Talk:Shepard tables

The illusion is strong, but can you prove they are the same?

Nice illusion, but am looking for proof that the parallogram's are identical in the main article. Image distortion is unfair - I can convert an ostritch to a cube with enough image processing. --2600:6C48:7006:200:B056:6066:1296:EF0B (talk) 02:19, 5 March 2019 (UTC)

The classic proof is to print out the image, trace one of the parallelograms and cut it out from a piece of paper. You will see that the same cut-out paper fits exactly over the top of both tables. It really is a strong illusion, isn't it? HouseOfChange (talk) 04:12, 5 March 2019 (UTC)

I tried that earlier. Need WP proof. I did capture image and rotated both 45 and 90 degrees - I still don't see it. Maybe you got me, but I want proof.

2600:6C48:7006:200:B056:6066:1296:EF0B (talk) 04:54, 5 March 2019 (UTC)

Try this. Enlarge the image so you can get a good look at it. Measure each side of one of the tables with a ruler, and write down their lengths. Now use the same ruler to measure the two sides of the other table. You will see that each short edge is the same length as the other, and each long edge is the same length as the other. There are several animations online also, showing the equivalence. One of them is linked from the article's External Links section. HouseOfChange (talk) 05:45, 5 March 2019 (UTC)

The problem is that the linked picture is clearly not of parallelograms. I expanded the picture and measured the pixels. The top line moves up as it moves to the right and the bottom line moves down as it moves to the right. The illusion is a real thing but in this case the picture is distractingly wrong. TheJabberWalking (talk) 06:15, 5 March 2019 (UTC)

I found a better version here, but I don't know if it's free or not. It wouldn't be hard to recreate. --Lasunncty (talk) 07:24, 5 March 2019 (UTC)

We already have an image showing the rotation at the Optical illusion page - see here. Mr Ernie (talk) 09:19, 5 March 2019 (UTC)

I've measured the two "parallelograms", and they are neither parallelograms - their "parallel" sides are of different lengths - nor equal. I presume that copying of the image has caused it to become distorted. The illusion is powerful and beautiful, but we should have a proper drawing of the image that does not cheat. Chiswick Chap (talk) 10:24, 5 March 2019 (UTC)

I've taken the original drawings and rotated the images and also cropped out all but the table top surfaces. I've added this to the article because I reckon there is no research here, original or otherwise, just a derived image :-). Thincat (talk) 10:56, 5 March 2019 (UTC)

Thincat that's a good addition -- it really is an amazingly strong illusion -- but to those demanding proof: the fact reliable sources are saying they're the same is the proof Wikipedia needs. valereee (talk) 11:13, 5 March 2019 (UTC)

I love it. All you have to do is hold a ruler up to the damn things and measure them. My ruler says they are not equal. But that is original research, and as valereee correctly points out, "the fact reliable sources are saying they're the same is the proof Wikipedia needs". Hilarious. However, in this instance we are actually looking at homemade approximations, not Shepard's originals – so reliable sources don't apply. Milkunderwood (talk) 06:59, 25 May 2020 (UTC)