Making 3D Pictures
by
The Red / Green Method
(Anaglyph)

Clive R. Haynes FRPS

 

What is a 3D image?
Our human optical system surveys a scene and transmits messages to our brain which are, in fact, the contents two simultaneous, two-dimensional images, spaced about three inches apart. In making a 3D image we need to convince our optical system that, just as in real life, the two flat images presented add together to make a stereoscopic (3D) picture.

In Victorian times 'bioscope scenes' viewed with a hand-held stereo viewer were very popular. Two pictures were taken of a scene and pasted together side-by-side for viewing through simple lenses. They were very good with excellent depth and clarity.

 

How to do it.
The first requirement is to take two pictures of the scene, one left, one right. Obviously it's best unless using a specialist stereo camera, to avoid scenes with movement.

The viewpoint for pictures for most normal scenes, using say a 35mm to 100mm focal length lens (35mm photography) need to be about 3 inches apart - this is the average distance between the left eye and the right eye. Special devices can be bought to do this with extreme precision, however it can be done quite successfully by the 'rock from one foot to the other' method. To do this take one picture (left) then rock from the left foot to the right so as to lean to the right - go about 3 inches (75mm) keep everything level and without twisting the camera, take the second shot (right).

'Toeing-in'
An improvement upon the simple 'move to the right' method can be made by 'toeing-in'.
This is what you do.

When taking the left shot, look at the centre of the scene area for a 'reference object' and remember where it is within the frame.
After moving the 3 inches to the right, pan the camera back into the original scene area (this is 'toeing in').
Place the 'reference object' in the same spot as it was in the left image.
Take the shot.
The slight 'pan' almost re-aligns the image elements, however, they remain displaced by 3 inches to the right and this gives the stereo effect.
The 'toeing-in' is to reduce the amount of image area a that spills beyond the frame of each shot when combined in the 'anaglyph'. In this way a more complete image area is retained. Actually it's far easier and more obvious to do than explain!

The two resulting pictures can be viewed as a stereo pair. If they are slides, then a double viewer can be used to recreate the 3D scene.
If prints, then I find that placing two 5" x 7" prints (one 'left', one right') side by side about 3 feet away and deliberately going cross-eyed to view them works crudely. Some people find that going cross-eyed is not at all easy - so it's not something that I would recommend.

Work in RGB (colour) mode
Finally, I recommend using a colour image (either scanned or DI capture) as, if required, the tone and contrast of the intended monochrome image can be adjusted to suit the subject and aid visual/spacial separation.
If you are using a monochrome source, perhaps black & white film, you must scan as RGB or if it's a Greyscale image convert to RGB before beginning the process outlined below.

 

An Anaglyph
A popular form of 3D picture is a two-colour image that recreates depth by making each half of the stereo pair a different colour, this is known as an 'anaglyph'. Usually red and green is chosen for this.

To achieve a result we need to make the left eye image red and the right eye image green. (This is the convention, it could just as easily be the other way around - provided we view it that way).

Digital imaging has made the production of anaglyphs simple.

The word anaglyph incidentally, derives from the Greek - anaglyphos, anaglyptos - in low relief - ana, up/back, glyphein, to engrave or carve. 'Anaglypta' wallpaper has the same derivation!

 

Making an Anaglyph (red/green) 3D picture

What follows looks long-winded because I've detailed every step. In practice the whole process is very quick and I can run through it all, once scanned in, to a final 3D image, ready to print, in under two minutes - honest!

 
1 Import/scan the 'left' and 'right' images into Photoshop in the usual way.
NB Should the images be JPEG files, convert them to .PSD (Photoshop) files by 'Save As'.
If the images remain as JPEG's, the 'Duplicate Channel' functions listed below will not function correctly.
 
2 Make any obvious corrections, bearing in mind that the two images will ultimately overlap each other to form a single optical composite.
NB Do not make any alterations to the pixel dimensions either by 'Cropping' or 'Transformation' at this stage.
If alterations are made, the 'Duplicate Channel' functions listed below will not function correctly.
 
3 Go to the 'left' image and de-saturate it. Do this via Image > Adjust > Desaturate
 
4 Go to the Channels palette and click on the RED channel (it highlights in blue)
 
5 Go to the horizontal right facing arrow at the top right of the channels palette and click. In the menu that opens choose 'Duplicate Channel'
 
6 In the dialogue box that opens go to the Document box and click on the drop-down arrow and choose 'New'
 
7 In the 'Name' box (not the 'document' box), type in the name of the image you are working on. (If you've already saved the left and right originals then choose another title otherwise the system will have great pleasure in telling you that the file name is already in use!). Click 'OK'.
A new monochrome image will now appear
 
8 Minimise or close the 'left' original. Keep the new monochrome image running
 
9 Go to the 'right' image and Desaturate it. Do this via Image > Adjust > Desaturate
 
10 Go to the Channels palette and click on the GREEN channel (it highlights in blue)
 
11 Go to the horizontal right facing arrow at the top right of the channels palette and click. In the menu that opens choose 'Duplicate Channel'
 
12 This time in the dialogue box that opens, go to the 'Document' box, click on the drop-down arrow and choose (click on) the file name you gave in step 7. Click 'OK'
 
13 Go back to the 'right' original and repeat steps, 10, 11, 12 this time selecting the BLUE channel in step 10
 
14 Minimise or close the right original
 
15 With the monochrome image running, created from the three separate colour channels (red = left, green & blue = right), go to > Image > Mode > RGB and as if by magic, the monochrome image becomes a red/green (and blue) anaglyph
 
16 Put on your stereo red/green spectacles and view in glorious 3D
 
17 All should be well but some adjustment may be necessary. See 'tweaks' on the next page.
 
18 Print the picture to matt paper as reflections from a glossy surface may be distracting
 
To continue, click on the link below
Continue

3D Gallery

Know - How Contents

HOMEPAGE

CRHfoto@aol.com