The subject of seeing unusual colors reminds me of a project I’ve wanted to do for years. Alas, it’s the sort of project that requires specialized equipment, and one would need to really be committed to do it properly. Here’s the gist of it:
Given that the human mind is capable of extraordinary feats of perceptual remapping and interpretation, it is plausible that we could map a wider spectrum into a narrower one. Specifically, suppose we set up three cameras: one that sees in infrared, another that sees in R,G,B, and a third that sees in ultraviolet.
Then we remap the resulting five-band image into a narrower spectrum: Infrared becomes red, red becomes orange/yellow, blue becomes cyan, and ultraviolet becomes violet.
Clearly information will be lost in the course of chromatic compression. But will our brains compensate? Will we be able to use our higher level knowledge of the world to perceptually reconstruct the full R,G,B portion of the spectrum in our minds, thereby freeing us to “see” the infrared and ultraviolet bands?
If so, this would be a way for people to see beyond the bounds of visible light.
I tried this out (well, the infrared end anyway) and posted the result on my blog. I’ve had trouble posting links here before, but if you click the link to my website it should work.
I mean, click my name and it should link to the correct webpage with the images.
I think this is reasonably doable. To approximate, I suggest buying a cheap 3ccd camcorder. It will contain a trichroic prism, which splits the light via a pair of dichroic prisms. It might be possible to split the prisms apart, remove their coatings and replace them with dichroic filters for UV/Vis and Vis/IR instead of R/G and G/B. Alternatively, if you don’t mind having IR, Green, UV, then add notch filters to remove red and blue in front of the lens (if you can find them!).
Alternatively, get a sport camcorder with a back-thinned CCD, like the Sony HDR-AS15 (to pick the first search result). As it can record at 90fps, you should be able to synchronise a spinning filter in front of the lens, and post-process to merge the interleaved colour frames. This would require removing the filter from the sensor, which could be fun. Given a brightly lit subject (with a broad-spectrum lamp), even if 99% of the light is blocked, you should still have enough light. You may need to stop down the visible to match the UV & IR to enable the white balancing to work.
Very cool Doug!!!!