Film-Tech Forum ARCHIVE: digital projection & perforations

This is topic digital projection & perforations 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=000132

Posted by Matt Hale (Member # 1518) on 10-09-2003, 04:03 PM:

So the screen went up in our new mixing theatre last night and we fired up our brand-new lcd projector (a Sanyo PLC-EF31N) and got a major interference pattern between the pixel grid of the image and the perforation grid of the screen. It looks sort of like wavy rainbow plaid. The only way to get rid of it is either zooming the image to about 75% of the screen size or defocusing the lens - neither of which is very appealing to us.
The screen is an ordinary matte white film screen with standard size perforations. My question is to those of you have digital projectors and micro-perf screens - does it get rid of the interference pattern or do more holes make it worse?

Also, my logic tells me that a micro-perf screen should attenuate the high-frequency sound from the screen speakers less than a normal perforated screen, but our installer insists its the other way around. Any experiences either way?

Thanks,

Posted by Richard Fowler (Member # 893) on 10-09-2003, 04:25 PM:

LCD projectors have more pronounced grid affect on perforated screens versus DLP / LCOS type of machines. The screen perforations, based on the hole pattern ( in addition to size )produced by the manufacturer's punching die, can also make this more pronounced. Micro perf is better but they do require EQ changes for audio due to the less net space for the holes. Microperf, if not a one piece sheet, will have a tendency to have the seam joints more noticable compared to standard perf. You should have demoed different type of projectors and screen samples. SXGA resolution or better is the why to go to future proof yourself.

Posted by Ian Price (Member # 14) on 10-09-2003, 06:21 PM:

I have this exact same projector. We have a 14' high Hurley screen with standard perforations. I have not noticed this "grid interference." But so far we have only been running NTSC video off of a Sony Betamax SP deck and one DVD. Running in this way, the pixels are much larger than the perforations.

Posted by John Pytlak (Member # 331) on 10-09-2003, 06:57 PM:

Whenever two regular patterns (the projector pixels and the holes in the screen) have about the same frequency, they can alternately reinforce and interfere with each other, causing moire. You either need to change the frequency of the pattern (change magnification, increase the number of pixels) or have less of a pattern (other types of projector, defocusing, use smaller holes) to make the pattern less noticeable. The screen manufacturers have been working to develop screen perforations that minimize aliasing patterns with digital projectors, while not significantly attenuating the sound or requiring extreme equilization, but ultimately, better projector technology is being developed.

Interference can also occur with regular patterns in the image itself (the classic moire seen in herringbone suits or venetian blinds on television or digital displays). In this case, anti-aliasing filtration is used to attenuate the offending image detail, at the expense of at least 1/2 of the resolution (Nyquist criteria):

Berkeley Moire

Rice University

CMU Aliasing

Stanford Lecture

Posted by Evans A Criswell (Member # 381) on 10-09-2003, 07:03 PM:

One thing that would help is if screen manufacturers would, instead of using a regular lattice, would use a random distribution of perforations. The lack of a rectangular or haxagonal (or other regular) pattern in the holes would eliminate the beat pattern formed by the pixel pattern of the projector interacting with the regular lattice of perforations on the screen.

Posted by Richard Fowler (Member # 893) on 10-09-2003, 07:19 PM:

The random aspect for the perforations are limited due to the fact since most screens are assembled from panels that are seamed together. There is matching involved so as not to see a obvious repetition. The random pattern is also limited to keep the gain reflectance even on the surface. Years ago there was one screen manufacturer which used a random pattern that was so broad that at about 4 meters or less the pattern was visible while watching the image.

Posted by Gordon McLeod (Member # 33) on 10-09-2003, 07:31 PM:

Also a random patern might interfere with consistent sound spread and attinuation

Posted by Evans A Criswell (Member # 381) on 10-10-2003, 10:45 AM:

Right. There's no easy solution. The reason that the screen perforations haven't caused too much of a problem for film presentation is due to the random distribution of the grains in the film media. I can only notice screen perforations if I'm less than 20 feet from the screen and the focus is very sharp when I'm watching film. The perfs give the appearance of watching an array of pixels.

The interaction between digital projection and screen perforations should only be a big problem if the spacings of the pixels of the image and the spacing of the perforations on the screen are similar. For example, let's say your digital projector has 1280 pixels across the image and you're showing a movie on a screen 24 feet in width. That's one pixel per 0.225 inch. Suppose your screen has perforations every 0.2 inch. Then you have a beat pattern that is the difference of the frequencies of these two patterns: 1440 cycles per screen width for perforations minus 1280 cycles per screen width for the image pixels, which is 160 cycles per screen width, or a moire pattern that will have 1 "wave" every 1.8 inches, which may be very noticeable.

Edit: For those of you who notice interference patterns between your projector and screen, please post your screen perforation pitch, screen dimensions, and projector resolutions. This is interesting.

Posted by Paul Mayer (Member # 355) on 10-10-2003, 12:23 PM:

Here's a thread from last year on the same subject for those that want to read further:

Digital projection and perforated screens

Posted by Frank Angel (Member # 248) on 10-10-2003, 01:43 PM:

Sure there's an easy solution, Evans, USE FILM! (Sorry, couldn't resist). [Big Grin]

Frank

Posted by Evans A Criswell (Member # 381) on 10-10-2003, 01:56 PM:

Wow. I'm beginning to see what being 35 can do for my memory. That thread that Paul is referring to last year is one that I started, and I had no recollection of it.

Posted by Matt Hale (Member # 1518) on 10-10-2003, 03:01 PM:

Oops... I guess I should have done a search of old threads before posting.

Anyhow, our installer brought a sample of microperf by for us to try and if anything it made it worse. Strange thing is the micro-perfs were smaller and denser by at least a factor of 10 than the regular screen, so I guess we just have great misfortune in hitting pixel/perf harmonics. Our old theatre had the same problem too, with the same resolution projector but a 10 foot bigger screen.

I wonder if throw distance (or perhaps throw:screen ratio) has anything to do with it. The other theatre at our old facility had a 30 foot longer throw, but only a slightly larger screen, and never showed a hint of a problem. As far as I know the screen material (and therefore perf density) was exactly the same.

After much fiddling we have reduced the pattern on our new setup from wavy rainbow plaid to a mild horizontal bar thats really only visible in large areas of solid colour, but the image is just a little imperfect.

In other words its not film...