Additive color printing brainstorming

[bvy]

Of course, I'm struggling a little to understand how (or rather why) this works. The instructions are that I project the image through a diffusion filter, and then overlay on the paper a graded red (green or blue) filter having different densities, representing different exposure times for red light, process the paper, then evaluate. Why is it that this is enough information to determine the correct exposure time for red? More specifically, why do the colors in the image not matter? For example, if I take a picture of a solid colored wall... ?

 

[Photo Engineer]

Well, red is the slowest speed, and so the other two layers swing around it regarding balance and exposure time.

PE

 

[bvy]

Thanks. I guess I'm looking for a more long winded explanation. Having trouble finding links or references. For example, when you diffuse the image, you're "averaging" all the colors -- speaking loosely. Then you take readings of that light through graded filters, one for each of red, green and blue. What's going on here? How does this work?

 

[Photo Engineer]

Ok, the method can fail. But, on average a scene is considered to average gray. This is how the balance is set for a color printer which uses RGB filters (or now LEDs). So, the image is set up with a standard and then everything afterwards is printed based on that average. It works most of the time. There are few reprints in a production photofinisher due to bad color.

PE

 

[bvy]

Ok, the method can fail. But, on average a scene is considered to average gray. This is how the balance is set for a color printer which uses RGB filters (or now LEDs). So, the image is set up with a standard and then everything afterwards is printed based on that average. It works most of the time. There are few reprints in a production photofinisher due to bad color.

PE

Thanks. Maybe that's just it. I'm having trouble understanding why even most scenes would or should average to gray.

 

[Photo Engineer]

It assumes a gray based on an average, but has a bias detector which we do not have and thus we must do it manually and that is why 3 color printing on color paper is so hard. Dye Transfer was a bit easier in that you could make 3 matrices and these had neutral scales which could be adjusted. This was rather easy but a bit expensive.

PE

 

[DREW WILEY]

The Phillips tricolor enlarger is far too weak for anything serious. It's based on rheostats - basically a Jurassic design.