Posted on 15 May 2011 at 08:22h

Congratulations to Azerbaijan for winning the Eurovision song contest. I didn't watch the whole thing but did see the start and was very impressed with the stage that had been set up in Düsseldorf. The whole back of stage seemed to be one giant TV screen. The way the Earth appeared during the Finish entry was particularly impressive. This morning I was curious enough to find out who had won and what the song was like so I looked it up on the Eurovision website. I can't say that I like the song because I don't like it however the lighting did catch my eye particularly the effect in the photo. I can't be sure but I don't think these fringes would have been visible to the live audience,  think it is a Moiré fringe due to interference between the pixels on the screen and the pixels in the camera, as the camera panned out the fringes shifted. You can see it here after 58s (turn the sound down if you don't like the music). I would like to see Pink Floyd or Jean Michel Jare with a light show like that.

Ever since writing this blog entry I have been thinking of how this effect is produced. First I tried taking a photo of my blank computer screen to see if I could get fringes. The first picture worked beautifully but after that the fringes just disappeared. I have no explanation for that but here is the photo when it worked.  As mentioned previously I think this is caused by the fact that both the object and image are made of pixels, when they line up you get a bright spot when they don't it's dark. Next i started playing around with dots. I drew a load of them in paint then superimposed them. You can clearly see the effect when the spacing of the dots are not the same in the screencast below. I guess this is a sort of negative version since when the spots coincide it gives a small spot but when they overlap you get a more prominent big dot. if you want to play with my dots you can download them here. Open in "paint" select one of the grids and slide over the other (make sure the selection is transparent).

Well I just had to try reversing the colours to see if I could get the true effect rather than a negative and it worked a treat. I can see this is a nice analogue of beats. If so the spacing of the fringes should be a function of the difference in the spacing of the two sets of dots.

Here is what you get if you change the spacing.  I tried using the beats equation to find the fringe spacing and got an OK result but not as close as I would have liked. I can see a nice practical coming out of this.