All times are UTC + 2 hours

Post new topic Reply to topic  [ 3 posts ] 
Author Message
 Post subject: Low-tech viewing for parallel image pairs
PostPosted: 21 Jan 2011, 13:07 

21 Jan 2011, 10:56
Posts: 5
Prototype - Copy.jpg [552.16 KiB]
Not downloaded yet
Until the various competing 3-D display system technologies sort themselves out (and also because I'm on a starvation budget), I'm sticking with parallel image pairs for most of my viewing and photo-editing. While anaglyph (red/cyan) remains popular with many people new to 3-D, there are some obvious problems when using tinted glasses with colour photos and video.

Chief among the anaglyph limitations is the distortion of true colour (a red rose looks blue!), as well as an annoying retinal effect with some bright shades of blue and yellow in the original image. Another common problem is incorrect pair alignment and convergence. The latter problem is not nearly so noticeable when looking at a non-anaglyph pair instead of a merged image overlap, because our eyes adjust. It is also correctable by simply taking more care when preparing the anaglyph.

Do-it-yourself interlacing for shutter glasses and lenticular overlays is obviously the next step, but it's complicated and seems mostly affordable on iPhone displays at this time.

Having grown up with a Stereo Realist camera which my father purchased 60 years ago, I became accustomed to viewing parallel image pairs on tandem Kodachrome slides in a binocular viewer. This format is still unbeatable for sheer simplicity, but it cannot be upscaled for the computer monitor, to say nothing of the impossibility for video. The short focal length (15+ cm) of stereopticon lenses makes this method impractical for viewing image pairs on a monitor.

That leaves two choices for parallel viewing: diopter prisms or mirrors. Parallel images can be "free-viewed" without glasses with a little eye-training, but not everyone can master the technique, and it becomes tiresome after awhile for those who can (I can't imagine watching a two hour film that way!). Free-viewing is probably easier in the "cross-eyed" format, where the Left and Right images are reversed.

The purpose of using prisms or mirrors is to deflect the images coming to each eye in such a way that a third, "virtual" image appears (there is a "one eye" system as well, and there is a pocket mirror technique, using one "flipped" image).

With mirrors, two pairs are used so that the images don't need to be "flipped". Front surface mirrors are preferred to avoid the refractive errors of glass. Front surface mirrors are unfortunately rather costly and cannot be wiped clean of dust or smudges without potentially damaging the ultra-thin metal coating.

I have concluded that prisms are the way to go. Berezin Photo (and other suppliers) sells some excellent 3-D prism glasses from NVP3D (Switzerland) for under US $20, but they are built for a child's face. Since I also wear reading glasses, I wanted something to wear over my regular lenses, and I also wanted to have a viewing platform with masking baffles to eliminate the visual interference and concentrate my focus only on the 3-D image.

Anachrome, a small company run by a 3-D pioneer, has some prism glasses which seem to fit the bill, except that there's no comfortable allowance for wearing them over a pair of reading glasses (if you go to the website you'll also find some anaglyph glasses of an improved design). However, they are pre-set only for cross-view image pairs, and all of the rendering I have done in my 3-D photo collection is in straight parallel. The output of most modern 3-D cameras is also straight parallel. Still, I think that the prism configuration for cross-view may be superior.

I also must mention another format for prism viewing: the over-under "KMQ" system developed in Germany in the 1980s. In this case, the images are "stacked" on the monitor, the Right above the Left. This presents some interesting possibilities for presenting dual "landscape" or wide screen formats, and I think that "over-under" may win out in the future, especially on lenticular displays. KMQ glasses may be purchased from Berezin and other suppliers in Europe. A nice overview can be seen at, and you can follow over-under software developer Thomas Kumlehn's discussions at, or on Facebook.

After all this, I decided to build my own prism viewer. I wanted the flexibility to experiment and I wanted the clearest possible optical quality for my work. Unfortunately, most glasses and viewers presently sold are necessarily a compromise to keep the cost down. I also think that some makers of passive glasses have not expected such a surge in interest in high-quality viewing equipment (except in gaming and HD TV). Previously, some of these glasses (the cardboard ones, especially) were toys.

My prototype is pictured in the attachment. After a number of hours brushing up on the mathematics of prism optics, I came up with a diopter strength which seems optimal for viewing my monitor from a comfortable distance. I ordered my prisms from a supplier of medical eye products and they required my eye doctor's authorization!

I wanted a flexible headband instead of eyeglass "arms" and I wanted an apparatus to allow me to shift the prisms around for the best adjustment. I also wanted the chance to test "cross view" (reversing the prisms) and "over-under" (orienting the prisms almost like the old camera rangefinders). What worked best was also the cheapest -- I simply bought a pair of eye protectors used by woodworkers and other factory employees for around US $5. I punched out the transparent face shield and replaced it with a 1.5mm sheet of black plastic, bought at a hobby shop and cut to the face shield pattern with two square-ish openings. I could now just glue the prisms in place, but I'm using tape instead, until I have tried all of the possibilities. I think that having mounting brackets for the prisms to slide in and out would allow the flexibility for parallel, cross-view and over-under.

And how does it work? Extremely comfortable, no eye strain, perfect vision. What more can you ask?

 Post subject: Re: Low-tech viewing for parallel image pairs
PostPosted: 21 Jan 2011, 13:40 
Site Admin
User avatar

29 Apr 2010, 17:51
Posts: 3148
Location: Sofia, Bulgaria
Sounds interesting, how much was the total cost of the project? I imagine the prisms were the most expensive part? And also what size of monitors does this viewing method support?

 Profile WWW  
 Post subject: Re: Low-tech viewing for parallel image pairs
PostPosted: 21 Jan 2011, 13:52 

21 Jan 2011, 10:56
Posts: 5
Yes, the prisms were the most expensive. They cost US $10 each. I'm sure that there are some less expensive ones available, maybe in Europe. The total cost was thus about US $32. Thank you for your interest! Are you also working on some 3-D experiments? I'll have a look at some of your posts.

Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 3 posts ] 

All times are UTC + 2 hours

Who is online

Registered users: No registered users

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to: