from Hacker News

Make a "Magic Eye" image using Excel

by noiv on 12/4/24, 7:36 PM with 56 comments

  • by ndr42 on 12/8/24, 11:37 AM

    I had my eyes focused to see the hidden image the same way I always do. But this time I was scrolling the picture on an iPad and the next one was coming up and it was directly in focus and the next one too.

    This was real incredible - no need to focus again - it is like living in this hidden world.

    If you never experienced this I recommend to try it - at least for me it was a wow moment.

  • by BiteCode_dev on 12/8/24, 12:02 PM

    ChatGPT can't read stereograms yet, so this would be a nice captchas, if only most humans could read them too.

  • by thih9 on 12/8/24, 10:04 AM

    Is there an app for “decoding” magic eye images like these? The output could be a 2 frame gif showing perspective from each eye[1], or a single image with depth of field simulated with some blur.

    Edit: found this https://piellardj.github.io/stereogram-solver/ , it works well for single layer images; but it isn’t great for 3d surfaces. Another one, though not online, seems better: https://github.com/MikhailPedus/AutostereogramSolver

    [1]: https://www.vexels.com/blog/stunning-3d-effect-with-gif-2-fr...

  • by gcanyon on 12/8/24, 1:37 PM

    Have stereograms advanced in some way over the past 30 years? I know how silly that sounds, but: I remember back in the day when the fad first broke, spending minutes, many minutes, trying "get" it. And having to focus (heh) on focusing my eyes. And sometimes failing. I haven't looked at a stereogram since then until now.

    And these were easy. The first one literally took 15 seconds to "drop in." The second maybe ten. The third was near-instant. The only one that gave me any trouble was the continuous function one like an egg crate, with no sharp edges, just dropping down and up. That took maybe twenty seconds, and once I recognized what I was looking for, it was easy.

    So has the technique changed/improved? Or has my brain changed?

  • by tzs on 12/9/24, 6:57 PM

    One of the obfuscated C contest winners in 2001 was a program that generated ASCII Magic Eye images of text strings, suitable for viewing on a terminal.

    It is the herman2 entry here: http://www.ioccc.org/years.html#2001

    Note that the source code is itself such an image.

  • by a1o on 12/8/24, 10:11 AM

    Is everyone able to see this? I have even played a game on itch that had these in real time, but these ones aren’t “working” for me.

  • by TheJoeMan on 12/9/24, 12:40 AM

    This is neat, I’m pretty good seeing magic eyes. Something happened on my second visit to the site though, I see the Pi almost double-vision - four vertical legs instead of two, and the top bar has a second narrower bar superimposed on top, essentially 3 depth layers instead of usual 2 layers. Is it possible there’s harmonic focal point distances?

  • by matsemann on 12/8/24, 9:59 AM

    > This one shows a surface: the graph of a multivariable function.

    Almost think textbooks should utilize images like these, worked very well.

  • by starshadowx2 on 12/8/24, 12:36 PM

    > You shoud see a flat background with one large copy of the letter π floating a little above the background (closer to you).

    To me these all look like they're reversed from what this says, like they're further away from me behind the flat part.

  • by wolpoli on 12/9/24, 9:10 AM

    Anyone else noticed the extra green dot in the pink (fifth from the top) image?

  • by aclindsa on 12/8/24, 1:10 PM

    Could we... make a gif of these?