x2160) are independently visible and measurable.

Q: Does projector qualifying as "4K UHD" mean it will have better image quality than a projector that doesn't qualify for that high
a resolution?

As with everything, it's more complicated than that.

Q: There are reports that some Sony projectors with 8.3 million pixels on each chip can't do a single pixel checkerboard
correctly. Are these "Native 4K"?

Now here's a tough one, I want to say both yes and no. On the one hand, yes it's native 4K because it uses 4096x2160 panels, on the other if it fails to get all those to screen independently, well then it's not.  I guess I'd say it is, with an asterisk.

[stanger89]

Why on earth did this get you an infraction ? Too much of a possibility for arguing ?

Q: Does qualifying as "Native 4K" require that 8.3 million pixels be displayed simultaneously?

A: in my opinion, yes. The standard up until now has been the number of pixels on the chip = the native resolution. Everything else could be considered " 4K compatible " .

Like we said above, consider the case of a scanning laser projector (there are a couple out there), arguably it only projects one pixel at a time, but if it were to scan 2160 lines and modulate fast enough to resolve 3840 pixels per line, I think it would qualify. 

Although at the same time, I agree that since such a projector can produce many resolutions, it doesn't have a "native" resolution.  Arguably you could say that that's "True" 4K, where as "Native" or not refers specifically to the imaging device used, and if it has a physical resolution of 3840x2160 (or 4096x2160).

[stanger89]

I could see how your thread/post might be interpreted as antagonistic towards Runt even though you didn't mention him by name. Based on his history, I am sure he hit the report button more than once. Also, I am sure he doesn't want clarification on the subject.

I will give him a small amount of credit, if he were the type to report posts, I'm fairly certain I'd be banned already, considering how many times I've confronted him.

OT - is there a way to multiquote with this forum software?

[AVSMike]

I do not consider any projector that uses 1080P panels a native 4K projector. Does not matter how many pixels are put on screen, when you have to shift and overlap them. You could take 720P panels and shift them multiple times to get the CTA definition, but that certainly would not be 4K.

These new faux k DLP's would be nice projectors, if they had decent contrast, even though they are not true 4K.

[ericglo]

I will give him a small amount of credit, if he were the type to report posts, I'm fairly certain I'd be banned already, considering how many times I've confronted him.

OT - is there a way to multiquote with this forum software?

You may be correct, but I have only had posts removed and infractions incurred on those DLP threads.

I agree with the individually addressable pixels independent of each other would be my criteria.

[stanger89]

You may be correct, but I have only had posts removed and infractions incurred on those DLP threads.

Interesting, maybe I'm just diplomatic enough 

[darinp]

Not sure what this means.

This was posted by longhornsk57 in one of the UHZ65 threads:

If you get 8 million unique pixels aligned properly at 60hz no matter how it's accomplished it's 4K.

I'm not sure if he actually thinks they are properly aligned with the XPR method, but that is my guess. I would respond there, but I was banned from that thread for correcting misinformation.

Now here's a tough one, I want to say both yes and no. On the one hand, yes it's native 4K because it uses 4096x2160 panels, on the other if it fails to get all those to screen independently, well then it's not.  I guess I'd say it is, with an asterisk.

I agree. I asked this because it is a tough one. Maybe 4K*, like you said. Basically, the pixel level is native 4K, but there may be a bottleneck elsewhere in the chain that can't actually support that much properly. If it is just a matter of going into the service menu and disabling pixel alignment then I would call it native 4K.

I agree with your other answers.

--Darin