Stereoscopic Photography...the Science of Solid Viewing

Projection onto a Screen and Transparency Viewing

Projection for this project means 35 mm positive film slides, Kodak 4200 projectors (now obsolete) with bright lamps and Schneider F2.8 150 mm lenses. Viewer separation of the two images is achieved by quadrature polarized light projection; with the viewer wearing matching glasses. Left and right images being polarized in quadrature. This simple polarization scheme requires very accurate alignment of the projection polarizing filters with a `standard viewing glasses’ mounted on a post. The viewers must wear glasses that match the standard and keep their head aligned precisely in a lateral horizontal line. A better system of polarization would be to use circular polarization; left and right being clockwise and anti-clockwise circularly polarized. (Some day I shall convert as that system is theoretically better.) Currently my slides are in plastic mounts and my lenses are for a curved field. For best results the slides should be in flat glass mounts with flat field projection lenses. Two types of screen have been used:-

1. Darlite silvered front projection screen. With this screen the slides are mounted to give a direct image on the screen facing the projector operator. With this screen the viewer must sit just below the screen and tilt their head back. (Head frozen laterally and tilted back = sore neck). With front metallic reflective screens the reflected image is in polarization quadrature with the incident light; hence projection filters must be set accordingly. Reflection does in a very minor way confuse the polarization, resulting in a cross-image effect that slightly confuses left and right. Projector focusing is done by viewing the screen image though binoculars to very fine tuning the focus.

2. Screen-Tech rear viewing screen. With this screen the slides are mounted to give a reversed (faint) image on the screen facing the projector operator. The viewer sits directly behind the screen in direct line with the projection beams. (No sore neck). The rear viewing screen does not change the image polarization but it does introduce some significant polarization confusion and cross image effect. Focusing is done by placing a white paper on the screen and focusing onto that.

With both screens the viewer sits very close; optimally 0.96 of the image width. At this position parallax and perspective coincide. In practice I go a bit closer say 0.8, as per the diagram. Because of this close viewing the screen image must be very clear.

Of the two projection methods, both using linear quadrature polarization, the front viewing screen gives the clearest picture with least cross-image. All screens suffer some image scattering and for rear view screens dispersion as well. Polarization of the images is never perfect so cross-image distortion must take place. All this degrades and confuses the image even with good focus, especially with the rear projection.

So far my projected images are rated (by me!) OK to good; by normal 2D projection standards they would be considered very good images. Stereoscopic images require excellent resolution because they are so realistic; any screen image fault may be considered subconsciously as an eye vision fault.

A suitable experimental projection system that gives good images with correct parallax and perspective is shown below. The viewer should sit between 0.8 and 1.0 metres

This is obviously a very personal projection system – only one person at a time can see it. However even this experimental system gives very good pictures. For a 35 mm positive slide taken with a good camera set on a secure rigid tripod contains a vast amount of information. Resolution wise it is about 12 Mega pixels with 3 layer deep subtractive colour. Dynamic tone range is hard to estimate – but with the proverbial `black cat in snowfield’ slides can get the texture of the cat’s fur and the texture of the snow. Prints do not come close to that performance.  Slides are just plain good pictures!

But great improvements could be made to render the pictures truly life like – by that I mean that the viewer feels that they are actually there in the scene.

1. Glass mount and accurately align the slides.
2. Use flat field high resolution long focus projection lenses.
3. Use good projectors with accurate slide registration and good condenser lenses.
4. Use bright lamps.
5. Use circular polarization – clockwise, and anticlockwise. (Theoretically better. Practically?)
   (Remember a silver screen reverses it upon reflection)
6. Oh whilst I dream superlatives – use large format 60x60 mm pictures.

Non-projection Transparency Viewing

The actual slide pairs may be viewed in a viewer such as the Twin Elite 2x2 viewer Berezin Stereo Photography.  Such a viewer has high quality lenses but the perceived image as a bit too small to give the correct perspective with this wide angle stereoscopic system.

The print viewer described elsewhere has a lens focal length of 300 mm, too long to give the correct angle of view and thus perceived perspective. If the focal length is reduced to improve the perspective it becomes very difficult to get light in to illuminate the prints. However if the prints are made as back lit transparencies this limitation is removed.

A back lit transparency  viewer taking  5x 7.5 inches ( 127 x 190 mm) with 250 mm focal length lenses has been tested. In this case the prints have to be placed one above the other (and staggered by 60mm) hence the convenient belt transfer system is precluded.  The transparencies are made by ink jet printing on Epson ink Jet transparency film with an image that has been increased in density double or triple times. This is an interesting line for development.