Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Make a "Magic Eye" image using Excel (divisbyzero.com)
115 points by noiv on Dec 8, 2024 | hide | past | favorite | 56 comments


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.


I drive Curiosity and that's how I often view our raw downlink images. They aren't stereograms, but the left and right eyes are next to each other on the report, so I just cross my eyes to get the stereo and scroll the page. Works great but despite my trying to spread the good word I think I'm the only rover planner who does this.


> “I drive Curiosity”

“What kind of SUV is that? He can’t mean… Oh, he does…! Holy s**t”


"My other car is Curiosity."


Wow! I just checked your CV (we were at Cambridge at the same time :) ), and I will never dismiss a Geography degree again!


Honestly... it turned out alright but I wish I had done something more rigorous in undergrad. I had to do a lot of math self-teaching to catch up for grad school. The most useful stuff I learned at Cambridge was from the caving club and from my mathmo / compsci housemates.

Also, thanks for the reminder to update my cv X-D


Too late.


Stupid funny-pages trick: "Spot the Differences", you can cross your eyes and the things that "shimmer" are the differences. https://en.wikipedia.org/wiki/Spot_the_difference#Solving_te...


It might be overkill but a VR headset would make that significantly easier for people.


We actually use NVIDIA 3D Vision 2 in 3D simulation environments during planning. It's great but quite hard to keep it working because NVIDIA doesn't support it any more. Our software falls back to analgyph for systems that can't run the 3D Vision and I end up using anaglyph a lot and it's totally fine.

We've experimented with VR a bit but it hasn't caught on yet. We've also been trying a few other 3D display technologies.


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


Voight-Kampff test but it’s just staring at stereograms with sweat glistening on your forehead.

It’s unfortunate that so many humans were initially retired, but it was also helpful in motivating the population to finally figure out Magic Eye.


Oh, so the image was a tortoise upside down? Now I see it


Tortoise? What’s that?


You know what a U+1F422 is?


You mean that all theses stuff on my screen are also physical beings? Lol. And while we are it, women and Real World exist ?


I'm trying man. I just don't get how it works


Focus your eyes as if you're looking at a more distant object. You know how if you look at something a little further away and hold your finger in front of your eye it doubles? With a magic eye, the image is repetitive, and you're trying to get the doubles to overlap and "lock" together, tricking your eyes into thinking they've focused correctly. The repetition is varied in such away that it makes the illusion of a 3d object.

The kind of weird thing is that you can achieve the same effect with any repeated pattern - floor tiles, a fence post, etc. But there's no underlying illusiory image, so you just see basically the same thing but everything else is focused wrong.


I used to love paper books of these as a kid. I was taught to start with my nose practically touching the page and slowly backing my face away without changing my focus. Eventually I could do it without doing so (and the skill never faded!) but it's the best way to get started


For me, the secret is to relax my eyesight a bit. Kinda try to ignore the image and think of something else. "Space out" while staring at it.... and suddenly boom, you're in


I think you can decode a stereogram into the depth map with like 5 lines of Python, but it's been a while since I've played around with this stuff.

Meaning, maybe GPT can't (yet) see magic eye images, but it shouldn't have trouble building itself a prosthetic that allows it to see them.


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...


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?


I recently learned that there are two views: https://www.reddit.com/r/ParallelView/ vs https://www.reddit.com/r/CrossView/

And there is definitely a difference between them. If you try to view a cross view image using parallel view, it will look weird and not be easy to focus. Maybe the egg crate image was different?

Here's a quick test: https://i.redd.it/g5ilwgk99r781.jpg

Parallel view is easy for me but it takes a bit of effort for me to see cross view. For cross view, I start by looking cross-eyed at my nose and then try to see the image without fully uncrossing my eyes.


I have the opposite experience. Cross view is easy for me but focusing parallel view is very difficult to impossible.

When I try to relax my eyes to look past the screen to start the parallel view (I think that's how it is done?) the image is too blurry to resolve. When I let my eyes adjust that, they fall apart to the separate images.


Thanks for the link, I saw the Tokyo parallel view beautiful crispy for the first time, now I can't do it again, I guess my eyes became tired and need some time


I think the egg crate was harder because there were no sharp edges to latch onto.


I've always been pretty quick to see the hidden image, but I could see these faster than ever - almost instantaneously, even the first one. I wonder if it's something to do with me having stared at a screen for a decade since the last time I saw a stereogram...


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.


That is incredible, thanks for sharing. For others trying to find it, it's "herrmann2", not herman2. To try it yourself, the following worked for me in WSL:

  wget https://www.ioccc.org/2001/Makefile
  wget https://www.ioccc.org/2001/herrmann2.c
  wget https://www.ioccc.org/2001/herrmann2.cup
  wget https://www.ioccc.org/2001/herrmann2.hint
  wget https://www.ioccc.org/2001/herrmann2.ioccc
  cat herrmann2.hint    #instructions
  make herrmann2
  ./herrmann2 < herrmann2.ioccc
My mind is boggled. I get the basics of how autostereograms work, but the fact that it works so well with ASCII text is mindblowing.

Edit: If you're lazy, I made a quick Gist that does the above, so you can just do

  curl https://gist.githubusercontent.com/justusthane/a8c27c72350793ac452b86cc665a268b/raw/72aa6bf37689ea0cbbb70fa70fbd758c3a1fde3a | bash
(Obligatory "It's not safe to pipe internet stuff into a Shell" disclaimer)


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.


When I was younger, my ophthalmologist was still testing for depth perception through binocular disparity (the parallax effect, I think). I could pass those tests (although my mother is in the 20% or so of the population that can't do it), but I never could see those stereoscopic images for some reason.

I haven't been tested for many years, but I assume I have lost this type of depth perceptiion due to monovision after cataract surgery. I still have functional depth perception, but it must rely on other clues these days. And I still can't see stereoscopic images.

At least there are tools now that can undo the process by which these images are produced.


I struggled at the beginning and was almost certain I was one of the small percentage who can't see them, but after working on it for a while I managed to see them. Now I use them to relax my eyes with the parallel mode. I remember not being able to properly align the two images on a stereo microscope without pulling outward one of my eyelids (which in turn adjusts one eye). I did the same for the magic eye until I was able to see the stereoscopic image form, then I no longer needed to do that.


I was able to perceive all of them, using the "focus on something behind the plane of the monitor surface" technique (rather than the "bring your face up close and slowly move away while keeping your eyes focused the same" technique, which I find has been a bit hit-and-miss).

I found that the π and ∞ stereograms were easier for me than the graph ones, as it's easier to snap to focus once you recognise the general outline of the shape. The graph ones were a bit harder for me, although that may just be the colours used.


Works easily for me. There are two types of stereograms (cross-eyed and parallel) so perhaps you had success with the other type. I believe this are parallel (aka "wall-eyed".)


They're a mix. I find the mathematical symbols ok but they don't jump out very much which seems to make my eyes reluctant to stay with the illusion. The first (green) surface one is one of the best magic eyes I've ever seen, but the second (pink) one I can't get to stay at all.


These posters were around in the 90s. After a few tries I “got it” and was able to see them. It was frustrating initially.

Popular enough to be on Seinfeld.

https://youtu.be/Uy9D0lO_0y0?t=38&si=f9zczm_noVd9hsmN

“Computers make these.. big computers”


I remember when these things first came out and they were everywhere. I've never ever gotten them to work for me.


I used to be able to see these but as my nearsightedness progressed at some point I lost the ability.


Seems not everyone, and it appears that seeing them is a combination of skill and vision.

Never been able to see them, but my eves have different enough vision that it is hard to make them work together at certain distances.


The intended way (when image floats above background, closer to you) - no. I can see images if I cross eyes and let patterns merge. However, this way images appear below background and are harder to recognize.


Every single time I tried it in the past, I failed. I never got them to work for me.


i get colourful snow and a headache trying to make my eyes do things. it just blurs and unblurs :'(


Working on my computer


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?


That’s an interesting way of looking at it. When this happens, it means your eyes are crossed too much. I solve this by relaxing and catching it before my eyes get fully straight. Doing this is usually easier than seeing it the first time.


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

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


> 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.


I always had this problem back in the day when they were in newspapers etc. I didn't really get what people were seeing, because to me it was all in reverse. I looked at these on smaller screens last night (phone and tablet) and I could see them! But just now I tried on my 27" workstation monitors and I got them reversed!

People have pointed out that these are "straight eye" rather than "cross eye" ones. So my theory is on a big screen these are too wide for my eyes or something. I can always go cross eyed (by looking at my nose), but I probably can't go "wide eyed".


You should focus behind it instead of looking cross-eyed.


I believe this depends on if you focus "beyond" or "in front" of the image


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


Could we... make a gif of these?



Man, that looks like the hardest possible gif someone could make and still have it work


Yeah I reckon the scrolling isn't making it easier to focus!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: