This thread is fascinating. It seems to basically say curving the screen can help, but striking the right balance of how much to curve is very difficult.
I rehashed my optics background and having a radius of curviture twice the throw distance should result in the light leaving the screen perpendicular to the wall behind it (at least I think). That should decrease but not completely eliminate hotspotting and should make it so anyone at a given distance should get the same image provided they are within the width of the screen (perfect for a couch). The only way to completely eliminate hotspotting with a high gain screen would be to have the radius of curvature slightly less than the throw and to have the PJ just behind and above your head. Obiviously that isn't practical as there would only be one sweet spot (perfect for a chair).
Practically speaking I would think having the radius of curvature 2 x throw would eliminate hot spotting for all but really high gain (or poorly constructed) screens, and to generalize, l think it's safe to say having a mild curviture of radius as described above for a medium to large sized 2.35:1 screen should create a fairly even/consistent image (in ftL) for a rather large seating area.
That's great, but it hasn't addressed the potential for pincushion effect or barrel distortion. Please correct me if I'm wrong, but I take it as pin cushion effect occurs when a PJ WITH an anamorphic lens casts a widescren image on a large flat 2.35:1 screen. This is especially likely to occur if the throw is less than 2 x the width of the 1.78 portion of the 2.35 screen. The opposite is barrel distortion, and that potentially occurs when a curved screen (particularly when the throw is long) is used WITHOUT an anamoprphic lens in front of the PJ.
So... the question I have is what would be the best combination of throw distance and radius of curvature to minimize the pincushion effect when the anamophic lens is used, minimize the barrel distortion when the anamorphic lens is not used, minimize hotspotting when a moderate to high gain screen is used and maximize the viewing area?
I would imagine that is tough to do, but having the radius of curvature at or a little more than 2 x throw when the throw is a little less than 2 x the width of the 1.78 portion of the 2.35 screen may strike the right balance.
What do you guys think? If you don't think you can make a defintive suggestion, please state the curvature of your screen (flat or curved and if curved what the radius of curvature is), which PJ you use, then length or your throw, which anamorphic lens you use (if you have one) and comment on your experience in your home theater.
This is particulaly important to me because I'm seriously thinking of building a DIY curved 2.35:1 AR screen with a sheet of 1.24 gain white laminate. Total cost should be about $150. Obviously it won't be as good as a curved Stewart screen, but people in the DIY screen subforum seem to like this material a lot, and it would be fun to make. There is a lot of potential, but if it doesn't work, at most I lose $150 and a little bit of time. Not a bad gamble