[Matthew Todd]

I've decided that I'd like to use a constant area projection scheme in my future dedicated HT (like Bjoern Roy's -- see these threads)
http://www.avsforum.com/ubb/Forum4/HTML/002498.html
http://www.hometheaterforum.com/uub/...ML/036131.html

I've been thinking some more about it and would like to use a curved screen like the torus if that is possible. The reason for this is that I could go slightly larger with the screen in this case. I'm going to be using either a Sony G90 or a Marquee 9500LC 9" CRT projector and was planning to use a 96" wide image for 16:9 stuff (before I thought of using a curved screen). Wider aspect ratios would have the same area, and therefore would be wider. Taller aspect ratios would have the same area and would therefore be narrower but taller. A 96" wide 16:9 aspect ratio image would be 8 feet wide. I'm thinking that if I can get away with curving the screen, then I will be able to get away with higher gain without the hotspotting/color shifting. Maybe I could go to 9 or 10 feet wide with 16:9 material.


I'd like to have the screen perforated so that I can place the front speakers (all identical) behind the screen.


Has anyone ever thought about doing something like this? Is such a thing possible?


Roughly, what would such a curved screen cost? Would the Torus be a fit for these requirements? What about the possibility of making one myself?


Here's a picture of a Torus (I don't want one quite that big). http://www.screens-usa.com/stewart/torus.html


Thanks for any help!


Matt

 

[KFung]

Yeah, I gotta say, constant area is a nice balance. You get the increased horizontal size of a 2.35:1 image without making your screen a lot larger, the only downside is that 4-way masking can be pretty costly. Possibly even more than a larger size 2-way masked constant height solution that many use.


Anyways, as it looks like you're going top-of-the-line price is not such a large concern.


That being said, a TORUS is really rather expensive and perhaps a bit of overkill at this time. It's definitely a special order (with a screen cost in the 4-5 figure range) plus the consulting costs to get the design optimized for your theatre (on the order of several thousand dollars). Trust me I asked. http://www.avsforum.com/ubb/biggrin.gif Of course if you're going all-out it's unbeatable for a CRT on a large screen, it's much better (and cheaper!) than double stacking a projector for higher brightness. Unfortunately, the screen can't be perforated because it uses a vacuum to hold the screen in its curve. Maybe someone will find a way to make a rigid TORUS that will accept perforations, like Cineramax dreams, but I don't think it will come along in your time frame. http://www.avsforum.com/ubb/frown.gif


If you really have a desire for a very large screen using a CRT (you might consider something like the JVC G-15) and can fit the center speaker beneath the screen (or possibly on the top and bottom for better center channel imaging) the TORUS is a great solution. I think you might find that you are able to go higher than 10' wide, perhaps as far as 11'.


Regards,


Kam Fung

 

[REW]

Kam,

Good points-but what the world needs now is a low cost DIY curved screen(nearly Torus)Peter has blazed this path and several (Alan)are trying.Lets get to work on this.Enough talk!

Lets have at least a separate topic to slay this dragon.

PS This would put Kam in CRT camp-sooner or later!
http://www.avsforum.com/ubb/smile.gif



------------------

Ron

"Your priorities will be different-its the weighting that counts!"

 

[Matthew Todd]

Thank you for the responses Kam and Ron.


I really am cost conscious, it's just that I want to get the best video that I can afford, so I'm going to splurge there and get a G90 or 9500LC.


It sounds like maybe building a curved screen myself may be the best solution until I have the dough to really splurge on a torus or its equivalent (although I really want to be able to place the front 3 channels behind the screen).


I've seen the thread about Alan's curved screen project.


Has anyone experimented with a DIY microperf curved screen, either just curved in one direction or a compound curve. How much of the benefit of a curved screen do you get if you only curve in the horizontal axis and not the vertical also?


If it comes down to it, I'd probably choose a flat microperf screen with constant area projection (that's what Bjoern Roy has I believe) over a curved screen with no perforations. I'm just trying to find a way to have both!


Matt


[This message has been edited by Matthew Todd (edited 05-30-2001).]

 

[KFung]

I'd have to say that DIY'ing a curved screen is risky if you don't understand the math behind how the TORUS works, you could easily end up with a screen that has too high of a gain for the curvature or vice-versa, or a vertical curvature that is optimized for a different gain than the horizontal. There is a reason why Sigma Design Group charges so much for their design fee (although I would argue that it's probably too much, they're used to larger projects I think). That's shouldn't discourage people from trying, but make sure you do your homework first! In any case it shouldn't look too bad no matter what you do, but it probably won't come close to a truly optimized TORUS. Just remember, a curved screen only provides a benefit if you use a moderate to high gain surface (even if it has a grey backing).


Matt,


You won't be disappointed with either of those projectors, they really are amazing! I only wish they weren't $10-15k outside of my budget!


Anyways, if you only want to curve in one direction (I would recommend a horizontal curve) you don't need to go the DIY route. I know Stewart can curve one of their fixed screens in one dimension and you can get microperf as well! Actually, many modern theatres go this route (although I've heard the curve is to improve focus on their short-throw lenses). You could take the gain up to 2.0 - 2.5 (that's just a guess, I don't know what the gain characteristics are for screen surfaces with such a high gain) and go for that 10' screen! It should give you somewhere in the neighborhood of 8 ft-lamberts with minimal hotspotting.


Regards,


Kam Fung

 

[Matthew Todd]

Thanks again for taking the time to reply Kam!


If I did go with a Stewart microperfed horizontally curved screen (I've seen these large ones in the local cineplex -- everytime I'm there it seems like I have this uncontrollable urge to go up and look closely at them) with a gain of say 1.8 to 2.5 how would that compare in terms of hotspotting and color shift to a flat (and smaller) 1.3 gain screen?


What I'm wondering is if the curve is only in one direction (horizontal) is that enough to overcome the problems of hotspotting and color shifting that you'd have with a high gain flat screen?


Just so I understand, is hotspotting caused by the combination of high gain screen and the fact that different portions of the screen are varying distances from the projector tubes? A curved screen would (depending on the amount of curve) eliminate or reduce the varying distances between tubes.


Is color shifting caused by the same combination?


Thanks!


Matt

 

[KFung]

Matt,


If done right a screen with a horizontal curve should be able to give results as good as a flat screen with a 1.3 gain surface. The objective would be to pick a gain that would not hotspot or colour-shift in the vertical dimension and then determine the horizontal curve to match. I'm not familiar with the gain fall-off characteristics of the higher gain screens so I couldn't recommend a particular gain. Your best bet (for optimum results) would be to talk to the people at Sigma Design Group (I think you can e-mail Gerald Nash at [email protected]). Their consulting fee for a horizontally curved screen is much less and they should be very familiar with Stewart screens. All the TORUS screens that Sigma Design Group designs are made by Stewart.


Hotspotting is caused more by the different points of the screen being at different *angles* to the light from the projector. High gain screens are more reflective (like a mirror) rather than diffusive (like a painted wall). So a high gain screen directs more light along the path of direct reflection (you know, angle of incidence = angle of reflection). At the edges and corners of a screen, light bouncing off the screen mostly goes away from the viewing position. Curving the screen reflects more of the light towards the viewer.


I believe colour shifting is similar to hot-spotting, and really only a problem with 3-gun CRT's and very high-gain screens. Because each colour comes out of a different lens in a different position, the angle with the screen for each colour is different and you essentially get a different type of hotspot for each colour. This leads to colour shifting because at some points the green will be brighter than the red and blue, or any combination of the above.


Regards,


Kam Fung