THE SILVER SCREEN

http://lennylipton.wordpress.com/2009/11/07/the-silver-screen-part-1/#more-297

http://lennylipton.wordpress.com/2009/11/09/the-silver-screen-part-2/

 

By lennylipton

Ah the silver screen: searchlights scanning the Hollywood sky, glamorous premiers, gorgeous actresses….  The silver screen is a term that has denoted the glamour and excitement of Hollywood since Chaplin twirled his cane. While to some it is the most visible sign of hope for the cinema for others it is a dreaded surface upon which to project those old standby 2D movies.  But there’s so much more to it than glamour – there’s dreadful science.  It’s a technology that ought to command the industry’s keenest minds, because, after all, that’s where a hundred and fifty million bucks wind up as a vibrating veneer of a hundred billion photons reflected into the eyes of tens of millions of photon consumers. That’s one big point in favor of the film industry – they have not dehumanized the customer to the point where he or she is called a consumer.  The customers are still the audience, people with feelings rather than human maws born to consume piles of chazarai made in China. 

In Hollywood, as elsewhere, technology decision are made based on incomplete information and herd instinct rather than on common sense, and science and engineering are basically common sense. It’s the scientific method that is the hope of mankind, not emotion and prejudice. There are many in the industry who reject projecting on a silver screen, and they have good reasons.  But how substantive are these reasons? 

The silver screen, actually an aluminum surfaced screen, is not only an enabler for the best method of projecting 3D images but it is the weakest link in the optical system for that method: polarized-light-image-selection.  While the Dolby system produces good images and does not require a silver screen it is not bright enough for the biggest screens (unless two projectors are used) and it uses dauntingly expensive eyewear that must be cleaned in the theater.  (Ditto XPand.) Most theaters projecting 3D use the polarization selection method whether of the circular variety (RealD and Master Image) or the linear variety (IMAX and most theme parks). 

In the past four years or so silver projection screens have improved.  When the first screens were installed for Chicken Little, the studio and exhibitor complaints centered on two issues: visibility of seams and mottling and a textured appearance on the surface of the screen.  People in the industry were also greatly concerned about how well these screens performed for 2D projection, not just because of the defects mentioned, but because their colorimetric characteristics are different from matte screens. And in this town technical and creative people will go bonkers if the images they have so painstakingly created aren’t accurately reproduced.  They can go as bonkers as they want but outside of LA the projection of movies is sometimes a smack in the face to every DP, art director, and colorist who cares about the movies.  (And let’s not get started on what happened to that image quality on LCD TVs.) 

Which stumps me becasue the Dobly system to my eyeballs does not meet spec.  One image is more highly saturated than the other, and you can see that oh so clearly with the violet-blue laden Avatar.

Silver screens are designed to conserve polarized light and have gain.  When polarized light is projected onto a nonmetallic surface, a matte screen, it is depolarized.  The reflected rays no longer have the orderly orientation of the electric vector (a component of the electro-magnetic wave construct that describes the physics of light) that defines polarized light.  When polarized light is reflected by a nonmetallic (a dielectric surface) it is depolarized and no longer useful for image selection.  A dielectric is a material that has closely-bound electrons.  It is a poor conductor of heat and electricity. It’s an insulator. So the same kind of material that has difficulty conducting electricity and heat will not preserve the characteristics of polarized light. Matte screens, usually made of vinyl, are dielectrics. 

Dielectrics, which do not preserve polarization characteristics, have reflection characteristics explained by Brewster’s Law that says that light rays that are at glancing angle to the surface undergo some degree of polarization. But the material that is most interesting for making silver screens is a metal (or a conductor).  Silver screens are manufactured by painting or coating matte screens with an aluminum pigment mixed into some kind of a binder or medium to be coated or sprayed onto the screen surface.  Motion picture screens are usually made of vinyl plastic 54 inches wide.  These sheets are welded together in vertical sections.   The vinyl is welded together and then painted with aluminum pigment. The weld is accomplished in a different ways but however it is done the weld has to be invisible because nobody wants to look at them.  They are easily as distracting as the guy sitting in front of you wearing a Dodgers baseball cap.

Before we dig deeper into the characteristics of the silver screen a word or two about matte screens: Matte screens are perfectly fine for 2D if the projector has a bright enough source of illumination, and they have a more or less “Lambertian surface” so that the incoming light is reflected or distributed evenly in space with pretty much the same brightness for any seat in the house. 

In addition they have very little shading.  As you look across the surface of the screen from your seat you will see very brightness change from corner to corner.  Much of the shading you will see is a result of the projector optical system and that’s called vignetting, but you can’t see it unless you are looking for it and the movie happens to be shots of the sky or a close up of a sheet.  So when you’re off in the corner in the worst seats in the house –in the front row, way over on the left or the right side – illumination holds up pretty well across the screen.  The closest part of the screen and the furthest part of the screen are pretty equal in brightness. Of course you are looking at a distorted image, but that’s another problem.  Some people seem to like sitting in the front row and some people like pickle parfait pie or chicken mint ice-cream.  Do you really subscribe to the seemingly enlightened old saw that says there’s no accounting for the other guy’s taste?  In your heart of hearts you know that if other guy doesn’t agree with you he’s warped. 

Burt we are interested in the silver screen. With the surface of the screen coated with metal it now has the ability to conserve polarization.  If polarized light is projected on it polarized light will be reflected because the metallic atoms of the surface have free electrons, and when light shines on its surface those electrons are able to vibrate in any which way to reflect polarized light.  

These screens ought to get pretty good conservation of polarization even at steep angles for the worst seats in the house but the ability to conserve polarization falls off with at the side seats.  Why does it happen?  I think it probably has to do with the binder which may be a dielectric or even (heaven forbid) birefingent.  I have heard it described in terms of diffusion, but I don’t think that’s right because a diffused metallic surface is still a metallic surface that has free electrons.  

Motion picture screens are a tradeoff between reflection and diffusion. If you were to look into a mirror that’s being used as a projection screen you will see a bright glob of light where the projection lens is.  If you add some diffusion to the surface you are able to form an image on the surface of the screen and you get rid of the so-called hot spot.  A silver screens has to be the right balance between reflection and diffusion and just like a mirror it can also have a hot spot.  But if the screen is curved the hot spot can be spread out over its surface so it cease to be visible.  Silver screens are typically curved into a section of a cylinder and this tends to mitigate the hot spot.  

Hot-spotting depends on the geometry.  For a long narrow auditorium with a long throw – in other words, the distance from the lens to the screen is great – when sitting in the middle of the auditorium, you will see very little hot-spotting.  But that’s not the way modern auditoria are designed.  They are designed to be more square-ish and have large screens.  A theater in a modern multiplex is the toughest geometry for a silver screen.

Silver screens, if they are well made and installed, can have minimal hot-spotting, but they still have what I call shading.  I make a distinction between hot-spotting and shading.  Although they may come about from the same reflective characteristics of the screen, shading has to do with an asymmetrical change in brightness across the screen and is typically dependent upon where you are sitting.  Shading happens quite noticeably when sitting in the worst seat in the house, say in the front row way on the extreme left or the right. In fact, the worst seats in the house for viewing a 2-D movie on a matte screen become even worse when viewing a polarized light stereoscopic movie on a silver screen from a bad seat. 

I haven’t gone to every theater that has a silver screen, and I can’t tell you for sure that they are all good screens.  But I do know that in the years since Chicken Little was released the manufacturers have learned how to make better screens so when you are looking at a 2-D or a 3-D movie you will not see the seam; and I do know that the manufacturers now know how to make screens that don’t have any splotches or texturing.  I have been in a number of theaters where there are good silver screens that work very well for 3-D and as well for 2-D projection.  The screen’s performance for 2-D is important so it’s reasonable for studios and theater owners to demand that it function well in both modes.  

A lot of work has been done on high-gain metallic-surface screens, not because of their polarization characteristics but because they are bright. The matte screen that I talked about earlier with the Lambertian surface has a gain of 1, let’s say, with light reflected more or less equally in all directions.  But in a theater that’s not what is needed because there’s nobody watching the movie up on the ceiling or down on the floor or way off on the sides.  A lot of the information about how to make a good high-gain screen can be found in patents that are assigned to Eastman Kodak.  The idea behind a high-gain screen is to take the unwanted light that goes off to the ceiling and floor and wherever, and to send it where it is needed – to the audience. 

Probably the finest example of a screen of that kind –the one I’ve used and I know about – is the Ektalite screen.  The Ektalite screen, which was offered in the 1970s and into the 1980s by Kodak, is possibly the finest stereoscopic projection screen ever offered.  It wasn’t designed to be a stereoscopic projection screen; it was designed to be a high-gain screen used in classrooms or conference rooms that was bright and rejected unwanted spill light from windows and overhead lights.  I’ve used several of them in different sizes, and I can verify what I’ve said because I measured the performance of one of them.  The Ektalite screen – on-axis, which is typically the way high-gain screens are measured – had about a 10:1 to 14:1 gain. (That’s 14 times brighter than a matte screen.)  Remember, we’re talking about an era in which people didn’t have flat panel displays and they were projecting using slide projectors or 16mm or 8mm projectors; so set up in a classroom the Ektalite screen was super, because over a 60-degree angle of view, properly placed, everybody in the classroom could see a bright image.  

It worked great for its intended purpose, and it achieved this by interesting technology.  The surface was aluminum foil, and it was textured by placing the two pieces of aluminum foil under high pressure through rollers, with oil introduced between the two surfaces.  The result was a micro-fine bark-like texture that served as the diffusing surface.  The Ektalite screen combined reflection and diffusion perfectly, but it would still have had a fierce hot spot if the aluminum foil had not applied to a compound-curve screen which was shaped like the inside section of a sphere.  It was made on dense foam core or some such plastic, and the foil was bonded to the surface.  It had a delicate surface and you had to be careful cleaning it.  But wow! what a screen. 

Modern screens – and I haven’t tested one in a few years so my observations may be out of date – although they conserve polarization, are not nearly as good.  An experiment I did in my days at StereoGraphics measured a standard aluminum painted vinyl screen’s polarization efficiency starting with linear polarizers that had a 5000:1 extinction ratio. That means that only 1/5000th of the unwanted image passes through crossed polarizers when measured on an optical bench.  But when measuring the dynamic range reflected from the surface of the silver screen on-axis, using the same polarizer over the lens and for eyewear, the dynamic range was reduced to 200:1.  Still good for stereoscopic applications, but this told me that improvement was possible because this painted vinyl screen was not as good as the Ektalite for polarization conservation. As I mentioned, Kodak wasn’t particularly interested in making polarization-conserving screens, but it turns out that they did with the Ektalite. The Ektalite screen had, on-axis, 1000:1 dynamic range using good 5000:1 linear polarizers.  The polarization loss on-axis was much less than vinyl painted screens, and off at the edges of the field (admittedly narrow) the dynamic range was about 200:1. 

When using circular polarizers on a bench I got numbers that were a few hundred to one for dynamic range.  And the same protection measurements will show a dynamic range of maybe 20:1 dynamic range on a vinyl painted screen – which is not too good. Hence, Real D adds ghostbuster to the ZScreen system to improve the effective dynamic range.  

I’ve recently seen both Real D (circular) projection and IMAX (linear) projection, and they are both good.  Do I prefer one?  IMAX projection with linear polarization used two gigantic pieces of 70mm film running horizontally through the projector, the size of small picture postcards, and the image looked fantastically good.  If it didn’t see this image quality improvement there would be no excuse for what I paid for the family to see Cloudy With A Chance of Meatballs.  I had sticker shock at the box office. With regard to the characteristics of circular versus linear and circular’s head tipping advantage, booth were fine.  You can only tip your head so much before the misalignment of homologous points causes a breakdown of fusion so the circular advantage isn’t all it’s cracked up to be, especially when starting off with linear’s much higher dynamic range. 

To me the Ektalite says more work can be done to improve silver screens.  They may be better than ever but the Ektalite performance tells me that more is possible.  True, the Ektalite had a narrow viewing angle and required a compound curve to get a super gain, but its performance points to possibilities. High-gain silver screens on-axis – which is the way they are rated – can have a 2.4:1 gain, but off-axis the gain is going to be very much less.  So these screens might work out to be an average of 1.5 gain or something like that, if you make some kind of a weighted average according to some reasonable formula for the various seats in the auditorium.  Same story for polarization conservation as a function of viewing angle. 

If we can have a screen that has a true 2 or even 3 gain over a board viewing angle, that would be an improvement for the stereoscopic cinema.  The present illumination for 3-D movies is a niggardly 4 to 5 foot-lamberts. That’s true across the boards for all DLP-based digital systems with the exception of the RealD XL system which is a twice as bright. As for the rest that’s pretty dim when you consider that the SMPTE recommendation for 2-D projection is 14 foot-lamberts.  I’m of the opinion that there is really no excuse for stereoscopic images being projected any less bright than 2-D images. I don’t think something special is going on that makes it okay to have low light levels for 3-D projection.  There is nothing additive going on between the two eyes in terms of illumination.  There may be something going on in terms of perception of image quality, possibly a reduction in granular noise and maybe improvements in sharpness.   

We really have two issues now with the silver screens, and they are (l) get the gain up, and keep it even across the seats so that wherever you are sitting you have a fairly high gain, and (2) make sure that polarization is conserved consistently for every seat so one can see a good stereoscopic image anywhere in the auditorium. 

Having said this I know that there are many people, and experts counted amongst their numbers, who simply dislike silver screens and find them to be unacceptable for 2-D projection.  I’ve watched many 2-D films on good silver screens and I think they look better that way, but to each his own. The way this issue is going to be resolved may be as follows:  Matte screens will someday become the minority with the ubiquity of the stereoscopic cinema.