Youtube video on color

Started 4 months ago | Discussions thread
Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 4,760
Re: Even More Fascinating

TheGrammarFairy wrote:

Ideally, spectral response curves should be similar to human vision but that's not practical...

I get the sense that there's a lot to learn from such a seemingly simple statement. Obviously, if it were possible to make the curves identical to human vision it would have been done long ago.

What is the biggest obstacle to making them similar? And are there other technologies or methods that would allow better curve matching? (I always hope the answer will be to shoot my exposures through some combo of RGB and CMY filters).

I don't know. But, you may notice that there is a lot of overlap between M and L. I would think it would be challenging to get good separation between red and green.

I would think we would rather model the sensor on the CIE 1931 color observer:

Note that this figure uses colors. So you would this weighting functions to project an observation in three primaries.

This article by Jack Hogan discusses it in some good detail.

The reason Jack looked into this was the original information from Phase One on the 'Thrichromatic'. It caused a lot of discussion and testing. Jim Kasson and Jack Hogan cooperated on modelling color responses for different CFA designs:
Personally, I think that the major change on the Thrichromatic may have been on the UV and IR filtering as I have seen examples where the Thrichromatic was showing much less purple fringing.

Best regards


Overlap on sensors is needed, else they would not be able to reproduce spectral colors.

Ideally, spectral response curves should be similar to human vision:

But, that is not really practical.

-- hide signature --

Erik Kaffehr
Magic uses to disappear in controlled experiments…

Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow