Color Negative Printing

When printing a color picture, the machine "reads" the entire frame

and chooses the filter combination that it considers fit for the

respective image. This is something I never managed to understand:

why? Why does the filtering depend on the image itself? <br>

Let me give you an example: if the picture contains 80% green and 20%

red, the machine will choose a filter that will give a perfect green,

but the red will be slightly off. If the frame contains 80% red and

20% green (but exactly the same shades as in the previous case), the

machine will choose another filter combination, which will result in

a perfect red and a slightly wrong green. When you look at the

negative you see that those two frames, apart from the colors'

distribution, are identical. They both have the same shade of cyan

and the same shade of magenta, yet the final results look different.

Why?

<br><br>

From all green shades in the universe, the photographed object has

one and only one. A unique one. This particular shade of green

corresponds one particular shade of magenta. One and only one. Add to

this the color of the mask and you'll get a unique color. There is

only one shade of green that produces that color on that negative. No

other color in the universe would produce exactly the same color on

the negative. If you select a filter that is exactly the

complementary color of the mask, plus a little more blue to

compensate for the color temperature of the light, you should obtain

the original shade of green. Right?

<br><br>

Let's imagine a little experiment. We put the camera on a tripod, and

take two identical pictures, at exactly the same speed and aperture,

on two different films. The light does not change during the entire

procedure. The films have the same speed and are of the same brand

(so that they have similar emulsions), but one of them is a negative

and the other one is a reversal film.<br>

From my point of view, the machine should choose a combination of

filters that would make the print have the same colors as the slide.

And this combination should depend only on the color of the mask and

the color temperature of the machine's bulb. But in reality this

doesn't happen, so I'm obviously missing something. What?

<br><br>

Is it possible to get on paper the same colors as on the slide? If

not, then why?

I'm going to speak to color printing in general, things are slightly different with a mini-lab machine. There are two more variables to take into account, paper and chemistry. Take the same negative and print on two different papers and the color will be slightly different. Same thing if you vary the chemistry slightly (including the variation that occurs due to reaction from processing paper).

Thus with the same enlarger, and filter pack you will get slight variations in color, thus you need to color correct.

The mini-lab should have its chemistry monitored regularly and should have that much variation from chemistry. I would imagine that mini-lab operators don't switch paper frequently either.

Thus for a given film, properly exposed, there should be little variation in the filter pack if you want print that would match a slide exactly. Unfortunately, most people don't correctly expose their negatives, and take photographs with color casts. The mini-lab operator need to compensate for this, otherwise too many customers would come back complaining that uncle bob's white shirt is yellow.

> Why does the filtering depend on the image itself?

When humans look at a scene there is some processing going on inside their heads to automatically adjust for color temperature, contrast range, overall brightness, image content, etc. The brain registers a "percieved" image, unlike he film which records only a raw image.

The printing machine is programmed to simulate human visual perception on a primitive level and its goal is to produce as "natural" looking prints as possible. Therefore the machine is allowed to make some semi-intelligent decisions about color filtering and exposure that should work best for a given picture.

OK, I agree that most people don't expose their films correctly and they want their prints corrected, but if I tell the lab <b>not</b> to correct my prints, I never get satisfactory results, although I'm absolutely sure that my exposure is right (I work mainly with slides and I always get good results).

<br><br>

A friend of mine used to work in a street lab and I used to have him process my films and pictures whenever I worked with negatives. He always asked me what part of the picture I'm interested in. One day I asked him: "What do you mean? I want <b>everything</b> to come out perfectly. I don't want to choose. I want you to get on paper exactly the same image that I would have got if I had used slide film." He answered:<br>

1. "I don't know what you would have got on slide. From this negative I can get thousands of different looking pictures, and I can't possibly know which one of them is the "correct" one."<br>

2. "I can't get all colors perfect. If there is a person in the picture, then I'll do my best to get good skin tones, forgetting about the other colors. If there is no skin tone, then I must decide which color is most important; if for example you have a lot of green in the frame, I use more magenta so that you get a good looking green. I can't get simultaneously a good looking green and a good looking red, for example."

<br><br>

I can understand the first part, although I still believe that if you set the right filters for your particular mask+bulb+paper+chemistry combination you should be able to get the image that is the exact result of your exposure.<br>

However, I don't understand the second part, and my friend couldn't explain it to me. He <b>knew</b> from experience that it is so, but he didn't know how to explain it.

<br><br>

In order to test this, I recently went to a lab that has a Frontier and I asked for a print from a slide. It was a picture of my wife, dressed in a red sweater, sitting on a yellow couch. The colors of the slide were all correct, the image looked as realistic as it could get. However, the print showed perfect skin tones and a perfect red (they were exactly as on the slide), but the couch was yellow-green. In order to get on paper the right red, they had to sacrifice the yellow. <br>

Why?

--"But if I tell the lab not to correct my prints, I never get satisfactory results, although I'm absolutely sure that my exposure is right"--

Vlad, I'm no expect. I also experienced asking the lab to not correct my prints and then getting unsatisfactory results.

Background. I learned using slides (mainly by reading John Shaw, nature photography.) I could swear that I eventually got so good that I could get the correct if not the desired exposure 8 or 9 out of every 10 slides. Then I switched to prints to see if my slide techniques worked with print film. I also asked the lab to not correct my print film.

Vlad, my first 8 rolls of print film had few if any frames correctly exposed. This was bewildering to me. Eventually, I realized two things. I had gotten lazy and stopped being extremely careful about minor details of exposing film. I was rushing and omitting steps. The convience of 1 hour processing made me shoot quicker so that I could more quickly see my handiwork. Secondly, I learned the lab was still adjusting my pictures. Eventually one lab tech told me how to read the printout on back of each picture. He said that no adjustment (called "autoexposure") would print on the back of each photo four Ns. I would see: "N N N N" When adjustment took place the printout would read something like "N-30 N+2 N N+4"

Now I verify when no adjustments take place and I take my time to expose each print. I've been getting a higher number of correct exposures. Mystery solved (at least for me.)