Color Filters to Improve IQ in Poorly Corrected Optics

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Online
ray_parkhurst
Posts: 3415
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:If you do this with a single image, you are accepting crosstalk between the channels. In other words, some of the energy in green channel is actually produced by blue light (that is not necessarily focused in the same plane as the green light).
Any idea the relative intensities? If we're talking 1-10%, I'm not too concerned. If we're talking 50%, then I don't see how the color filters could produce decent color fidelity.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

The graphs I've seen can reach 50% but it varies greatly between brands.

Here is a graph:
https://i.stack.imgur.com/m62Ox.png

But color filters resolve this problem completely, because you are treating the sensor as a monochrome sensor. Signal strength in all channels is from just one color.

Online
ray_parkhurst
Posts: 3415
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote: But color filters resolve this problem completely, because you are treating the sensor as a monochrome sensor. Signal strength in all channels is from just one color.
Only true if you were able to remove the Bayer color array. With the array in place, it doesn't seem that either your color resolution nor fidelity are improved much.

I suppose you could characterize the leakage of each color into the other channels, and do some sort of correction to flatten the response, making the sensor appear as monochrome even with the Bayer filter array in place. Convolution of the opposing filter responses would cause significant potential errors though since the wavelengths present are not known and can't be deduced.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Ray, if you have a pixel-shifting sensor, the problem would be solved.

Online
ray_parkhurst
Posts: 3415
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

Lou Jost wrote:Ray, if you have a pixel-shifting sensor, the problem would be solved.
True. Still probably better to use a monochrome sensor.

Lou Jost
Posts: 5944
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Yes, definitely.

Post Reply Previous topicNext topic