Lenticular 3D Print Attempt!

Lenticular 3D Print Attempt!

Not everything works, but I think I learned something from this, and had some fun!

I’ve been very pleased so far with the lenticular project. There is a fully mounted and framed version of ‘The Perfect Boiled Egg’ ready for Helen, which looks great (IMHO). So I started to think… could I use this method to print a true 3D image, just using my home printer, with folding?

The answer is YES but the practical difficulties are severe and illuminating and viewing the 3D print are challenging, to say the least. Overall I’d rate this as only a limited success, but I thought I’d write it as a blog anyway. With better practical skills, it might have been spectacular!

The diagram below shows the principle. It is like the lenticular prints I’ve talked about but the angle of each mountain changes away from the centre, by adjusting the lenticle width. Looking at the diagram, the left eye only sees the blue lines, the right eye only the red. So, if you have a Left Right stereo pair printed on the blue (left) and red (right) sets of lenticles and look at the print from the right point, you will get a full 3D image! there are limitations… you have to be at the right distance and ideally have the print set for your particular interocular distance, although I think using 8cm for this will work for everyone.


A bit of geometry and trigonometry enables the lengths to be calculated. I set up a simple excel sheet for this, and pasted the result into the figure (sorry left and right are transposed here). This was for the geometry shown on the figure. The folding is EXTREME! Lenticles are around 10cm wide and only 3cm apart at the base. Stiff paper isn’t happy being folded like this… and thinner paper won’t stay flat if you have the slightest errors. But I marched on, with that idea in mind that I’d fold it up and this beautiful 3D image would just pop out…

I prepared a test image of snowdrops, with two views taken 8cm apart, pasted below…


If you cross your eyes slightly to overlay the images, this will pop into 3D…but I wanted to do it without the uncomfortable eye-crossing.

The other EXTREME aspect of the print is the anamorphic stretch. To the eye, each lenticle will only look 3cm wide, but has to be printed 10cm wide, so it needs a 330% stretch sideways. The print will also be very long, so lots of joining needed (not using expensive roll feed for this!). The lenticles also need to be given a trapezoidal distortion, so the image looks like a constant width when viewed obliquely.

The images were combined in photoshop as before, but each lenticle was given thetrapezoidal distortion using the skew tool.


This is the ‘ready to print’ image! As soon as I put it on the print module, I realised I had been way too optimistic - it was far too large and would have needed 17 A3+ sheets of paper. Unfortunately, you can’t just scale it, it needed to be redone as a smaller image, with the same interocular spacing and viewing distance…

So, did it work this time?…. er, no, or only with optimistic imagination. The whole structure was too floppy and innaccurate. Rather than try harder, I tried simpler. Just two lenticles! Here is the ready to print image.


This worked really well. At the right distance, there was the square 3D image sitting in the middle…but a bit too smalll, only 5cm square. Onwards and upwards… recalculate for 6 lenticles… 3 left, three right, with the peak in the middle…


This also worked, but only if I stuck the paper to stiff card to keep it flat, and even then, the viewing angle was a bit limited…but the image was real 3D, you could move your head from side to side and see round it (a bit…). I can't show the actual 3D image here, you have to take my word for it!

So, sort of success… at least good enough to stop me trying it again, but I certainly wouldn’t recommend it as a display method. If you had a lot of patience and assembled the image out of pieces of stiff card and worked out how to get the illumination into the valleys and held the viewers head in exactly the right spot, it would work brilliantly. Yeah, right. However, it was fun, and really kept me thinking about anamorphic distortion and getting the left right right (or is that left?)

Thanks for reading. I hope you found this interesting, or at least amusing!