I have never been able to see what is hiding behind the variegated patterns of the “magic pictures”.
No matter how hard I strained my eyes, nothing worked.
Today, instead of seeing, I strained my brain, and this is what happened:

Idea
The main requirement for viewing “magic pictures” (stereograms) is the ability to focus the view “behind the screen”. In theory, in this case, each eye will be directed to different parts of the “magic” picture, and two identical images will fall into the brain, only shifted horizontally. Let's try to simulate this process.
Implementation
We arm ourselves with the free Paint.Net graphic editor (or another suitable one), and open the test image:
')

A small experiment showed that “seeing” people distinguish a latent image even in a black and white picture. Therefore, select the menu
Adjustments - Black and White , we get a black and white noise a la TV :)

Now we will create the second layer, it will be an “image from another eye”:

Open the properties, please display the difference between the layers (
Mode -> Difference ):

The result is a black rectangle. But as soon as we begin to gradually shift the image in the upper layer (
Move selected pixels , then cursor arrows) - the noise returns, and after a while something sensible appears from it:

We continue to move (and feel at the control of the
scanner :)









From these “cuts” one can understand which part of the latent image is in the foreground and which part is in the background. Probably, the combination of these pictures creates a sense of the three-dimensionality of the object in the “seeing” (the theory can be
read on iXBT ). The silhouette of the hidden image, however, is encoded separately: a few more clicks of the arrow - and voila:

The described method easily copes with captcha:

Result
Now you can see with your own eyes
[censored] on the main Habr ! :)