Film-Tech Forum ARCHIVE: screen size and resolution

This is topic screen size and resolution in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=001294

Posted by Scott Norwood (Member # 30) on 11-27-2012, 09:01 PM:

My experience with DCI equipment has been limited to 2K projection and small/medium screens (30' wide or smaller). I have never seen 4K DLP projection, and am unaware of any venue in this area that has the capability. My personal feeling is that 2K is "adequate" on this screen size, but not as good as a top-quality 35mm print.

Assuming 4K source material, at what screen size is the difference between 2K and 4K noticeable? At what screen size is 4K necessary? Is there a real and noticeable benefit to 4K on the smallest screens, or is it really just for extra-large screens and special venues?

Posted by Victor Liorentas (Member # 5272) on 11-27-2012, 09:40 PM:

My screen is 32' wide,our new Christie 2220 is 2k being served with a Doremi Show Vault IMB and as nice as it is my best 35mm prints look better to me overall.
The digital has it's strengths for sure but up close I can see the cute little pixels and stair stepping on text and round objects.
It's 4k upgradable and if we ever get there I would hope to see the stair stepping disappear.Also I hope the morie I can see up close would go away with 4k.

Posted by Harold Hallikainen (Member # 5405) on 11-27-2012, 10:26 PM:

We have 2k and 4k projectors onto a small screen (maybe 20 feet) in our conference room. I can clearly see the difference between 2k and 4k. The 2k looks great, but the 4k is visibly better.

I wonder how much screen size plays into the resolution question. With a larger screen, people tend to be farther away from the screen, so the larger pixels appear smaller, possibly about the same size as when they sit closer to a smaller screen.

There's a story about an early public demonstration of television by RCA. Sarnoff (head of RCA) apparently objected to the visible scan lines the day before the demonstration. He congratulated the engineers on fixing the problem the next day. They had moved the chairs back.

Harold

Posted by Carsten Kurz (Member # 5396) on 11-28-2012, 06:40 AM:

It is only a matter of relative viewing distance. Even a 20ft screen can benefit from 4k.

Screen perforation moire does not necessarily go away with 4k - also microperf is only necessary for closer viewing distance. I start to notice our standard perf from below 1screen width distance.

Microperf, however, is a real problem with sound, as it cut's out more high frequency than todays tweeters can compensate.

We have a 'traditional' longish auditorium, and our patrons usually sit in the back part, as we rarely sell out our +400 seats. So, realistically, very few people would actually benefit from 4k here.

- Carsten

Posted by Bajsic Bojan (Member # 4928) on 11-28-2012, 07:52 AM:

Have not seen 4k projected with 4k material so i cannot say. But we have a 20" screen and while flat looks ok, the scope image is worse than 35mm and i'm waiting for the day 4k is mandatory [Razz]

Posted by Bobby Henderson (Member # 840) on 11-28-2012, 09:01 AM:

The quality limits of a video projector's lens and the perforations in the movie screen both will work to hide the jaggy look of the movie's pixel grid. Those factors can be bad enough in a given theater where the projected image just doesn't look like it is all the way in focus. If you have a really good projector and screen setup, the pixel grid of 2K imagery will become obvious if you get close enough to the screen. It can especially show up on graphical items like lettering in the end credits.

Doing a little simple math one can determine the pixels per inch resolution of the projected image by dividing the native resolution of the image and the screen size.

If you have a 30' wide screen and are throwing a 2048 pixel wide image onto it the projected image will have a resolution of 5.68 pixels per inch. If native 4K material is projected in 4K resolution onto the same screen it will have a resolution of 11.37 pixels per inch.

Viewing distance is important. I can design a 14' tall x 48' wide billboard with as little as 6 ppi resolution if the only people viewing the board are motorists on a busy freeway. That full scale image is still going to be 3456 pixels in width. Normally I design billboard images with a full scale resolution of 25 pixels per inch, making a 14' x 48' board image 4200 X 14400. Static, printed billboards are different from projected moving images; flaws will show more easily in s static, printed image. Lots of billboards are getting converted over to LED-based displays. The limited resolution is clearly visible in these boards. Cost per pixel keeps the electronic billboards down to pretty coarse, low levels of resolution.

In graphics work if the intended viewing distance gets closer the native resolution in the finished project must go higher. Most properly designed vehicle wraps are designed with a full scale resolution of 72ppi. A large poster may need a mix of 150ppi Photoshopped material and vector-based logos, type and other sharp graphics. When you get down to a standard page sizes the photo resolution requirements can go to 300ppi and beyond (and you better be using vector-based art for type).

Posted by Antti Nayha (Member # 4980) on 11-28-2012, 10:36 AM:

Obviously, neither ppi nor viewing distance alone determines the human-discernible resolution – it’s the combination of the two. There’s no reason why a small screen couldn’t benefit from 4K, provided that the patrons sit close enough.

(Warning: Some theory coming up, feel free to skip to the end of the post if you’re not in the scientific mood right now.)

See this presentation by Mikko Kuutti: How many pixels?
(Very informative, although I’m pretty sure that most people will find his preferred 90° viewing angle way too close to the screen, unless we’re talking about films specifically made for giant 15/70 IMAX screens.)

Also check out this paper by Sony: Does 4K really make a difference?
(Again, good information… but could be a bit biased, since Sony was the only manufacturer who had a 4K system in the market at the time.)

If you want to try it out yourself, just download and project this 2K test pattern that I made.

From close up, you will clearly see that the pattern consists of individual white pixels separated by a black grid. As you slowly move farther away from the screen, the black and white pixels will start to merge to gray. At some point, you will only see a solid gray screen with no visible texture whatsoever. Now, this is the 2K/4K threshold distance for your auditorium and your eyesight – in other words, sitting any closer from here, you would theoretically benefit somewhat from upgrading to 4K.

Personally, it looks like I can just about make out the pixel texture when sitting closer than about 2.5 x screen width, if the focus and convergence have been adjusted properly. If my trigonometry is correct, this just happens to match the 1°/90–1°/60 range of smallest human-discernible detail mentioned in Mr. Kuutti’s presentation.

That means I should (again, in theory) be able to see the 2K/4K difference in most auditoriums. Then again, of course, you won’t find a lot of real-life films that contain such high-contrast, single-pixel test patterns… [Cool]

It’s probably impossible to say where the precise threshold for any real benefit falls, since it also depends on the content.

For example, some time ago I saw the recent 4K restoration of The Bridge on the River Kwai projected in Sony 4K. The screen was rather small, but I was also sitting close to the screen, and I could definitely see the 4K difference. The amount of detail was unbelievable – closer to 70mm than any 35mm projection that I’ve seen, really. It’s truly unbelievable how much detail you can squeeze out from a good 35mm negative.

Then again, the 4K projection of Dr. Strangelove – watched in the same auditorium, even sitting a bit closer – didn’t seem that much sharper than a good 35mm projection. There’s simply a lot less detail on the original high-speed B&W negative compared to Kwai.

As mentioned before, the limits of 2K are indeed most visible in graphical stuff such as titles. (Non-anti-aliased DCI subtitles being the worst case.) The rest is a bit more subconscious: you might not consciously realize that the image you’re watching lacks resolution, but it could still make more of an impact if you saw the same thing in 4K. But just try selling that to an audience…

Posted by Kirk Futrell (Member # 5006) on 01-09-2013, 12:51 AM:

Would microperf or possibly non perf screens make a very noticeable difference with a 30' screen at 4k?

Posted by Lyle Romer (Member # 1266) on 01-09-2013, 01:00 AM:

I saw Skyfall on a 30ish foot screen in Sony 4K projection. I was about 1.5 screen heights back. In bright scenes I could see the "screen door" pixel grid. If I can see that at 4k, I'll see it better at 2k.

Posted by Steve Guttag (Member # 268) on 01-09-2013, 05:53 AM:

On a 30ish foot screen with 4Km the pixel size is going to be on the order of 1/12 of an INCH. So your eyes can see that sort of detail at 20-feet? Impressive! The perforations MUST drive you crazy because they are a lot bigger! Even Microperfs are bigger.

Perhaps this theatre is 3D capable and had the dreaded 3D lens still in place...which means not only was it not 4K but less than 2K and depending on the alignment during the show...far less.

Posted by Stephen Furley (Member # 1277) on 01-09-2013, 06:49 AM:

I've only seen 4k once, a short clip of something at the Pictureville in Bradford during the Widescreen Weekend last year. I agree with Harold, it did look better than the best 2k which I've seen. Of course, it's possible to make bad transfers at any resolution, I've seen a number of poor 2k ones.

I'm not sure of the size of the flat screen at Bradford, but probably thirty something feet.

Posted by David Zylstra (Member # 4230) on 01-09-2013, 11:58 AM:

quote: Carsten Kurz
Microperf, however, is a real problem with sound, as it cut's out more high frequency than todays tweeters can compensate.

Not really a big problem, I recently did a small 4 screen retro and I tried a micro-perf from Harkness on one screen where I could do an A/B comparison with another of the exact same size and audio equipment. The only real difference between the 2 was the frequencies the screen affected - I forget what the shift was but the level change was not significant - of course there could be other psycho acoustic affects that were not measurable via D2.

Posted by Dave Macaulay (Member # 813) on 01-09-2013, 05:41 PM:

On a reasonably small screen the difference is visible, but not astounding. It's very dependent on viewing angle, effectively the distance to the screen vs its size. On a demo with a 30' non-perf screen and perfect focus (a lot of time spent doing the scheinpflug adjustment and convergence) the pixels disappear with 4K and are noticeable but not really objectionable with 2K.
How much 4K material is really going to be released? I've been using the Barco 4K RED camera demo and a couple of 4K trailers for tests on 4K installations. The demo roll is much higher resolution (really quite impressive) than the trailers, I wonder if they are actually 2K scans converted to 4K somehow.

Posted by Sam D. Chavez (Member # 1841) on 01-09-2013, 08:05 PM:

quote: Carsten Kurz

"Microperf, however, is a real problem with sound, as it cut's out more high frequency than todays tweeters can compensate."

Interesting.

I was under the impression there was more total hole area vs standard perf. so more air and less vinyl as it were so better HF transmission. One reason I came to this conclusion is the light level dropped a significant amount.

I'm happy to stand corrected on this, but only by Guttag.

Posted by Steve Guttag (Member # 268) on 01-09-2013, 09:09 PM:

Sam,

You are WRONG...WRONG, wrong, wrong! (there happy) [Wink]

It is true that micro-perf screens generally have more HF roll off associated with them. Some manufacturers will publish specifications on that. Harkness is one such manufacturer:

http://www.harkness-screens.com/projection-surfaces-matt-plus.html

Posted by Carsten Kurz (Member # 5396) on 01-09-2013, 09:21 PM:

The hole area is more or less the same between standard and microperf, at least for the screen material I have datasheets for. But the diffraction and attentuation is of course very different between the two.

E.g. standard perf 1.25mm,57.000holes/sqm, 7% hole area
microperf 0.5mm, 300.000holes/sqm/6% hole area

Harkness has a sample attentuation graph for different perforations in their datasheet:

http://www.harkness-screens.com/pdfs/data-sheets/Perlux_140.pdf

They have two different MP screens, one actually has a higher attentuation than SP, while MPS has a lower attentuation.

- Carsten

Posted by Sam D. Chavez (Member # 1841) on 01-09-2013, 10:58 PM:

It had to happen some time, and I'm happy to have it be you to point it out.

Oh well.