Film-Tech Forum ARCHIVE: NEC 2000C Moire Pattern

This is topic NEC 2000C Moire Pattern - Possible Causes? 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=002166

Posted by Jim Cassedy (Member # 4115) on 01-04-2015, 04:22 PM:

I'm doing a preview later tonight at one of the screening rooms where I
sometimes work. "The other guy" just informed me that the bulb on their
NEC 2000C projector was recently changed, by a professional service tech
and that ever since then, hs's noticed a slight moire pattern on screen
that he's never seen before. (And I've worked there many times and if
the moire was there previously, I know I would have seen it.)

The otherprojectionguy sez that nothing else was changed or adjusted
(other than the normal bulb-change-alignment) when the new bulb went in.

I said I'd go down early & takalook, but quite frankly, I'm not sure
exactly what I'm looking for. Does anybody have any ideas?

It's an NEC 2000C- - and I beleive they have a 2K bulb inside, since
it's only a screening room with a screen around 12-15ft wide, if I recall.

He sez he's seen it on both flat and scope. The projectorthingy is
equipped with Dolby 3D, which hasn't been used since the bulb change.
(and the wheel should be well outta the light path & probably OK)

I'm heading down there shortly, so if anybody has any ideas, please chime in!

THANKS! -Jim C-

Posted by Mark Gulbrandsen (Member # 72) on 01-04-2015, 07:55 PM:

The moire pattern is generated when the pixels and grid of the DMD's in the projector are at or near the same pitch of the screen perfs. Some times this can be minimized by getting the projector perfectly level with the perfs and or making the image slightly larger or smaller if any of this is possible and some times it can't be gotten rid of without changing the screen. Newer screens from Harkness and Severtson take this into account with modified perfs and there is less a chance of the moire.

Mark

Posted by Jim Cassedy (Member # 4115) on 01-05-2015, 12:33 AM:

So, Mark- There's never been a moire pattern at this location before.

I did notice that the projector has been moved very slightly from its'
previous position, and that might have been enough to cause the
moire to appear, from what you are telling me.

I didn't see your message till I got home, but in the course of checking
things, I did move the lens focus & zoom slightly and that seemed to
clear up 98% of the moire. When I first got there, you could see it all
over the screen, especially if you put up a peak or 50% white field.

After 'fiddling' with the lens settings a bit the moire is now confined
to one small spot in the lower left hand corner on a 'white screen'
and it's totally invisible during actual content.

I'll go back later this week when I have time & play with the projector
position & check the level, but my time was limited today.

As I mentioned to you once before, I still really want to go to the
"NEC Projector School" for some advanced operator training or
something like that.

Barco & Christie Tech's are a dime-a-dozen here in SF, but NEC techs,
not so much so. And I regularly deal with over half-a-dozen NEC's, so
it would be worth my time.

I don't care what the school costs- - that's not a problem. I'm still
trying to find a hole in my schedule big enough for me to attend,
and I'm still unsure exactly how to get into one.

Thanks for the moire advice

Posted by Marcel Birgelen (Member # 6801) on 01-05-2015, 01:50 AM:

The moire pattern/zebra stripe issue is present across all brands of DCI projector manufacturers. I guess it would help if screen manufacturers would create a slightly "random" perforation pattern for new screens.

Posted by Carsten Kurz (Member # 5396) on 01-05-2015, 07:16 AM:

It may be that the moire has not been noticed before. Maybe someone carefully adjusted focus so that is was just not visible. And maybe this focus was now touched, even if they claim it wasn't. I guess it is also possible to shift a bit of focus with the bulb alignment. So, I would check the pattern, and if it is indeed visible, readjust the focus slightly for all formats to get rid of it again.

- Carsten

Posted by Jim Cassedy (Member # 4115) on 01-05-2015, 10:39 AM:

quote: Carsten Kurz
It may be that the moire has not been noticed before.

Oh ho ho ho! Not very likely. Not only would I have noticed it on one of
the many occasions I've worked that room, but one day last summer I
spent 18hrs in that room working with a group of optical engineers who
were contracted by a consortium of several of the large theater chains to
evaluate different types of 3D glasses provided by several manufacturers.

The day before the test a complete alignment& re-bulbing of the projector
was done. On the day of the test they hauled in two huge road cases of
scientific equipment, and spent the first several hours just calibrating
everything. They also had two of those incredibly expensive
(around $12,000 us)screen meter-things that they used to establish
'basline' data for their tests, and also to take readings every two hours
to certify that the light levels were not 'drifting' any during the course
of the day which would have affected their test results if they didn't
factor those corrections (if any) into the data they were collecting.

95% of the work they were doing involved 'shooting white light'
on the screen.

Believe me- - if there had been a moire or ANYTHING wrong with the system
that day, they would have stopped everything.

I had to sign a non-disclosure agreement, so I couldn't take any pictures
and exactly what they were doing and who they were specifically doing it
for is still proprietary information. It will eventually be released
in some form, and they promised they'd let me know when that happens.
A professional photographer took a sh_tload of photos of the set up
that day, as part documenting the way they collected the data.

I hope I can share those once the info is published.

(It was a long, but easy day for me. - I got paid a wholebuncha
money
to basically just stay out of their way all day. Nice!)

Posted by Steve Guttag (Member # 268) on 01-05-2015, 11:38 AM:

That said...I can GUARANTEE that there was Moire. There is no way around it with a fixed grid screen and fixed grid projector. All you can do is minimize it. Zooming the lens can often push the grids so they are less objectionable. This is one of those scientific facts you just can't dance around...two superimposed grids (regular grids)...will Moire. What you are left with is how apparent it will be based on how the two grids "beat" off each other.

Posted by Ken Lackner (Member # 1002) on 01-05-2015, 12:29 PM:

Um. Maybe you should rephrase that, Steve. [evil]

Posted by Marco Giustini (Member # 4544) on 01-05-2015, 01:00 PM:

Jim,

Forget about the lamp.
Put a white pattern on screen, de-focus the picture and I bet the moire will disappear.

Posted by Mark Gulbrandsen (Member # 72) on 01-05-2015, 01:07 PM:

I agree that it's always been there to some extent. It is a fact of DLPee projection. The fact that your projector was moved from it's normal position may have exacerbated it. Throw up the alignment test pattern and be sure the projector is where it belongs on screen. Once you verify that then tweak zoom and focus slightly to minimize it.

Mark

Posted by Jarod Reddig (Member # 6363) on 01-05-2015, 01:34 PM:

It is indeed common knowledge that morie is the interaction between the screen perf pattern and the DMD grid. One way to get around it is to have the screen cut at an angle at the factory if its an option. Or as you posted adjust fine focus to minimize.

Posted by Marcel Birgelen (Member # 6801) on 01-05-2015, 04:59 PM:

I still don't get why the likes of Harkness, who now push out almost exclusively "digitally optimized" screens aren't able to deliver "randomized" perforations. This would get rid of the problem for once and for all. Is it because it would affect the tension strength of the screen rather negatively or am I missing something else?

quote: Jim Cassedy
Believe me- - if there had been a moire or ANYTHING wrong with the system that day, they would have stopped everything.

On the other hand, even if there was some slight moire, it shouldn't really have affected their measurements.

Posted by Jarod Reddig (Member # 6363) on 01-06-2015, 12:04 AM:

It probably has to do with how sound would interact with a non-uniformed perf pattern but its just a guess.

Posted by Steve Guttag (Member # 268) on 01-06-2015, 05:59 AM:

At best it would be a semi-random pattern. The way most screens are made by seaming it together, one has to be careful about having the pattern not mate up well when seamed. Then there is the problem of an irregular yet periodic pattern is actually more visible than a regular one. Your eye isn't going to fixate on constant/uniform...but it WILL notice something that is different.

Stewart Filmscreen effectively has an irregular pattern and they don't normally show Moire very much (must less than other screens). However they get their random perforations due to human error. Their screens are seamless and the perforations are applied after manufacture...as a result they tend to weave a bit. However, sometimes they weave into each other and you get some overlapping perfs. This can be every bit as visible as a seam. Again, your eye will pick off differences more than the same. Note, Stewart also rotates their perf pattern to not be on the same grid as a digital projector to even further minimize Moire.

Overall, I still like Stewart the best but they just aren't worth the 4-5 times the expense in a typical movie theatre...in a screening room, they can sure make a movie "pop" like no other screen I've seen.

Posted by Marcel Birgelen (Member # 6801) on 01-06-2015, 04:59 PM:

I've always asked myself how they get the perforations in there. I imagine it like them using something like big rollers with the perforation patterns on them. In this case, it will be impossible to create a truly random pattern with this technique.

You definitely want to avoid something like a wallpaper pattern occurring, but even if the perforations are semi-random and the repetitive pattern is sufficiently spaced out (like only occurring like two times on a certain screen) it should still work I guess? Because I doubt the average brain would be able to correlate this data into something of a repetitive pattern.

Posted by Carsten Kurz (Member # 5396) on 01-06-2015, 07:58 PM:

I guess they could use laser cutting for the perforations nowadays, but it would probably be more expensive than ordinary mechanical perforating.

The random pattern should be not such a big problem. There is a lot of 'classic' theory available, e.g. http://en.wikipedia.org/wiki/Dither

- Carsten

Posted by Marcel Birgelen (Member # 6801) on 01-07-2015, 04:11 AM:

Laser cutting sounds like a plausible solution, but wouldn't laser cutting create burned edges around the perforations?

Posted by Carsten Kurz (Member # 5396) on 01-07-2015, 04:30 AM:

No, laser cutting can produce very clean edges, surprisingly clean.

BTW - we have a wall covering in our old cinema made from PVC in the fifties. It has about the same perforation pattern as our screen. They could do these perfs perfectly even back then. I guess all 'roller-type' perforaters would produce visible bands when using random patterns.

Heck, the best way to do it is to hang an unperforated screen, then switch on your laser projector and play a 4k dither_perf DCP ;-)

- Carsten

Posted by Mark Gulbrandsen (Member # 72) on 01-07-2015, 01:29 PM:

Some screen manufacturers have already modified their perf pattern for DLPee. Severtson is one of them. I have always had way less moire on a Severtson screen then on any other surface.

Mark

Posted by Carsten Kurz (Member # 5396) on 01-08-2015, 10:12 AM:

Any info in how far their perforation differs? Is it more random, or applied at an angle?

Wondering wether you could actually hang any 'straight' perf screen slightly angled and then let the masking take care of it? You loose some screen estate, but then this standard screen material is far less costly than those optimized versions. Wondering how much of an angle is needed to reduce moire noticeably?

- Carsten