[Archangel]

Did those that saw the HS20 at CEDIA....think that the Screen Door Effect was still significant ?


Saw two reports and got two different opinions.


According to Secrets of HT &HF

Quote:

Sony had both of these projectors set up side by side with a black curtain separating the projectors and screens, but arranged in a way that you could easily view both images simultaneously. Both projectors looked good, but the HS-20 looked better, as you might well expect. The major issue with LCD-based projectors is the screen door effect and I must say that I find both the HS-10 and HS-20 to have somewhat worse screen door than the Sony 11HT I use on a daily basis. I was aware of the Screen Door Effect on both projectors at a viewing distance of approximately twice the screen width. Of course, you can’t see the actual grid, but, from my experience, the screen door effect does make the image seem somewhat grainy, and perhaps less real, compared to DLP images of similar resolutions. Nevertheless, given the HD resolution, and the excellent brightness and contrast ratio for and LCD projector, I have no doubt that the HS-20 will be even more popular than the HS-10, and rightly so.

Then according to Projector Central on the HS10....which I presume will only be better for thew HS20:


Quote:

With regard to the screendoor or pixelation effect, it is virtually absent. At a viewing distance of 1.5x the screen width, pixel structure is not visible at all in the video image. It is barely perceptible in rolling credits and subtitles, but only if you have 20/15 vision. So the short of it is that the screendoor problem does not exist on the HS10.

So which is it ???

 

[CMRA]

They are basically one in the same. Both mentioned they couldn't detect the gridlines. One went on to say that SDE ( LCD technology ) was apparent due to the graininess of the picture that would not be an issue with a DLP projector. Obviously he finds this objectionable. And, has very good eyes too.

 

[gmshoemaker]

Quote:

Originally posted by Archangel Then according to Projector Central on the HS10....which I presume will only be better for thew HS20: So which is it ???

I would expect it to be as good or better than the HS10. From my experience with the HS10, it is non-existant at 1.5 x screen width. I have 20-20 corrected vision. Your milage may vary.


Gary

 

[TomHuffman]

This is one of the strangest subjects regarding the HS10/20. A few people report some SDE. I own the HS10 and I can see none, nada, zip. Obviously, some people are seeing something I cannot.

 

[HTCrazy]

I agree it's weird that people would report a problem with the HS10 screen door. I stop seeing it any further than 5 feet from a 9 foot screen.

 

[tbodin]

Same here. It's something I feared before I bought based on all the discussion of the issue. My concern was gone within 30 seconds of hooking up the HS10 (WOW!), and I have never noticed it since.

 

[BigScreen]

I've seen the HS10 on two different occasions with a variety of source material (DVD, HDNet, and standard broadcast), and I could not see any screen door effect at all.


The broadcast demo was in a situation where I couldn't get closer than 2x screen width, but the first two sources were viewed from a variety of distances including a couple of inches away.


I've seen screen door effect on some LCD and DLP projectors which would prevent me from paying any price, it was just too distracting.


I'm looking forward to seeing the HS20 in person, hopefully they won't be on allocation and unavailable...

 

[dna]

Screen door is reduced with the use of MLA which I understand the HS20 uses. MLA is used in some other LCD projectors such as the PLV-70/HD20. (MLA also is used primarily to increase brightness.) Screen door can be almost eliminated at any reasonable viewing distance through the use of slight adjustment in focusing, enough to blur pixels together but not enough to reduce pixel resolution. (Both MLA and LCD focusing have been discussed in hundreds of posts on the digital projector forums, if you want more details.)


For example, on the PLV-70, when correctly focused, with my 20/20 vision I am unable to see any screen door beyond about .5X screen width. This suggests to me that even someone with 20/10 vision should not be able to see screen door at distances greater than 1X screen width on the similar (resolution, MLA) HS20.


I am not suggesting that people who are complaining about HS20 screen door at 1.5X or greater viewing distances are making this up. However, I don't believe they are actually seeing screendoor.


All LCD projectors have varying degrees of FPN (Fixed Pattern Noise) which is caused by LCD drivers that varying in somewhat in gain from driver to driver. This causes a pattern in the picture that some people are more aware of than others. FPN is easily seen in light slowly moving picture areas, such as sky. I suspect that what some people are referring to as screen door is actually FPN, especially if they can still see it in a high-res LCD projector with MLA that is correctly focused.


DLP projectors (both single and 3-chip) do not have FPN since they do not have analog drivers. LCOS does but JVC provides driver gain calibration that can be adjusted in the service screen. (I don't know about SXRD but I was not able to see any FPN in the couple demos I saw.)


-David

 

[rogo]

And I, personally, think that the "screendoor in the sky" effect is, indeed, screendoor and is not fixed-pattern noise. I think the pixel grid is visible in those scenes because they are very light / bright so the contrast of the wire grid makes it very visible. It happens in white areas, too, but in blue areas it is especially bad because the red and green sub-pixels are dark or barely lit, making the "grid portion" of the image that much more visible.


I am not personally that experienced with FPN patterns -- which I understand are often somewhat acute on D-ILA/LCOS projectors -- but I think screendoor is screendoor. And it's obvious that the darker the image, the less you'd be able to detect the grid / screendoor.


Mark

 

[dna]

Hi Mark!


Could you tell us a bit more about the LCD screen door you see? What was the viewingDistance/screenWidth ratio? What projector(s)? Did they have MLA? What was the fill factor? Was the focusing adjusted to just fill in (blur) the space between pixels?


Here's what puzzles me about the ability some people say they have to see screen door at "normal" viewing distances: Typical current LCD fill factor is about 87%, as I remember. The ratio of pixel length or width to the screen door gap is about 6.7:1. Even without MLA and defocussing and assuming perfect convergence, you have to be 6.7X closer to the screen to see screen door than to resolve individual pixels. Right?


How can it be that people can see screen door at normal viewing distance? If one can see screen door on a 720P projector at 1.5X, that means they can resolve 720P resolution at 10X viewingDistance/screenWidth ratio . This is vastly beyond the 60 degree eye resolution (for those with 20/20 vision) ability that tests show humans can resolve.


I am eager to have someone explain how it is possible to resolve lines between pixels that are 1/6.7th as thick as pixels at viewing distances appropriate for the resolution of the projector (not to mention projectors that are defocussed enough to fill in between pixels).


-David

 

[rogo]

I'll comment more on this later, but I want to start with this:


AFAIK, no LCD microdisplay has anything remotely resembling an 87% fill. That's about what the DLP fill is.


LCD fill in that environment is probably in the 60% realm.


And I also think this fancy math about human vision is misleading. You don't have to resolve the grid itself. You simply have to have something move across the grid for it to be apparent. Or you can have something bright enough to illuminate the fact there is a grid, without given you "true resolution" of the grid.


I can pick up screendoor on most LCD projectors with ease at 1.5-1.75 screenwidths away. On certain moments of content, I can do it at 2 screenwidths away. It is a mistake -- in my opinion -- to conclude that these types of content suddenly introduce fixed-pattern noise when the apparent grid that results (or, more accurately, the clear visual information that a grid is present) is already clearly known to be a function of the inter-pixel gaps,

 

[TomHuffman]

My vision is normal. I can see SDE on the HS10 only by putting my eyes a foot or two from the screen. Unless there's some weird unit-to-unit variation in this model or some others have Clark Kent eyes, I'm very skeptical about this claim.

 

[gmshoemaker]

Quote:

Originally posted by TomHuffman My vision is normal. I can see SDE on the HS10 only by putting my eyes a foot or two from the screen. Unless there's some weird unit-to-unit variation in this model or some others have Clark Kent eyes, I'm very skeptical about this claim.

Chuckle. I can see it on the white Cineza warmup screen at 1.5. (I usually find something to do during that eternity.) Otherwise, I don't see it unless I am real close. Maybe it is like rainbows on DLP, some people see it and some don't. Although I suspect if we all went out and bought $10k projectors we would suddenly see screendoor on HS10s.


Gary

 

[darinp]

I think we are talking about 2 things here:


1. If I put up a static image and then look for screen-door I will be able to see it clearly at viewing ratio x.

2. While I'm watching movies I sometimes become aware of digital pixel structure or screen door at viewing ratio y.


Maybe I'm wrong, but it sure seems like people are talking about different things. Otherwise, it is hard for me to explain what some people claim for how close they have to get to an HT1000 to see the pixels. I think they are describing where they can actually see the individual pixels and pick them out, as opposed to being aware that there are lines between them or that there is a digital makeup there, etc.


The point at which you are aware of lines between the pixels should vary by images (almost solid single color images with movement making them easier to see). If they don't vary by image makeup and movement, then I think you are talking about a different test than I am.


--Darin

 

[dna]

Mark,


I searched the web and the forum to see if I could find the fill factor percentage for the 20HD (PLV-70) that I spent some time watching awhile back at JimmyR's house. So far no luck. However, you are right, my figure of 87% is on the high side; LCD projectors "have fill factors from 60% to roughly 80%" .


As long as the fill factor is greater than 50%, I maintain that if you can see screen door, you are sitting closer to the screen than necessary to see all the resolution on the screen. Just move back until you can't see the screen door and you'll still be close enough to see all the picture detail your projector can display.


You didn't answer my question on whether you've seen LCD with MLA that is "correctly" defocussed. It does make a big difference. There is no reason to focus an LCD projector on the screen door. You do not gain picture resolution but you do significantly reduce screen door.


-David-NA

 

[EHUFF]

I am in agreement with dna and have properly defocused my 10HT for 3 years with no ill effects to the viewing experience, IMHO. I have never been distracted by SDE using this method.

 

[Victor C]

I have better than 20/20 eye sight. I can occasionally see patches of pixels even from 2X of width away. I almost never saw any pixels from 2.5X away. I don't like to defocus my HS10. I like razor sharp images. For screen doors, I might have seen a trace of it on some DVDs even at 2.5X away. It appeared as a very short, thin, dim and a bit darker than background vertical line on with bright sky as background. It is strange, I have not seen any screen door on HDTV. I have tried looking for it on PBS HD loops. They have a lot of outdoor scenes. Bright blue sky and ocean, but no screen door at all. That leads me to suspect that the "screen doors" I saw on DVDs may be in fact noises. I don't know, is that even possible?


In conclusion, I am not much bothered by the pixels I can see from 2X distance. No one in my family can see that. I am not much bothered by the "screen doors" either. Again, none of my family member can see that either. In fact, none of my many guests could see them either.

 

[TomHuffman]

gmshoemaker:


Chuckle all you want, but what I report is not abstract theory or idle speculation. It's what I and every person I've had at my home actually perceive. No one sees a SDE, even on the dreaded start-up logo.


BTW, if readers of this thread consider the original quote:


"you can’t see the actual grid, but, from my experience, the screen door effect does make the image seem somewhat grainy, and perhaps less real, compared to DLP images of similar resolutions."


you can see that we may be talking past one another. Even this critic admits he cannot see the grid. However, the screen door is the grid! If you can't see that, then you can't see SDE. The author complains about "grainy" and "less real" images than comparable DLPs. This may be due to the subtle, residual visibility of pixel structure, or it may be due to something else entirely.


However, whatever it is, it is not a SDE, at least as that phrase has been used in the past when describing digital images. I've seen it in older generation models and it really resulted in looking at an image as though through a screendoor. It's a really obnoxious effect, and the HS10 has none of this.

 

[gmshoemaker]

Quote:

Originally posted by TomHuffman gmshoemaker: Chuckle all you want, but what I report is not abstract theory or idle speculation. It's what I and every person I've had at my home actually perceive. No one sees a SDE, even on the dreaded start-up logo.

I wasn't laughing at you. I agree with you. My chuckle was at your reply to Rogo's theoretical explanation of why we should see it, when all HS10 owners on this thread, contrary to "theory", have said they don't. Sorry for the miscommunication.


Gary

 

[rogo]

Look, I know what the grid is. I also own an LCD projector so I live with the grid. I also can stand up close to the screen and see how wide the grid really is.


I also have seen the HS20, although it was probably focused pretty dead on, and it has screendoor. I don't know when I would lose the screendoor on it though.


On the Z1, I can still detect the screendoor against bright material at 1.75-2 screenwidths away. Is it objectionable? No, but it's there. At 1.5 screenwidths away, it is objectionable -- to me.


I have very good vision, up to 20/13.


On the HS20, I'd probably just focus the damned thing. I'd rather avoid defocus -- which is no panacea anyway -- and admit that I can't sit that close. That's my preference. I will probably get an HS20, too.


I personally think that the higher-res LCDs have less-apparent screendoor all else being equal because the finer the grid, the harder it is to see at distance -- even if fill factor is the same as a bigger grid. Ultimately, you either see squares or not -- the appearance of the grid. If the squares are smaller -- even if the grid is the same width -- your visual acuity makes it harder to resolve the grid.


I have to say that people who cannot see the grid on an LCD projector without jamming there face up against the screen either have their projectors fairly well out of focus or do not see very well.


In fact, against a solid background of light color, I use the grid to focus the projector. And I can do this from a pretty substantial distance away and often get the focus dead on using this trick (note that the focus tool of the HS-10 has no analogue on the Z1, it's optical using source).


Mark