I must admit that I didn't think there any point to making a tutorial on this topic, because it is so ridiculously simple. However after talking to friends who are photographers, and even more after seeing a product at my favourite online camera shop that automatically generates stereoscopic images, I realized that maybe this isn't as well known as I believed. (And for the record, that camera accessory costs over $200CDN and according to most reviews does not generate results even close to this free alternative)
This following method works for most camera lenses, although I do find the best results are for low focal length lenses. In fact my best results are with an 8mm fish-eye lens!
Step 1: Obtain a camera, with a suitable lens. As I wrote above, I prefer an 8mm fish-eye lens, but please do not feel restricted to this choice. I encourage you to experiment with different lenses and different settings, and see what effects can be created.
Step 2: Line up you photo as you normally would, but use the Manual mode if possible. You can use an automatic mode to get the right exposure and aperature, but it is important that the settings now change between photos. So once you have the shutter speed, aperture, and focus correct, switch to full manual mode and lock them all in.
Step 3: I like to shoot these photos through the viewfinder rather than LiveView, as it is easier to line up the resulting photos later. Take one photo with the viewfinder against your left eye, and then shift the camera and take a second photo with the viewfinder against your right eye. Ideally the camera and lens should not rotate at all, but only move parallel to your eyes.
Step 4: Download the photos to your computer, and load them into your favourite editing software such as GIMP or Photoshop. The first photo should be duplicated into a second layer (because Photoshop won't allow the background layer to be transformed) and the second photo loaded into a third layer over the first photo.
Step 5: Apply a transform to the second layer (you may have to turn off the third layer's visibility to do this) and scale its width to 50%. This can be done either with the mouse, or by typing 50 into the appropriate control box. Drag this compressed image to the far left of the photo, so that the left side of the image is flush with the left side of the canvas.
Step 6: Now repeat this process for the third layer, scaling it to 50% width and then moving it flush with the right side of the photo. You now have a completed stereoscopic image, with much better results and for $200 less than the camera shop gadget.
The final photo can then be imported to virtual reality software, and used to create stunningly realistic environments from very simple photos. So I encourage all of you to try it out and see how it works for you, and you might just be amazed at how good the results can be with a little experience!