Color Sytem Engineering part I
This is the port of part 1 from Photo Net posted years ago.
I hope it reaches a suitable audience here as it did there at that time.
To carry over some comments from PN, yes, there are better ways to design photo products today. But, since they are not available to us, I used the methods available to Hanson, Vittum and other early workers in the field.
And, I am not the best of persons in this area, just as I am not the best to teach emulsion making. However, I can do it and I am willing to do it!
We are going to design a hypothetical color negative film. Much of what I say is applicable to reversal films. Major exceptions include masking and contrast.
First, we have to have an aim. We work backwards from the print. Since any viewable material will have a toe and shoulder, and since the eye integrates them into the overall experience to obtain the final result, and we know there is a 1:1 density to exposure relationship in the real world, we know that the print material will have an overall contrast greater than one. This value actually works out to about 1.5 - 1.7 based on B&W experience. Using the same B&W experience, this puts the paper's origianl contrast at between 2 and 2.5, so by simple division we come up with a film contrast of about 0.6, the range being 0.5 to 0.7 on average.
We also know that an average scene can encompass a range of about 1.8 log E or 6 stops. (6 * 0.3 = 1.8)
So, we have a potential aim curve of slope 0.6, and straight line portion of 1.8 minimum. Lets give the customer some latitude and extend that curve to 2.4 or 3.2 Log E for over and underexposure latitude.
Ok, progress. Now, if we want overexposure and underexposure latitude, the EI has to be correct, but the threshold speed must be faster by the amount of latitude you wish to allow underexposure without being completely on the toe.
Lets assume a 100 speed film is our aim. Then the emulsions have to be far faster than 100 to achieve this final speed. First, the EI is going to be 100, second, the absorber dyes and antihalation are going to eat up speed, and third, turbidity of couplers and overlying layers are going to eat up speed due to internal reflection and absorption of light.
Pracitically then, lets assume that the blue layer, on top, has to be 200 speed, the green layer coming second, has to be 300, and the red layer on the bottom has to be 400 to account for these losses. This means 3 different emulsions with graded sensitivity, grain size, halide content and etc. And, being on the bottom, the cyan (red) layer will be the slowest to develop due to grain size, halide ratio, and diffusion effects, so we have to keep them in mind when designing the entire package.
Now, we coat our first single layer experiments and get 3 B&W coatings of B, G, and R emulsions, and they show great development in an MQ, but practically no development in a color developer. Reasoning this out, we find that without coupler to scavenge oxidized color developer, a color developer is very slow in development due to equllibrium effects. Oxidized color developer build up acts to restrain further development, so we have to either coat couplers to start with or rely on MQ tests.
Relying on MQ tests is dangerous, as MQ results are based on visual silver density, but color formation is based on mass of silver developed. One mole of silver can produce either 1/2 or 1/4th of a mole of dye depending on coupler type. This is a critical factor.
So, you decide to go with color tests for your single layers and coat single color R, G, and B coatings with C, M, and Y dye forming couplers. (which ones you pick is a real problem. You have an entire group out there testing them for activity, dye stability and dye hue for product families)
You coat and process your single color experiments and find that the results give you good D-min and a D-max of 3.0, with a latitude of 0.9 Log E. Whoops, you want 1.8 to 3.2 for your film. Oh, the contrast is about 1.5 as well. Back to the drawing boards. It seems that to achieve even the minimum of 1.8, you will need to blend at least 3 emulsions in each layer to get the latitude you want. For example, in the blue layer, the fast component will be ISO 200, with a latitude of 0.9 log E, and the medium will be ISO 25 to blend in with the shoulder of the fast, and the slow will have to be ISO 3 to blend in with the shoulder of the medium. In other words, blending 3 emulsions make the curves additive and they must be separated in speed by the speed width of the latitudes of each emulsion.
So, by coating a fast medium and slow each at 1/3 of the desired silver rate, we get a curve that is 3 x .9 or 2.7 log E in total latitude. Just about where we want to be. And the contrast is about 1/3 the original or about O.5. (Looking at those emulsion speeds, do you see why a color negative film can be exposed at 200 or 50 if the ISO is rated at 100? Hint.. Hint.. Hint..)
By now, we have 3 single layer color coatings that give us a latitude of about 2.7, contrast of about 0.5 and mid curve speed of about 100 B, 150 G, and 200 R, with threshold speeds of 200 B, 300 G, and 400 R.
We now coat a coating with R first, G, then B in a 3 layer package and we process it. What do we see? Blue looks normal, G is about 1/2 the contrast and about 1/2 stop slower, and R is about 1/3 the contrast or less and about a stop slower. The dye images look like mud. I would call it Cyan, Grape and Pumpkin. Not Cyan, Magenta, and Yellow. Something seriously went wrong.
Well, enough for now. I'll wait for comments from you all and answer questions as we go along in this thread. Lets see if there is any interest!