Magic Eye Maker
Type a word, pick a shape or add a photo, choose a pattern, and send the hidden 3D picture to someone to find.
Make a Magic Eye picture
Everything happens in your browser. Your photos and words are not uploaded to our servers.
Your picture
Making the picture…
Drawn for this screen. Look through it until the two dots become three.
Is it doubled?
If you see four dots, or the shape looks smeared sideways with double edges, your eyes locked two repeats apart. Look away, relax, and try again: stop at exactly three dots.
1. What to hide
2. Pattern
3. How it is viewed
4. Save or send
1080 × 1350, looks right on a phone and on a computer
Every saved picture has a strip under it with the two guide dots and “made with owl3d.com”, and a small “owl3d.com” hidden in the 3D, bottom right, that pops out when you see the picture.
How to use it
- Choose what to hide. Type a word or a name, pick a shape, add a depth map, or try a photo.
- Choose a pattern. Colour dots are the easiest to see. "Your word" fills the picture with a name; "Your photo" uses a photo as the wallpaper.
- Look at it. The picture is drawn for your screen. Look through it until the two dots under it become exactly three (four, or a doubled, smeared shape: relax and try again). Press "Show what is hidden" to check.
- Save or send it. Pick a size and download, use your phone's share button, or copy a puzzle link: the person who opens it gets the same puzzle with the answer hidden until they ask.
What each input does
Words. Short words in capital letters are the easiest to see. Up to 24 letters; long text is split over two lines.
Shapes. Fourteen simple shapes, from a square (the easiest) up, and 22 sculpted scenes we modelled for this: animals, sea creatures, dinosaurs, space, words.
Depth map. A grey picture where white is near and black is far, such as one made by a 3D program or a depth-map tool. Tick the box if yours uses black for near.
Photo (experimental). The tool works out the depth of your photo on your device, separates the person or pet from the background, and smooths it. Be ready for one thing: a Magic Eye picture shows a shape, not a face or colours. Whoever finds it sees the outline of a person or a dog standing out, not who it is. One person or animal against a plain background works best; a crowded group photo becomes one bump. To keep the family recognisable, choose the pattern "Your photo": the picture is then tiled with small copies of the photo, and the 3D shape hides inside it.
Sizes, and why they matter
The pattern repeat has to be about the right physical width for the eyes to lock onto: around 3 cm on a computer screen or on paper, less on a phone, and never more than the distance between your eyes. So each size here is designed for where it will be seen:
| Size | Pixels | Made for |
|---|---|---|
| Phone screen | 1080 × 1620 | Looking at on a phone held upright |
| Share | 1080 × 1350 | Chats and social media: works on phones and computers |
| Computer screen | 1920 × 1080 | A monitor or TV, full screen |
| 3300 × 2400 | A4 or Letter at 300 dpi, 31.5 mm repeat |
If a picture is hard to see on a big screen, it is usually too large: look at it smaller or from further away.
Frequently asked questions
How does a Magic Eye picture work?
The pattern repeats across the picture. Where the hidden shape is, the repeats are drawn slightly closer together. When your eyes line up two neighbouring repeats, the closer ones appear nearer to you, and the shape stands out. This tool uses the method published by Thimbleby, Inglis and Witten (1994), with its fix for hidden surfaces, and computes each line at several times the resolution so curves look smooth instead of stepped.
Parallel or cross-eyed: which should I choose?
Parallel ("look through") is how most Magic Eye pictures are made. Choose cross-eyed only if you, or the person you send it to, find crossing easier: a picture made for one way looks inside out the other way.
Can the person I send it to see the answer?
With a puzzle link (words and shapes), yes: they press "Show the answer". With a picture file, they can add it to the Magic Eye solver, which shows the hidden shape.
Is my photo uploaded?
No. The photo, its depth and the picture are all made in your browser. The depth model (Depth Anything V2 Small) is downloaded once, about 27 to 50 MB, and kept by your browser.
What is the mark on saved pictures?
A strip under the picture with the two guide dots and "made with owl3d.com", and a small "owl3d.com" hidden in the 3D in the bottom-right corner. It is part of every saved picture.
Learn more
How to see Magic Eye pictures · Magic Eye pictures to try · Magic Eye solver.
Magic Eye is a registered trademark of Magic Eye Inc. Owl3D is not affiliated with Magic Eye Inc.