How Does a Tilted Screen Stay in Focus and Preserve Geometry?

[Scott Jentsch]

The Warehouse Cinemas in Frederick, Maryland opened on Friday, September 4, in the space that was formerly the Hoyts Frederick Towne Mall 10, which closed in 2011.

It appears they kept the sloped floors instead of converting to stadium seating, and elevated and tilted the top of the screen toward the viewers. Here is a photo that they provide on their website:

skyvue.jpg

I'm sure the screen image is a Photoshop addition, but in a news article, they are quoted as saying that the screen is tilted at a 12 degree angle to provide better sight lines.

I haven't done any calculations to see how far forward the top of a 20 or 30 foot high screen might be compared to the bottom when tilted 12 degrees, but it would seem to me that this would cause focus and geometry issues.

Has anyone seen this before? Is it as bad an idea as it seems like it might be?

[Randy Stankey]

Not saying how this is done. Just theorizing that it is possible.

Have you ever seen a view camera, like a 4x5, that has movable lens board and film holder?
You can tilt and rotate the lens board and the film holder such that they are not parallel. The idea is to allow you to correct for perspective issues. For instance, when you are taking a photograph of a tall building and you want to cancel out converging or diverging lines of perspective. If you are standing on the sidewalk, looking up at a tall skyscraper, the top of the building looks narrower than they bottom even though we know that the building is actually straight. Using a view camera, you can tilt the lens board and film plane so that the building appears to be straight in the resulting photographic image.

Theoretically, this could be done in a projector. After all, what is a projector but a reverse camera? Right?
If you tilt the lens in such a way that the image is keystoned in the opposite way, you can make it appear like it is straight.

There's a bit of math involved in figuring it out but it's mostly just basic trigonometry.

Look up "Scheimpflug Principle" for more info: https://en.wikipedia.org/wiki/Scheimpflug_principle

Maybe this isn't how it's actually done but it could be done this way.

I suppose some sort of electronic chicanery could be involved, as well.

[Steve Guttag]

Being the one that installed that theatre...We tilted the projectors UP and then lens shifted down (we also located the projectors on the booth floor so that they started as low as possible). So the relative tilt of the screen was no more than 5-degrees (it all depended on what the available lens shift could accommodate with preference being given to focus over keystone). Most cinemas have a 5-10 degree projection angle down and can only compensate for about 5-degrees with lens shift (and didn't at all for film unless they used PC-Cine adapters). So the keystone there is no worse than any other typical site.

As for the screens leaning...they aren't bad, I can tell you that. I'd rather have a sloped floor to one with risers and without sight-line issues.

And yes, that is an "Artist's Conception" photo...none of it is real though the real thing is pretty close to the artist's conception. If you go to their Facebook page, there are enough photos there to give you a feel for how it came out (pretty darn good).

They are all Barco 4K laser projectors too (SP4K-C series). All QSC Q-SYS audio, all QSC SC-424 4-way stage speakers. So even the non-ATMOS theatres have pretty impressive systems.

Randy, Scheimpflug is indeed how one can get uniform focus though it doesn't help you on keystone. Lens shift will do that but it does mess with linearity, which most people won't notice. I didn't take the picture below but that's me in the front row as I was adjusting the centering/focus of the lens.

[Frank Cox]

And they have an actual scope-sized screen, too!

I really don't like the idea of having a flat-sized screen and making Cinemascope smaller than Widescreen. Call me a curmudgeon...

[Scott Jentsch]

I'm glad to hear that the end result is a good one! I've always been bothered by the fact that the trend to widely spaced rows of recliners has caused the best seats in the house to be lost. You can choose to be too far away, or too low relative to the screen, because there is a walkway where the best seating would be, or the designers chose to split the difference with two rows. I don't care to be eye-level with the bottom of the screen.

Maybe this type of installation will catch on, especially in houses that are still sloped and haven't been ripped up for stadium seating yet...

If you have any photos that we could use, I'd love to put them on the Photo Gallery page for the theater:

https://www.bigscreen.com/Marquee.ph...26&view=photos

We will give you credit for the photos (including copyright), with any captions you wish to add.

[Steve Guttag]

Frank, only the three larger auditoriums are scope screens. The rest are 1.85. But that is due to the shape of the rooms.

[Marcel Birgelen]

Interesting concept. It's nothing new to tilt the screen slightly forward in traditional sloped auditoriums, but I've not seen it done at this kind of steep angle. The only thing I might be a bit worried about is, how much the screen is going to sag after a few years of use.

The other question I have is: Did they build a baffle wall behind the screen? And is it also at an angle?

[Randy Stankey]

Being the one that installed that theatre...We tilted the projectors UP and then lens shifted down (we also located the projectors on the booth floor so that they started as low as possible)...

...

Randy, Scheimpflug is indeed how one can get uniform focus though it doesn't help you on keystone.

Cool! Thanks! Good to know!

I saw the picture, considered the setup like a puzzle to be reverse engineered in my mind then thought about what I'd do if I was trying to take a photograph under comparable conditions...a projector being a "reverse camera." It's not a 100% correct model but I thought it would be close enough.

If I was taking a photo under similar conditions, I probably would have angled the camera and tilted the lens and/or film plane to get the solution I needed. Although I didn't specifically think about dropping the projector nearer to the floor, I guess it would have been part of the equation if I was solving the problem in real life.

The things I didn't know were whether the image plane was movable and how the lens could be shifted.

Moving the image plane probably wouldn't be a good idea in a cinema projector. Even if it could be done, I wouldn't really want to do it because it would introduce a lot of other problems into the mix that I wouldn't want to deal with.

Shifting the lens would have been my first thought but I didn't know how that would be done. It could be some adjustments built into the projector or it could be some kind of attachment like a "LensBaby." Built-in adjustments would have been my first thought but just how much shift is available is unknown to me.

I like solving this kind of puzzle with my view camera. (A Cambo monorail style camera.) If I do a little studying and some practice, I can get to the point where I can do most of the calculations in my head. Maybe, I'll have to do a little bit of math on the back of a pizza box to make sure but 90% of it is just 9th Grade algebra.

Just a bit of Kentucky Windage. Right?

I wish I had more time to do things like this instead of sitting in a shop, hooking up wires and junk. This is the kind of thing I enjoy doing and piecework is...boring!

Nice work! Maybe, some day, I'd like to come see it.

[Steve Guttag]

Scott,

By and large, I leave it to the owners of the theatre to release whatever photos they want presented to the public. I did contribute to the warehouse for the AFI/Silver but did get explicit permission for the director of the AFI/Silver to do so. Also, most of my personal photos of a site are MUCH more booth related since, from a support standpoint, they can be more helpful in directing someone via phone since I (or another tech) can essentially see what they are seeing. It is rare that I take any really decent photos of the auditorium beyond standing in the middle and snapping a quick shot of all four walls. In this space, the Dolby ATMOS house was an exception since they are required for Dolby ATMOS certification. The photo I posted above was from the Warehouse Cinema's Facebook page so it was already public.

Randy, another piece of the puzzle is also telling the screen/frame people how much longer/taller to make the screen since now you are on the hypotenuse of the frame. So a little lite-trig is required and backed up with a CAD drawing. Then again, doing curved screens is more tricky since you have to provide arc-length (s = rθ where θ is in radians) and then also consider how large the image is in the center as well as the edges. My initial presumptions were that I could get 5-degrees of projector tilt and still compensate with lens shift and I was pretty close to that (a little better actually). Lens shift is tricky as every lens model is different. One thing we could have potentially done is gone to the "B" series lens mount as that allows for the larger lenses, which have a larger inlet pupil and consequently more lens shift. Then again, nobody is going to to notice the keystone in these theatres any more than they do in any other theatre. A down side to lens shift is that you are putting the image off center on the lens and you do get some geometric distortions there (curved lines) but again, how bad it is is really lens model dependent.

[Randy Stankey]

Randy, another piece of the puzzle is also telling the screen/frame people how much longer/taller to make the screen since now you are on the hypotenuse of the frame. So a little lite-trig is required and backed up with a CAD drawing...

...A down side to lens shift is that you are putting the image off center on the lens and you do get some geometric distortions there (curved lines) but again, how bad it is is really lens model dependent.

Sure! Just tell them that the screen's shape is a trapezoid defined by the intersection of a rectangular plane and a cube at given relative orientations to each other.

Then wait for the "deer caught in headlights" look to appear on their faces.

I also thought that there might be some barrel distortion caused by shifting the lens but, like I said when I was referring to the Scheimplfug principle and moving the camera/projector around in conceptual space, I knew that there would be some effect but couldn't explain the specifics without actually working through the problem.

When I mentioned the possibility of some sort of electronic chicanery, I was thinking about the idea of using some sort of image processing to counter-distort the image inside the projector to cancel out the barrel distortion effect. Although I am sure that it's theoretically possible, I don't think it would be practical. It would be computationally and logistically cumbersome, plus it would probably cause loss of image quality.

It's probably better to just let the proverbial shirt tails hang down than to waste time trying to fix something that probably doesn't need to be fixed.

Like you say, 90% of the people probably wouldn't notice. The other 10% don't keep a straight edged ruler in their back pockets when they go to the movies, anyhow.

[Sean McKinnon]

That really interesting Steve. Was this screen tilt thing an Architect, or owner idea? Do you feel it is an improvement overall over a sloped floor and "normal" sightlines? Did it allow you to put in larger screens?

[Steve Guttag]

It was definitely Owner driven. I do feel it improves the sightlines on a sloped floor. Screen size was dictated by the theatre sizes and exits.

[Marcel Birgelen]

The thing about the baffle wall and speaker placement is a trade secret?