Optoma UHD60 & UHD65 4K/UHD HDR DLP Projectors Now Available - Page 15 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1730Likes
Reply
 
Thread Tools
post #421 of 2011 Old 06-14-2017, 10:29 PM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
TheronB is offline  
Sponsored Links
Advertisement
 
post #422 of 2011 Old 06-14-2017, 11:38 PM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by TheronB View Post
This is irrelevant to what is being discussed. The shifted camera sensor shows more detail with a better lens because it is capturing information of unlimited resolution (real life) , not projecting information of limited resolution (DMD micromirrors). The amount of information coming through the lens and hitting the camera sensor is unlimited resolution, limited only first by the lens resolving power and then sensor resolution. That is why a better resolving lens gives better results in this camera example, because the light being captured is unlimited resolution.

In the case being discussed with projectors, the resolution being projected through the lens will always be limited to 2716x1528 with the 0.67" XPR DMD as this is all the DMD can physically project through the lens as limited by the micromirrors. The shift mechanism does not increase this, it only increases the speed and modulation offset of projection. So even with XPR enabled 2716x1528 is all that is ever projected through the lens at any one time. Thus, as lenses are analog if you have a lens capable of fully resolving 2716x1528 with this DMD, that is all you need for maximum 4k quality with this projector as the combination to 4k happens in your eyes/brain, not the lens.
Dave Harper likes this.

Last edited by Ruined; 06-15-2017 at 12:04 AM.
Ruined is offline  
post #423 of 2011 Old 06-15-2017, 12:07 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
Our eyes are pixel shifting sensors, and if they are to capture the added resolution XPR provides, the lens must be able to resolve it. If the lens isn't good enough, the artifacts from each "pass" of the XPR flashes will blend together, in our brain like you have been saying, and negate the resolution gain.

The point you're arguing discredits your argument.
RLBURNSIDE likes this.

Last edited by TheronB; 06-15-2017 at 12:19 AM.
TheronB is offline  
Sponsored Links
Advertisement
 
post #424 of 2011 Old 06-15-2017, 01:09 AM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by TheronB View Post
Our eyes are pixel shifting sensors, and if they are to capture the added resolution XPR provides, the lens must be able to resolve it. If the lens isn't good enough, the artifacts from each "pass" of the XPR flashes will blend together, in our brain like you have been saying, and negate the resolution gain.

The point you're arguing discredits your argument.
No, it does not.

It's actually very simple.

The DMD can only ever project 2716x1528. No more, no less. The lens, thus, only ever has this limited 2716x1528 resolution projected through it in the projector example. Unlike your camera example where the lens has unlimited resolution being captured through it.

If the lens can fully resolve 2716x1528, then the lens does not limit the DMD. Fully resolve means the lens can perfectly or near-perfectly display the resolution from edge to edge of the frame. If there are significant artifacts in the picture it is not fully resolving the resolution.

That's all there is to it. The rest of the argument is meaningless as the lens' only job in the DLP XPR projector is to display 2716x1528 as perfectly as possible - no more, no less, regardless of whether XPR is enabled or not; even if XPR is disabled I am sure you don't want artifacts and blurry resolution, and if it is enabled all you are doing is the same 2716x1528 resolution with an offset modulation. In your camera example the lens needs to deliver as much of real life's unlimited resolution to the sensor as possible, which is a monumentally more difficult job.

Last edited by Ruined; 06-15-2017 at 01:15 AM.
Ruined is offline  
post #425 of 2011 Old 06-15-2017, 01:18 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
In that said lens there will be artifacts that wouldn't affect 2716x1528 but would affect 4K. The point at which things like spherical aberration affect real resolution is undeniably different. Those interpixel/blended spaces must be up to the 4K task.

That's all there is to it.

Last edited by TheronB; 06-15-2017 at 01:26 AM.
TheronB is offline  
post #426 of 2011 Old 06-15-2017, 01:20 AM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by TheronB View Post
In that said lens there will be artifacts that wouldn't affect 2716x1528 but would affect 4K.

That's all there is to it.
??? Invisible artifacts? Never heard of that one before.

Either the artifacts are there, or they are not there. They are not going to be invisible when the lens is tested at 2716x1528 then all of a sudden appear when XPR is enabled.

Last edited by Ruined; 06-15-2017 at 01:23 AM.
Ruined is offline  
post #427 of 2011 Old 06-15-2017, 01:26 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
See my edited post, and you can see an example of what you dismiss.
TheronB is offline  
post #428 of 2011 Old 06-15-2017, 01:29 AM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by TheronB View Post
See my edited post, and you can see an example of what you dismiss.
If there are artifacts that are causing pixels to bleed into the space between pixels, you will be able to see this when testing a lens for quality at 2716x1528. And it will be visible at 2716x1528 and thus adversely impact the quality of both the native resolution and XPR. The Sony 4K 665ES LCOS projectors are a great example of seeing this type of between-pixel artifact (in that case caused by panel alignment instead of lens) impact the native resolution negatively.

That is the whole point, all you need is a lens that can fully resolve the native resolution and you are set. If you get a subpar lens that has significant artifacts, then yeah it will be a problem both for the native resolution and XPR 4k. But for the last 5 pages using a mediocre lens was never the argument. The argument was using a lens that could fully resolve the native resolution for the 0.67" XPR DMD.
Dave Harper likes this.

Last edited by Ruined; 06-15-2017 at 01:39 AM.
Ruined is offline  
post #429 of 2011 Old 06-15-2017, 04:31 AM
 
RLBURNSIDE's Avatar
 
Join Date: Jan 2007
Posts: 3,901
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 2012 Post(s)
Liked: 1408
Quote:
Originally Posted by Ruined View Post
If the lens can fully resolve 2716x1528, then the lens does not limit the DMD. Fully resolve means the lens can perfectly or near-perfectly display the resolution from edge to edge of the frame. If there are significant artifacts in the picture it is not fully resolving the resolution.
You're not getting it, the 2.7K DMD gets shifted and thus there are two sets of independent light paths to consider. So, the lens needs to resolve both of them. Meaning, if they are offset by a 1/2 pixel, then the lens needs to be effectively 4K capable in order to not smear the detail from one of the shifted images away.

This is pretty straightforward.
RLBURNSIDE is offline  
post #430 of 2011 Old 06-15-2017, 05:59 AM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by RLBURNSIDE View Post
You're not getting it, the 2.7K DMD gets shifted and thus there are two sets of independent light paths to consider. So, the lens needs to resolve both of them. Meaning, if they are offset by a 1/2 pixel, then the lens needs to be effectively 4K capable in order to not smear the detail from one of the shifted images away.

This is pretty straightforward.
You are thinking in a digital mindset, which is where the logic is breaking down.

I will make it more basic:

A lens is round.

A lens has highest resolution in the center, and lower resolution near the edges.

A lens is fully analog, meaning there is no "pixel grid" like a digital device has

A rectangular image from the DMD is projected on the round surface of the lens

The resolution of that image is 2716x1528

If that resolution is fully resolved by the lens, this means it is fully sharp with no significant artifacts, blur, etc when this resolution is projected through it

Finally, when the image is shifted half a pixel, it will make zero difference because the lens is not a digital device. If it was completely sharp in position A, when position B is only half a pixel away it is also going to be completely sharp there as well

Where a lens would have to be "4k" is if the native resolution of the DMD were actually 3840x2140 and thus the pixels would be much smaller than the pixels projected by the 2716x1528 DMD. Even though our brain interprets smaller pixels with XPR, the lens never actually has to resolve smaller pixels with 2716x1528 XPR.
Dave Harper likes this.

Last edited by Ruined; 06-15-2017 at 06:56 AM.
Ruined is offline  
post #431 of 2011 Old 06-15-2017, 06:37 AM
 
RLBURNSIDE's Avatar
 
Join Date: Jan 2007
Posts: 3,901
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 2012 Post(s)
Liked: 1408
The fact that a lens is analog doesn't mean it's perfectly smooth, or smooth enough to not smear fine detail above a certain spatial frequency. That's the question.

Forget about XPR for a second. Think of two independent 2.7K DMDs overlaid in the same space, trying to get through the same lens at the same time, but offset 1/2 pixel diagonally from one another. If the lens can't resolve 4K, then it can't resolve both XPR images simultaneously. And that's the leap you're missing.

The fact that they aren't displayed simultaneously is irrelevant (for incoherent light sources). What matters is the spatial offset between them, geometrically. Think of it like a filter trying to let through two patterns without corrupting either one. It has to let through all the detail of each offset 2.7K imager independently.

For the purposes of how XPR doubles the spatial frequency of the projector, it's not the entire pixel extent which matters, it's the center of the pixel. That represents the sample point. So if you concentrate all the light to a point, and want to figure out "will that point have the same focus as its neighbours on a screen, after passing through the lens", then you have to consider the potential aberrations of all the sample positions. Of which there are 8 million, not 4 million.

Like I said before, in physics there is no free lunch. You don't get to borrow the sharpness of a 2.7K lens to apply it to two offset 2.7K imagers simultaneously. The fact that wobulation is used is totally irrelevant to the lens optics. If you are trying to resolve 4K, you need a 4K lens.

It's not the extent of the pixels that matters, it's the number of discrete sample positions / square mm.

Or, in other words, the minimum distance between neighbouring samples which determines their spatial frequency. And that's UHD. It doesn't matter if you're not sampling all those positions at the same time, but you do need to sample them at their correct offsets from each other, in order to distinguish them. Your eyes' ocular microtremours operate at 83hz so XPR working 120hz is more than enough to fuse two completely distinct images at different offsets into a sharper whole. The size of a 2.7K pixel vs a UHD pixel doesn't matter, what matters are the distances between the pixel centers.

And with XPR, there are 8 million distinct pixel centers to consider, offset at the same distance from one another as UHD pixel centers. This is of course by design, right? This is all a gross simplification I'm sure but it's enough to get the gist of it. The pixel extents are besides the point. Think of pixels like mathematical points with no extent. What matters for the lens is the lateral distance between those points, as they pass through. And the lateral distance is the same as that for a native UHD display, or roughly.

Last edited by RLBURNSIDE; 06-15-2017 at 07:04 AM.
RLBURNSIDE is offline  
post #432 of 2011 Old 06-15-2017, 07:05 AM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by RLBURNSIDE View Post
The fact that a lens is analog doesn't mean it's perfectly smooth, or smooth enough to not smear fine detail above a certain spatial frequency. That's the question.

Forget about XPR for a second. Think of two independent 2.7K DMDs overlaid in the same space, trying to get through the same lens at the same time, but offset 1/2 pixel diagonally from one another. If the lens can't resolve 4K, then it can't resolve both XPR images simultaneously. And that's the leap you're missing.

The fact that they aren't displayed simultaneously is irrelevant (for incoherent light sources). What matters is the spatial offset between them, geometrically. Think of it like a filter trying to let through two patterns without corrupting either one. It has to let through all the detail of each offset 2.7K imager independently.

For the purposes of how XPR doubles the spatial frequency of the projector, it's not the entire pixel extent which matters, it's the center of the pixel. That represents the sample point. So if you concentrate all the light to a point, and want to figure out "will that point have the same focus as its neighbours on a screen, after passing through the lens", then you have to consider the potential aberrations of all the sample positions. Of which there are 8 million, not 4 million.

Like I said before, in physics there is no free lunch. You don't get to borrow the sharpness of a 2.7K lens to apply it to two offset 2.7K imagers simultaneously. The fact that wobulation is used is totally irrelevant to the lens optics. If you are trying to resolve 4K, you need a 4K lens.

It's not the extent of the pixels that matters, it's the number of discrete sample positions / square mm.

Or, in other words, the minimum distance between neighbouring samples which determines their spatial frequency. And that's UHD. It doesn't matter if you're not sampling all those positions at the same time, but you do need to sample them at their correct offsets from each other, in order to distinguish them. Your eyes' ocular microtremours operate at 83hz so XPR working 120hz is more than enough to fuse two completely distinct images at different offsets into a sharper whole.
Maybe our definition of "fully resolve" is different.

When I say fully resolve, I mean fully - as in, all of the pixels are sharp, defined, as is the space in between them, without significant artifacts. The entire 2716x1528 grid is represented.

If there are any abberations present, they will be able to be easily evaluated using a 2716x1528 source on the lens. They will not just suddenly appear with a 4K source w/ XPR.

Thus, I maintain that if you can get a fully resolved 2716x1528, it should be no issue whether its position A on the lens or position A+0.5 pixels. If there are issues with 2716x1528 regarding pixel definition or artifacts you'd be able to see them at 2716x1528 when evaluating the quality of the lens.

Finally, if you fully believe your theory, you actually need a lens that is capable of resolving better than 5K, not 4K. The "composite" resolution of XPR is 5432x3056 when overlapped. The algorithm in the projector then takes the 3840x2160 source and manipulates the overlaid images to best represent 3840x2160 in the virtual 5432x3056 resolution.

Last edited by Ruined; 06-15-2017 at 07:17 AM.
Ruined is offline  
post #433 of 2011 Old 06-15-2017, 07:32 AM
AVS Forum Special Member
 
3DBob's Avatar
 
Join Date: Aug 2014
Location: Southeastern Michigan
Posts: 2,432
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 1078 Post(s)
Liked: 577
In the end, if the image looks good and gives you the impression of good 4k, that is all, of course that matters. I think the better example of resolution is not the limit of the lens resolution, but how it compares to standard 4K LCD or OLED resolution. Where projected 4K and standard TV generated 4K differ is the size of the screen. If you have a 60" 4K TV and you watch it next to a projected 60" image, then you'll see the difference in pixel resolution, but only if you get about a foot from the screen. And at that point the discussion is mute anyway since you will probably never watch a 60" projected image, but most likely a 120" or greater image and the pixels are going to be "huge" by comparison. It's just a matter of sitting back far enough to get the illusion that standard 4K TV offers on such a big screen. Mixing photography lens limits with projection of video pixels is, well, apples and oranges. But, I do enjoy lively technical discussions.... Kind of reminds me of the early days of LCD projection TVs where the screen-door effect drove people nuts so lens manufacturers created a slight lens blurring effect to minimize it and people thought that the smoother look was great. Of course, the muted lens resolution of the individual pixels was no longer an issue--the issue became how the image looked to the audience.
3DBob is offline  
post #434 of 2011 Old 06-15-2017, 07:41 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
The artifacts may not show up as such. They could be seen as smearing or just lack of fine resolution detail. It doesn't need to be a flare or ghosting or vignetting that shows up so easily.
TheronB is offline  
post #435 of 2011 Old 06-15-2017, 07:47 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
Quote:
Originally Posted by 3DBob View Post
In the end, if the image looks good and gives you the impression of good 4k, that is all, of course that matters. I think the better example of resolution is not the limit of the lens resolution, but how it compares to standard 4K LCD or OLED resolution. Where projected 4K and standard TV generated 4K differ is the size of the screen. If you have a 60" 4K TV and you watch it next to a projected 60" image, then you'll see the difference in pixel resolution, but only if you get about a foot from the screen. And at that point the discussion is mute anyway since you will probably never watch a 60" projected image, but most likely a 120" or greater image and the pixels are going to be "huge" by comparison. It's just a matter of sitting back far enough to get the illusion that standard 4K TV offers on such a big screen. Mixing photography lens limits with projection of video pixels is, well, apples and oranges. But, I do enjoy lively technical discussions.... Kind of reminds me of the early days of LCD projection TVs where the screen-door effect drove people nuts so lens manufacturers created a slight lens blurring effect to minimize it and people thought that the smoother look was great. Of course, the muted lens resolution of the individual pixels was no longer an issue--the issue became how the image looked to the audience.
This is true, but compression artifacts may probably be the biggest culprit of lost detail with non-UHD disk sources. GIGO as they say.

Also, the best reason for 4K is being able to sit closer to a big screen IMO. Contrast/HDR and color are more important overall.
TheronB is offline  
post #436 of 2011 Old 06-15-2017, 08:07 AM
Newbie
 
Join Date: Jun 2017
Posts: 7
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 0
Prices still starting at 2200 here in Berlin
trimak is offline  
post #437 of 2011 Old 06-15-2017, 08:21 AM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
TheronB and RLBURNSIDE,

I am curious what you think of the example I posted earlier about eye charts and reading glasses:
Quote:
Originally Posted by darinp View Post
You can try the 3 lines yourself with this image. Move back until the E3 looks like a blocky 8. Can you still make out the E and the 3 on the first 2 lines?



I don't need reading glasses to see the standalone E or the standalone 3, but I do for the E3.
I didn't think I could make things much simpler than that, yet Ruined continues posting his nonsense about being right because lenses are analog. So are reading glasses. I doubt he'll understand this example either.

This discussion was never about whether a lens that is perfect for the native resolution will be good enough for eShift on. It was about whether it matters whether the 2 images go through the lens at the same time or at different times. Just like with the glasses example, it doesn't for real world lenses.

It seems to me that Ruined has changed his argument a little bit to something more like if a lens scores a 10 out of 10 for sharpness with the native resolution it will score at least an 8 for eShift on, and that will be good enough, or maybe not. Real world lenses will obscure the native images and scored by how well they deliver those images to human viewers (which is their job) they will do worse delivering the images with eShift on than eShift off, even though both scores might be perfectly adequate. Whether the eShifted elements go through the lenses at the same time or different times is basically irrelevant to those scores just like with the reading glasses example where it is irrelevant whether the E and the 3 go through the glasses together or separate.

--Darin
RLBURNSIDE likes this.
darinp is offline  
post #438 of 2011 Old 06-15-2017, 08:30 AM
Advanced Member
 
Join Date: Jun 2014
Posts: 946
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 485 Post(s)
Liked: 125
That one and the one with the overlapped squares illustrate it clearly IMO.
RLBURNSIDE likes this.
TheronB is offline  
post #439 of 2011 Old 06-15-2017, 10:31 AM
AVS Forum Special Member
 
3DBob's Avatar
 
Join Date: Aug 2014
Location: Southeastern Michigan
Posts: 2,432
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 1078 Post(s)
Liked: 577
Of course in projection, only a small portion of the lens is used--check out a shifted lens and you will see what I mean. Because of the speed of light, a wobbled light source will never go through the lens at the same time, but our eyes are so slow that it appears that it does. Doesn't really matter about the hz either. Lighted pixels do not hold and release all the light from the chip at one time, so light of each pixel is going through the lens at different times (we are talking at the speed of light here). The lens offers no confusion to the light--our eyes do.
3DBob is offline  
post #440 of 2011 Old 06-15-2017, 10:44 AM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
Quote:
Originally Posted by Dave Harper View Post

This isn't even relevant to how eShift works since your "E" and "3" are on totally separate lines, above and below one another.
If you are so concerned with them being in different places on the display then just cover 2 lines at a time and use the scroll bar on your browser to put the one you want to see in the spot you are looking at. From what you and Ruined have argued over and over, if the E3 is displayed by flashing the E every even millisecond and the 3 every odd millisecond, then if the glasses are good enough for the E by itself to a human and the 3 by itself to a human, they will be good enough for the E3 for a human if the E and 3 are flashed extremely quickly in sequence, since the E and the 3 never go through the reading glasses at the same time. I think you guys have repeated that good enough for the E means good enough for the E3. That is not true, whether the E3 are flashed extremely quickly in sequence or displayed at the same time. It is only true if they are flashed so slowly that your eyes make them out as separate images, which isn't how eShift works.

This example is definitely relevant to whether the quality required of lenses (whether in projectors, glasses or human eyeballs) for humans is based on what goes through them in an instant in time, or what is perceived to go through them in an instant of time.

--Darin

Last edited by darinp; 06-15-2017 at 11:20 AM.
darinp is offline  
post #441 of 2011 Old 06-15-2017, 11:10 AM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
Quote:
Originally Posted by 3DBob View Post
Of course in projection, only a small portion of the lens is used--check out a shifted lens and you will see what I mean. Because of the speed of light, a wobbled light source will never go through the lens at the same time, but our eyes are so slow that it appears that it does. Doesn't really matter about the hz either. Lighted pixels do not hold and release all the light from the chip at one time, so light of each pixel is going through the lens at different times (we are talking at the speed of light here). The lens offers no confusion to the light--our eyes do.

I'm not sure exactly what you are trying to say. Our eyes confuse things, but the amount the lenses obscure the images (which any real lens does) affects whether are eyes will be able to see the intended detail.

If you are saying that the lens only has to be good enough for what goes through it at one instance then this is incorrect for displaying images to humans. The composite of all the images that went through the lens is how the lens performance is determined. For humans it needs to be good enough for the composite image, just like in the following the glasses needed to read the last line are the same whether the E and 3 go though the glasses at the same time or different times, as long as they flash fast enough that humans don't know whether they are ever up at the same time.

The glasses needed to read the first line are not the same as the glasses required to read the last line. Same with the lenses in eyeballs. If the E is put up every even millisecond the visual acuity required to see an E3 on the last line instead of an 8 is the same whether the 3 is put up every even millisecond or every odd millisecond.

--Darin

Last edited by darinp; 06-15-2017 at 11:14 AM.
darinp is offline  
post #442 of 2011 Old 06-15-2017, 11:34 AM
 
RLBURNSIDE's Avatar
 
Join Date: Jan 2007
Posts: 3,901
Mentioned: 16 Post(s)
Tagged: 0 Thread(s)
Quoted: 2012 Post(s)
Liked: 1408
Quote:
Originally Posted by 3DBob View Post
Kind of reminds me of the early days of LCD projection TVs where the screen-door effect drove people nuts so lens manufacturers created a slight lens blurring effect to minimize it and people thought that the smoother look was great. Of course, the muted lens resolution of the individual pixels was no longer an issue--the issue became how the image looked to the audience.
That's funny, I defocus my Gear VR ever so slightly to try and blur the subpixels together, otherwise I can clearly see that a white pixel is actually a four separate pixels in a pentile pattern. That's another cool benefit of a single-chip design, the subpixels are always perfectly aligned and overlapping in space. And a good reason to consider that 3-chip LCDs, DLPs, and LCoS have the potential to be better for high res VR helmets too, for the same reason. It would be far less jarring if, given the same RGB resolution, those RGB subpixels were indeed (perfectly or not) overlapped.
RLBURNSIDE is offline  
post #443 of 2011 Old 06-15-2017, 11:44 AM
Advanced Member
 
Join Date: Nov 2016
Posts: 539
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 261 Post(s)
Liked: 202
when i get my unit; if the 4K imagery is not apparent to me; the projector will be returned simple as that. I"ll let you know. = )
We shall see how good this "XPR" 4mill to 8mill pixel magic/trickery works for 2160p!
RLBURNSIDE and 3DBob like this.
am2model3 is offline  
post #444 of 2011 Old 06-15-2017, 12:06 PM
AVS Forum Special Member
 
Dave in Green's Avatar
 
Join Date: Jan 2014
Location: USA
Posts: 8,354
Mentioned: 152 Post(s)
Tagged: 0 Thread(s)
Quoted: 3787 Post(s)
Liked: 3001
Quote:
Originally Posted by sage11x View Post
So... I think all you guys are great... just fantastic human beings... but maybe we need to give the lens talk a break. Just a time out. Reconvene at a later date. You know, let the air out of the room a bit.
I disagree. This discussion is an example of what makes AVS Forum so great. Each side seems to be frustrated that the other side can't see what to them seems obvious, which is understandable. But the discussion is staying focused on a technical issue and is not devolving into a name-calling contest. I think in some ways both sides are in agreement on many points, so perhaps it would be productive to try to narrow it down a little and keep the focus on the primary area of disagreement.
Dave Harper likes this.
Dave in Green is offline  
post #445 of 2011 Old 06-15-2017, 12:10 PM
AVS Forum Special Member
 
3DBob's Avatar
 
Join Date: Aug 2014
Location: Southeastern Michigan
Posts: 2,432
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 1078 Post(s)
Liked: 577
Quote:
Originally Posted by darinp View Post
I'm not sure exactly what you are trying to say. Our eyes confuse things, but the amount the lenses obscure the images (which any real lens does) affects whether are eyes will be able to see the intended detail.


--Darin
I think we are saying the same thing, but I was trying to make a point that a lens is only as good as the resolution detail of a single pixel, doesn't matter how many pixels add up to make the total image. They all come through the lens at different times. Where they come through the lens might make a difference between the sharpness of one pixel over another, but typically the pixels come through the "sweet" part of the lens unless you are really lens shifting the light to an outer portion of the lens. As for detail, it can only be as wide as a pixel, not less than a pixel as we can't fully see it then. But our eyes sometimes confuse contrast pixels versus detail pixels. As in high-pass sharpening, unsharp mask or clarity sharpness in processed image photography. You think you can see more detail, but all you're doing is enhancing each pixel with more contrast and isolation against the neighboring pixels, not providing more detail. And, most of us don't see all that well anyway. There is a phenomenon, though, with 3D images that "apparently" doubles the resolution in your mind over 2D images. There is a subtle shift in detail between one eye and the other, which combined appears to give the illusion of more resolution. I've noticed it early on in watching 3D, and sometimes will watch a 2D movie in converted 3D to up the illusion of detail. It has been discussed along the way in other threads as well.

Regardless, once the reviews come in, we will all get a better feel for the projectors, and like all humans, opinions will flourish good and bad.
3DBob is offline  
post #446 of 2011 Old 06-15-2017, 12:14 PM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
Quote:
Originally Posted by Dave in Green View Post
I disagree. This discussion is an example of what makes AVS Forum so great. Each side seems to be frustrated that the other side can't see what to them seems obvious, which is understandable. But the discussion is staying focused on a technical issue and is not devolving into a name-calling contest. I think in some ways both sides are in agreement on many points, so perhaps it would be productive to try to narrow it down a little and keep the focus on the primary area of disagreement.
If people want to move this discussion to the thread I created for it about a year and a half ago, I am okay with that. That thread is here:

https://www.avsforum.com/forum/24-dig...l#post39064946

I went through all this before. To begin with it was mostly me arguing against a big crowd. There are some pretty technically savvy people on the >$3.5k forum and in the end I think most of them saw that I had been right the whole time. The composite image is what matters.

I'm glad Dave Harper brought up the eyeglasses thing. I think that explains things very well. If anybody thinks that with eShift images the eyeglasses only need to be good enough for the native resolution of the projector, since that is all that goes through the eyeglasses at one time, then they need to keep thinking.

--Darin
rocklee likes this.
darinp is offline  
post #447 of 2011 Old 06-15-2017, 12:25 PM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
Quote:
Originally Posted by 3DBob View Post
As for detail, it can only be as wide as a pixel, not less than a pixel as we can't fully see it then.
While that seems logical, I think I proved that isn't the case with all 3 of these images:

1:


2:


3:


In the first image the small white square is detail that can be 1/4th the area of the 2 overlapping pixels that were used to create it (like eShift). Or the same image could be made with 7 pixels. It doesn't matter which method is used to create it, the lens requirements are essentially the same (assuming idealized 100% fill ratio pixels).

In the third image the black between the E and the 3 is finer detail than either the E or the 3 contains. That image could be created with a single projector, or 2 projectors could be used to create it. That is, you could have one projector that can only display the large E and one that can only create the large 3. Each could be perfectly fine for displaying their character alone, but as soon as you want to put them together to show an E very close to a 3, the requirements for how sharp they need to be have humans see what is intended go up. That E can be smeared and still look like an E, but have that much smearing when both projectors are going and you get an 8 instead of an E3.

I know it might surprise some that detail that is smaller than what ever goes though the lens at a moment of time is possible, but that is what eShift is all about.

--Darin

Last edited by darinp; 06-15-2017 at 12:30 PM.
darinp is offline  
post #448 of 2011 Old 06-15-2017, 01:15 PM
AVS Forum Special Member
 
3DBob's Avatar
 
Join Date: Aug 2014
Location: Southeastern Michigan
Posts: 2,432
Mentioned: 31 Post(s)
Tagged: 0 Thread(s)
Quoted: 1078 Post(s)
Liked: 577
Darin, we are in agreement. Would be good to look at the Benq specs for a moment to see how they are handling the lens issue. Not saying that Optoma is doing something similar, but I have to think the UDH65 will have a similar lens.

http://www.benq.us/product/projector/ht8050/features/

https://www.optomausa.com/projectorproduct/uhd65
3DBob is offline  
post #449 of 2011 Old 06-15-2017, 01:35 PM
AVS Forum Special Member
 
Ruined's Avatar
 
Join Date: Jul 2002
Posts: 6,848
Mentioned: 35 Post(s)
Tagged: 0 Thread(s)
Quoted: 2635 Post(s)
Liked: 1265
Quote:
Originally Posted by 3DBob View Post
Darin, we are in agreement. Would be good to look at the Benq specs for a moment to see how they are handling the lens issue. Not saying that Optoma is doing something similar, but I have to think the UDH65 will have a similar lens.

http://www.benq.us/product/projector/ht8050/features/

https://www.optomausa.com/projectorproduct/uhd65
I believe BenQ offers more lens shift range in their model which (if so) leads to a significantly more costly lens.

I wouldn't buy a ht8050 anyway though as it is first gen and missing some key features, BenQ promised a fall refresh that will add HDR and a better DI. It will likely be more costly than the Optoma UHD65. Chassis is a lot nicer on the BenQ, too.

Then again with the new Optoma UHZ65 laser 4k announcement the BenQ next-gen lamp may be a tough sell, unless you have placement issues.

Last edited by Ruined; 06-15-2017 at 01:41 PM.
Ruined is offline  
post #450 of 2011 Old 06-15-2017, 03:10 PM
AVS Forum Special Member
 
darinp's Avatar
 
Join Date: Jul 2002
Location: Seattle, WA
Posts: 4,735
Mentioned: 17 Post(s)
Tagged: 0 Thread(s)
Quoted: 543 Post(s)
Liked: 749
Quote:
Originally Posted by 3DBob View Post
Darin, we are in agreement.
Sorry that I misinterpreted.

Thanks,
Darin
darinp is offline  
Sponsored Links
Advertisement
 
Reply Digital Projectors - Under $3,000 USD MSRP

Tags
dlp , hdr , Optoma , uhd , uhd60

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off