[popo]

There are some other references to this on the 'net but I thought I'd make my own anyway so I can check out what certain filters do.







Upper colour bar is a standard 600D response to a torch I have. The three mono lines below that are the RGB split out. Below that is a 450D with the removable filters removed, thus passing IR and maybe some UV too.



The amplitude isn't calibrated to anything but should be relative within a colour channel, when it's not clipping anyway. I need some way to work out what wavelength each point is...

[Sylvain]

Please don't burn me, but would an enlarged spectrum thru removal of filters translate to a wider range of colours in the visible spectrum too?

[Brightcolours]

[quote name='Sylvain' timestamp='1335350005' post='17746']

Please don't burn me, but would an enlarged spectrum thru removal of filters translate to a wider range of colours in the visible spectrum too?

[/quote]

Not sure how you mean this, but the normally invisible colours get stored as blue and red in the RAW data. So they do interfere with the visible colours.

[popo]

The three channel colour filter is supposed to represent human vision, so as such that is all it could ever hope to deal with. By removing the filter, we can increase sensitivity within the visible spectrum (assuming you use a different sharper filter to cut off IR/UV) but the balance will start to deviate from what the eye sees, even if mitigated by colour balance adjust.



Also it is not quite as simple that IR/UV are only stored in red or blue channels respectively. It can be quite clearly seen that the green channel is still weakly sensitive to IR, while the blue channel does cut out the nearest IR, it increase in sensitivity again at deeper IR. On the UV end, possibly there is some residual sensitivity in the red channel. Of course, none of this is a real issue with the original filters in place. This only becomes interesting if you're going beyond that. This was primarily for me own information as I recently got some really cheap IR filters off ebay and I wanted to see what they really do.



I should add I found a problem with my test method above. I forgot that for most lenses the IR focus point is not the same as for visible. I did focus in visible, so the IR spectrum presented will be blurred.