VERY STRANGE - Apparent CCD color splitting
 

So this is odd. I was shooting with a GL2, a passed-out guy across the street. About 75 feet away, with zoom pretty close. A bicycle passed in front of the guy, and the image suddenly jumps in its apparent position. Now, I noticed this because it looked like the camera moved, but upon frame-by-frame inspection, I saw that the image jumps down, turns gree, then blue, then jumps back up to red, then after the bike passes, it's normal.

See the sample here: http://www.quidditypark.com/COLORSPLIT.mpg ... it's about 9 megs.

So can anyone tell me the physics/mechanics of this issue?

Thanks,

DJ

I am having trouble playing your video for some reason. Is there any specific codec I should download?

Fascinating!

If you repeat the process but shoot through chain link fencing, and focus on an object on the other side of the chain link fencing, the light from the object passes by the out of focus chain links but it is ever so slightly softer. I can imagine that if I was to "dolly" the camera across the front of this chain link fencing I would see a "wave" of sharp>soft>sharp>soft focus - leaving the impression of the object "moving". Here you and the object stay still but the cyclist - my chain link fencing - is moving. The result is this wave of movement. But! Your movement is UP and DOWN! Weird . . . Is it at all possible that the cyclist had some semi-transparent windshield? Is this what we are seeing? An aberration as the object light is diffracted UP and DOWN?

Failing all this, could the sync hing of the 3 CCDs gone? UGH!

Another thought is, is this the resultant effect of the warning that Canon say about NOT have the OIS - steadying mode - switched on while on a tripod?

Truly fascinating!!!

You had ANYTHING like this before?

Grazie

I'm glad you found this as interesting as I did. I've never had anything like it before, and after looking at it a few more times, it really seems like there is not a refracting lens.

Also, check out the fact that it isn't just up and down, it is in a slightly triangular configuration. Is this not the same as the CCDs? It's two low--the green and blue--and then red is actually higher than the baseline. The colors don't look as vivid as they do on a monitor, but the refraction or color separation is very clearly visible.

DJ