Hi guys,

I’ve been looking at the pictures in the Anaglyph group and I’m a bit concerned. A fundamental principle of stereophotography and 3D anaglyphs, is that it is a composite image made from two original images, which are taken at a slightly different angle. These are then combined through the red and blue filters to create the illusion of dimensionality. This can be done digitally using either two cameras that fire off simultaneously (a stereo camera rig) or with one camera using what is known as the “cha-cha” method (taking one shot and then shifting the camera and taking a second one).

You cannot make a 3D picture using a single photo and just displacing the blue and red layers. If the red and blue at the foreground of the picture is displaced by the same amount as in the background, then there is nothing to trick the eyes into creating a 3D effect. The displacement of the red and blue channels is normally bigger the furthest your subject is from the camera, and they meet towards the front of the picture (what is known as the stereo window). What is also critical that the red and blue images are only displaced horizontally but stay vertically on the same plane.

You can see this effect on the two pictures I’ve uploaded to the group: the Hyacynth picture is taken with a single camera using the cha-cha method, whereas the B&W Syntagma one was taken with an original stereoscopic camera that took two simultaneous images on a glass plate.

Having said this, there are techniques and tools available for converting 2D to 3D pictures, but that involves using pixel-shifting algorithms and a LOT of hard work, to simulate the two different images that can then be combined to create the 3D anaglyph. But I don’t think these have been used here.

I would be interested to find out what techniques have been used to create the other pictured posted to the group.

Regards
George

The method I use is as follows:
Cha-Cha to get the L and R images. (L one first then rock to the right about 6 cms to get the R image). I try to select something in the mid-foreground that will be the aligned ‘window’. I also try to have some horizontal lines in the image so that I can easily correct spin-error.
1. Open L and R in photoshop, and convert each to greyscale (image>mode>grayscale)
2. Convert L to RGB (image>mode>RGB) (do not do this to R)
3. Select L image. Open channels display (window>channels). Highlight blue & green channels (use shift key).
4. Select R image. Select the whole thing (select>all), copy it (edit>copy).
5. Select L image. Paste (edit>paste). Highlight RGB channel (select it).
6. Align the L and R images: Highlight only the Red channel. Go to the RGB channel and click the box on the left, to show the eyeball. Only do that, NOT select the RGB. Asa result, you should see an eyeball against the 4 channels, with only the Red channel celected.
7. Pick a point in the ‘centre’ of the picture to “line up”. ZOOM into this portion of the picture (Ctrl + repeatedly), and select the ‘move tool’ then using the arrows on your keyboard, nudge the red image across until it perfectly lines up at the chosen ‘central point’.
8. If your originasl photos were ‘twisted slightly’ from one to the next, you may need to rotate the red slightly to correct for this. The method I use is: Crtl-T for edit transform. Find the centre marking, click on it and place it directly over the chosen ‘central point’ from previously. this will cause rotations to be around that point. Now zoom out (Ctrl – repeatedly) until you can see beyond the edge of the picture. Place your mouse to one side – it will change to a rotate shape. Click and drag up or down to rotate the red image. Using horizontal lines in your images, rotate until they line up perfectly.
9. Cropping: Using the crop tool, select the ‘whole’ portion of the merged image, press ‘enter’ to complete then crop.
10. Save as a jpg image – it is now ready to be ‘viewed’ with the suitable paper glasses (red left eye, and cyan right eye).
Keith Richardson