Question: You’ve included anaglyphs in several of your posts over the years, and I always enjoy them. I happen to have a pair of red/cyan 3D glasses from my early days in online gaming. I hang out on several photographic forums, usually just trolling for general info, and see a lot of anaglyphs. But I’ve never read a detailed explanation of how anaglyphs are made. You obviously know how to do it, so could you explain the process for a beginner? Thanks! — EAA, London, UK
Answer: Well I just happened to include an anaglyph in last week’s post, so I’ll reuse it here to answer your question. If you want to create your own anaglyphs there’s a few different methods. All involve photographing (or creating from scratch) two images of the same target object(s). The pros recommend a dual camera setup something like this: https://www.3d-image.net/ but that’s an expensive way to go. So I’ll “focus” on the single camera method …
Mount your camera on a tripod for the first photo. Ensure you have good lighting across the target object and a sharp focus. After the first photo, move the camera sideways by 60 mm (about 2.5 inches) perpendicular to the line of sight, and rotate as needed to keep the target centered. The 60 mm shift approximates the average distance between human eyes. Readjust the focus as necessary. Then take the second photo using the same exposure settings.
The next step is to combine those two photos into a composite image, side by side, using an image editor. I use Corel Photo-Paint. The composite is referred to as the “stereo image” for the anaglyph. It will be your input to any number of free online anaglyph generators. I like 3Dthis because of its intuitive interface. Some online anaglyph generators require the two photos to be uploaded separately, and you may want to use that option if you’re inexperienced with image editors.
For the anaglyph in this post, no photos were taken and no cameras were involved. It was created from scratch using my image editor. All I needed to do was create two copies of an image, with one copy having all the objects shifted horizontally to simulate the parallax seen by two eyes 60 mm apart. The amount of shift (in pixels) will determine the perceived depth of the 3D image.
If you’re working with an image containing multiple objects that have the same size (as in my example using the three Moons) you can enhance the 3D effect by scaling the objects. The scaling rule is: the size of the object should be inversely proportional to its distance. That will also contribute to the perception of depth. You’ll need to play around with anaglyphs until you get the best 3D. Like anything, there’s a learning curve.
Here’s the composite I submitted to 3Dthis:
[Note that the vertical white line dividing the halves of the image is not required for a stereo image. I added it to to make visual comparison of the two halves easier.]
If you look closely you’ll see the two halves of this stereo image are not identical. In the right half, the two largest Moons are shifted slightly to the right by amounts inversely proportional to their diameter, as if the camera had shifted to the left. That’s the simulated parallax. The smallest moon doesn’t shift much at all, so appears to be “at infinity.”
3Dthis generates a standard color anaglyph that can be seen using 3D glasses with a red lens for the left eye and a cyan (blue-green) lens for the right eye. The eye behind the red lens sees only the cyan-tinged image, and the eye behind the cyan lens sees only the red-tinged image. This is equivalent to real 3D when perceived by the eye-brain system, because each eye is getting a different signal. Of course, without the glasses, you will see both the red- and cyan-tinged images partly overlapping.
That’s all you need to know to start creating your own anaglyphs — with either a camera or an image editor. Have fun experimenting with this niche area of visual media! One final note of caution to save you some time … there are online anaglyph generators that claim to produce results from a single image using AI. I tried a few of them and was disappointed with the results. Theoretically it’s possible, but AI isn’t quite there yet.
Next Week in Sky Lights ⇒ What a Meteor Shower Looks Like from the ISS