Predicting color balance and print exposure times.

[stevewillard]

I have an idea for precisely predicting color balance and print exposure times for color negative papers. I am looking for your considerations.

First, let me provide you with the circumstances. I am a fine art landscape photographer. To maintain quality control and consistency from print to print, I make a master copy of each image I have. I use the master copy as a target when printing subsequent prints. I have found that making precise copies of the master print to be very difficult if not impossible without wasting lots of paper and time. I can do a better job by making test prints than I can with a color analyzer, however both still fall short of my goal.

My overall strategy is to apply the analysis at the print level rather than at the negative level as color analyzers do. This should eliminate all variables attributed to paper batch and processing. I will be using a print densitometer for taking readings which far more accurate than a color analyzer. So here is my grand plan.

1. I will “rough in” a test print that is close to the master print using the manual test strip method. I can get close in two sheets of paper.

2. I will then select a critical area on the Master print and take a RGB reflective density reading and a white light reflective density reading. I will then repeat this procedure on the test print in the same spot.

3. The differences in the white light density readings will tell me the time change that I need to apply to the test print to achieve the same density as the master print. I have already done this using a linear interpolation and found that this method to be very accurate provide the prints are reasonably close in density.

4. The difference in RGB readings will tell me the color head changes I need to make and the times changes I need to make as a result of changing the filtration. This time will be added to the white light time noted above.

Clearly, I will have to quantify the impact of unit changes in print density for both white light and RGB readings would have on the color head and final print times. Once I have done this, then I can develop an equation which inputs all the densities and computes time changes and color head changes needed to generate a print that matches the master print with a greater degree of accuracy and much less paper and time.

So does this seem possible? Has any body done something similar to this?

 

[jd callow]

Wow it sounds like a lot of work so that you can knock a big chunk of the human out of your prints. (not to be a heretic, but this is why they make lambdas, frontiers and other digital devices) I think your eyes and brain can do as good a job. I would think that you might lock yourself in and miss the opportunity to improve a given image. I have some popular prints that I have had to print time and again over the years. The early prints are good, but the later ones are almost always better. Had I used a system you describe my prints would be consistent, but not as good as they could be.

 

[Nick Zentena]

Which analyzer are you using and how often do you recalibrate it? With mine if I wanted to really match things I'd recalibrate after a certain number of prints. Say 10? Or at the very least at the beginning of each new session. That way any aging in the chemicals is adjusted for. With mine recalibrating is very simple. Just stick a grey negative in the enlarger and make a print. The analyzer then reads the dry print [actually a test strip] and adjusts itself. That's with my Colorstar 3000 but I think all the newer models do the same.

 

[Fotohuis]

Exactly, we are working with the FEM-Kunze CFL-4012, same principle. Best is to make a photo with the grey card on your film and indeed the analyser reads the dry print (test strip) and will adjust itself for that used combination of film, chemicals and paper batch, which you can store under a channel with a brief description. In 2-3 cycles you have a 100% neutral print.
To be more specific: I am talking about the AD2000 (densito) and CFL-4012 (C-analyzer) which will cost around Eur. 800,00 (new price).

http://www.fem-kunze.com

 

[Bob Carnie]

Have to agree with Mr Callow on this one , are you printing images or grey cards? Any image should be within 1-3 tests and you may discover a whole different and interesting image, by being off colour.

 

[pentaxuser]

I have an idea for precisely predicting color balance and print exposure times for color negative papers. I am looking for your considerations.

First, let me provide you with the circumstances. I am a fine art landscape photographer. To maintain quality control and consistency from print to print, I make a master copy of each image I have. I use the master copy as a target when printing subsequent prints. I have found that making precise copies of the master print to be very difficult if not impossible without wasting lots of paper and time. I can do a better job by making test prints than I can with a color analyzer, however both still fall short of my goal.

My overall strategy is to apply the analysis at the print level rather than at the negative level as color analyzers do. This should eliminate all variables attributed to paper batch and processing. I will be using a print densitometer for taking readings which far more accurate than a color analyzer. So here is my grand plan.

1. I will “rough in” a test print that is close to the master print using the manual test strip method. I can get close in two sheets of paper.

2. I will then select a critical area on the Master print and take a RGB reflective density reading and a white light reflective density reading. I will then repeat this procedure on the test print in the same spot.

3. The differences in the white light density readings will tell me the time change that I need to apply to the test print to achieve the same density as the master print. I have already done this using a linear interpolation and found that this method to be very accurate provide the prints are reasonably close in density.

4. The difference in RGB readings will tell me the color head changes I need to make and the times changes I need to make as a result of changing the filtration. This time will be added to the white light time noted above.

Clearly, I will have to quantify the impact of unit changes in print density for both white light and RGB readings would have on the color head and final print times. Once I have done this, then I can develop an equation which inputs all the densities and computes time changes and color head changes needed to generate a print that matches the master print with a greater degree of accuracy and much less paper and time.

So does this seem possible? Has any body done something similar to this?

I'd have to agree with the other respondents on this one and maybe go one extra stage. Admittedly I am fairly new to colour printing but recently I repeated a print I had done 2 months before on the same batch of paper. My colour analyser reproduced exactly the same colour balance and against the master print I had kept it was identical.

The analyser was a secondhand Paterson PCA2060 so no rolls-royce of an analyser.

The real problem arises when you change to a new batch of paper even of the same make but then the analyser only needs a small amount of re-calibration. Maybe one or two test prints against the master print to resolve the balance.

Tiny differences in balance may be detectable by instruments but probably not by human eye and normally the customer wants another print for someone else or has given away his original and wants a replacement. It would be unusual for any critical comparison to be made.

If a customer/friend were to be that critical then it is likely that he/she is also in the colour business and if he/she were, then being critical he/she would want to produce their own prints.

Pentaxuser

 

[Ed Sukach]

Exactly, we are working with the FEM-Kunze CFL-4012,...
http://www.fem-kunze.com

Visited that site. It seems *very* interesting, but as am not fluent in German and the "English" version is not availabe, I'm left with many questions.

Is there some sort of "automatic" method of comparing the original test card to one of the images of it printed out?
This would be, IMHO, one step ahead of the Colorstar 3000 (which I also use), but all the information here is not easily understood.