Next generation Foveon sensor

Started 7 months ago | Discussions thread
saltydogstudios Regular Member • Posts: 282
Re: Next generation Foveon sensor

Scottelly wrote:

Roland Karlsson wrote:

D Cox wrote:

Roland Karlsson wrote:

D Cox wrote:

But if you go for narrow-band detection, as in a Bayer sensor, you lose the colour quality of the Foveon sensors.

That is an interesting comment.

The Foveon sensor has a huge overlap.

In theory, this means that if you have an area of similar hue, then you can resolve more hues. So, when an RGB camera shows green leaves, a Foveon camera might show leaves with a nice texture.

That is theory. I do not know if anyone have tested the theory.

I posted a pair of images of a graded spectrum filter a while back, and they do show what I expect, as do normal photographs.

The problem is that the only Bayer camera I have access to is an obsolete model (NEX-5N), so people could argue that current cameras are different.

I want to shoot the images again at some point, taking more care over stray light.

I think Sony had a camera with both blue and bluegreen (emerald) filters in the Bayer array, in order to improve green hues.

I wonder what that was like to process. That makes me think of a sensor with red, blue, green, and yellow filters instead of red, blue, and two green filters per set of four. Someone suggested that was unnecessary though. I think a CFA with even more colors might make sense, such as a filter with all the green filters they have now, but red, pink, yellow, violet, blue, and light blue.

Hmm. Considering the intended output is RGB and that the eye is only sensitive to RGB - in that there the eye has only 3 color receptors, I don't see the point in a camera that has more color receptors than the eye?

My ex girlfriend was getting into polymer clay at some point & I decided to investigate color theory a bit. Whether or not you could indeed get any color by mixing just 3 colors.

What I found was that - sure, theoretically yes RGB monitors and CMYK printers could produce any color - within reason and with limitations.

But when it came to mixing clay, or paint - you didn't get a pure color, a bit of greyness creeped in. They type of greyness that would eventually turn into a muddy brown if you kept mixing.

Basically color mixing was triggering two receptors in varying amounts by showing it two different frequencies. But it wasn't as good as triggering the two receptors from one single frequency.

So you couldn't just buy 3 colors of clay and mix any colors - they would always come out a bit muted.

Again - RGB monitors and CMYK printers don't seem to suffer from this as much.

What you're proposing is - ignoring the color names - that we introduce a sensor that has pixels dedicated to specific color frequencies. So you get more different color frequency readings - but still spaced out in such a way that to get full color information you need to combine across multiple colors.

"Well this pixel triggered this much burnt ochre, and that pixel triggered some amount of cornflower blue"

Do we gain more information from this than we lose?

-- hide signature --
 saltydogstudios's gear list:saltydogstudios's gear list
Ricoh GR Digital Ricoh GR Digital III Sigma DP2 Merrill Sigma DP3 Merrill Epson R-D1 +5 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow