Film-Tech Forum ARCHIVE: Keystoning?

This is topic Keystoning? in forum Film Handlers' 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=1;t=006658

Posted by Richard Greco (Member # 2032) on 10-24-2004, 11:39 PM:

What is it? How does it affect a digital image? Any places on the web where I can learn about digital projection and keystoning? Thanks Guys

Posted by Daryl C. W. O'Shea (Member # 1303) on 10-25-2004, 12:26 AM:

Would you settle for film projection and keystoning?

Or how about just keystoning?

Google, define:keystoning

Posted by Brian Guckian (Member # 1678) on 10-25-2004, 07:05 PM:

The placement criteria for digital should be just the same as for a film projector. Remember that effectively, the lens is looking at a DMD panel, LCOS array, or whatever instead of the film frame.

Also, don't rely too much on optical lens shift or electronic image distortion to correct placement errors. In the case of the former, there is a limit to how much adjustment is available. The latter should be avoided as much as possible as it introduces artifacts into the image.

One convenient feature of most digital projectors is a precise masking control, which acts in the same manner as an aperture plate for disguising angled sides of the image where the rake or the horizontal offset is noticeable.

However, you still must get as close as possible to the centre line of the screen in both planes. [Smile]

Posted by Dominic Case (Member # 1830) on 10-26-2004, 02:18 AM:

SMPTE's engineering guideline EG 18 gives tolerances for keystoning, recommending no more than 3%, but allowing that up to 5% is generally OK. It also gives formulae to work out what % keystoning you will get for a certain distance off-axis.

SMPTE members can get these standards and guidelines, although I note that EG18 has been officially withdrawn (but the keystoning principles are still valid).

Not sure if the withdrawal relates to the keystoning issue, as digital projection does indeed have technology to compensate up to a point - or more likely, it it's withdrawn in the context of equal access legislation regarding line-of-sight and angle of view etc for disabled patrons.

Posted by Leo Enticknap (Member # 534) on 10-26-2004, 02:56 AM:

That's one of the few good things about digital projection, IMHO. Even the little £3k Proxima LCD projector we use for portable shows in venues without film projection has a keystone correct facility which, at quite severe rakes and/or horizontal angles, seems to work as if by magic.

With film, of course, there's absolutely bugger all you can do to actually correct the distortion if the architectural constraints of the building mean you have to project from a nasty angle - a very common problem with the big 2,000 seat cinemas that were twinned and tripled in the 1970s in Britain. A common cheat is to cut a skewed aperture plate which compensates for the angle of the keystone and makes the edges of the picture look straight relative to the masking - but of course the picture itself is still distorted.

So this SMPTE guideline sounds to me like something which would simply be impossible to comply with in certain auditoria. Maybe they're working on something which takes account of architectural limitations.

Posted by Carl Martin (Member # 1146) on 10-26-2004, 03:12 AM:

the guideline shouldn't be monkeyed with for that reason. keystoning doesn't look any better just because it's unavoidable. the guideline should set out what's acceptable. if it can be met, great. if it can't, it can't.

quote: Dominic Case
SMPTE's engineering guideline EG 18 gives tolerances for keystoning, recommending no more than 3%, but allowing that up to 5% is generally OK.

how is percentage keystoning figured? is it the additional width at the spread end compared to the width at the near end? i'd like to calculate a figure for our main auditorium. i know it's very unacceptable, but i want a number for how very.

carl

Posted by John Pytlak (Member # 331) on 10-26-2004, 05:28 AM:

The free download of Schneider Optics Theatre Design Pro software allows prediction of any keystone distortion:

http://www.schneideroptics.com/software/theatre_design_pro/

quote:
Theatre Design ProTM (TDP) is a sophisticated software package useful to designers of Motion Picture Theatres and other facilities utilizing projection of various types. TDP calculates picture size, lens focal length and image shape due to keystone distortion and screen curvature

Posted by Jim Spohn (Member # 1547) on 10-26-2004, 09:50 AM:

Richard, the people responcible for policing theatres with excessive keystone are....."the keystone cops" so watch out they know where you are!

Posted by Dominic Case (Member # 1830) on 10-26-2004, 08:44 PM:

quote: Leo Enticknap
So this SMPTE guideline sounds to me like something which would simply be impossible to comply with in certain auditoria.

The guideline is titled "Design of Effective Cine Theatres". It's meant to be used when the theater is being built as well as fitted out. It also covers things like line-of-sight, staggered rows of seating (looking between two heads in front of you instead of through one), angle of view (pity the patron at the end of the front row) etc.

quote: Carl Martin
how is percentage keystoning figured?

The guideline lists a number of trigonometric formulae to work it out - from almost any starting point. Too hard to type it all in here. And far too hard to understand! Looks like the Schneider download that John "Google" Pytlak found is the way to go.

Posted by Carl Martin (Member # 1146) on 10-27-2004, 02:12 AM:

i think you think i'm asking a more complex question than i really am. i'm not asking to calculate the keystoning given projection angle and throw and all that. i'm asking, given a trapezoid on the screen, how do i measure it and come up with a percentage?

i'm not an smtpe member so i can't check the guideline. the schneider program will come in handy if ever i do need to answer the more complex question.

thanks,
carl

Posted by Dominic Case (Member # 1830) on 10-27-2004, 02:38 AM:

quote: Carl Martin
given a trapezoid on the screen, how do i measure it

I think the percentage distortion is simply the ratio of the difference between the two sides, or the top and bottom. So if your rectangular aperture projects on the screen at 20 ft at the top and 21 ft at the bottom, the vertical distortion is (21-20)/20 x 100% that is, 5%.

The SMPTE formulae are for predicting the problem before building the theatre, based on angles.

Posted by Steve Guttag (Member # 268) on 10-31-2004, 08:43 PM:

Dominic, there is nothing in the document formerly known as EG-18 that was out of date or non-factual.

Steve

Posted by John Pytlak (Member # 331) on 11-01-2004, 09:07 AM:

Engineering Guideline SMPTE EG 18 listed 15 references for theatre design parameters. Much of EG 18 was based on the paper "Some Comments on the Design of Large-Screen Motion-Picture Theatres" by Will Szabo, published in the March 1976 SMPTE Journal, Volume 85, pages 159-163.

Posted by Dominic Case (Member # 1830) on 11-01-2004, 04:58 PM:

quote: Steve Guttag
Dominic, there is nothing in the document formerly known as EG-18 that was out of date or non-factual.

I agree. And of course it's just a guideline, not a standard. However, I wouldn't want to be recommending something in SMPTE's name that they themselves have withdrawn.

from http://www.smpte.org/engineering_committees/
NOTE: EG 18-1994, Design of Effective Cine Theaters, has been withdrawn by action of the Committee on Theatrical Projection, P3, at the meeting on 2003-03-05.

Posted by Steve Guttag (Member # 268) on 11-01-2004, 07:04 PM:

Dominic,

I can certainly understand that. Then again, I still stand behind EG-18 as the best single document that explains how to design a movie theatre properly for picture and, to a lesser degree, sound.

I still use the document since it is still correct and will properly guide architects to design theatres properly. Without it, there is really nothing left to use that is based on studies and facts. I've never had a theatre with sight line problems or other complaints that followed it, stadium seated or slope floored.

I defy anyone to show any part of this document that is incorrect...and as you say...it was just a guideline, the lowest of the documents that the SMPTE offers. Upon its withdrawal, I thought a lot less of the SMPTE and those that work within it.

Steve

Posted by Brian Guckian (Member # 1678) on 11-02-2004, 07:15 PM:

quote: Dominic Case
it's withdrawn in the context of equal access legislation regarding line-of-sight and angle of view etc for disabled patrons.

My understanding is that this was one of the reasons for withdrawal. I would imagine that a big problem these days is getting people who can commit time to update these documents. Certainly the engineering and architectural skills are out there to do so...look no further than the outstanding contributors to this very website for instance.

I'm sure EG-18 could be reinstated in an updated, expanded version (which fully incorporates the very valuable information in the most recent version) - perhaps it's the logistics of doing so that could be the problem. I'm just surmising.

Posted by Steve Guttag (Member # 268) on 11-03-2004, 12:21 AM:

There are members of the SMPTE P-3 committee on this forum, John P and myself included. There is indeed discussion on a new version of EG-18. The problem is that the only way that I can see a new version could come out is to remove information, not necessarily add to it. Nothing in it has ever been shown to be wrong; in fact, just about everything in it was the result of testing or otherwise factual information. In fact, it was a bit too good, if you get my meaning.

Steve

Posted by Daryl C. W. O'Shea (Member # 1303) on 11-03-2004, 12:26 AM:

Are there any other standing standards, practices, or guidelines concerning building design? By the sounds of things, EG-18 was withdrawn solely out of fear of liability, ie. "the SMPTE suggested we build our auditoriums this way, it's their fault".