Using a scanner as densitometergreenspun.com : LUSENET : Large format photography : One Thread
I have in my office an AGFA ARCUS II scanner. I shoot on 5x7 plates. How can I use this scanner as densitometer?
-- Martin Kapostas (email@example.com), February 15, 2001
I am making the following assumptions based on your question.
1. You have a transparency adapter and are scanning 5 x 7 negatives.
2. You have a computer. Might you also have photoshop?
If you have Photoshop, you can scan a calibrated negative, or a negative with known density values into Photoshop. Using the eyedropper tool, you can make a correlation between the percentage value of the eyedropper and the known negative density value. Then you can make your own density chart.
If you don't have photoshop, a used densitometer would probably be less expensive.
-- Joe Lipka (JoeLipka@Compuserve.com), February 15, 2001.
Hey, that's a clever idea. You'd have to be sure to turn off any autoexposure algorithms that the scanner software might have. And everything Joe said. You can scan a step tablet.
It might not get you densities accurate to 0.0001, but would probably give you a lot of useful information nonetheless.
FYI, in case no one is familiar with it, Beyond the Zone System gives directions for making a densitometer from your spot meter. Kind of the same idea. The scanner might be a little more consistent and wouldn't require hacking together a stand and a close focus attachment.
-- John H. Henderson (firstname.lastname@example.org), February 15, 2001.
This'll only work if you can be sure that the scanner doesn't use any auto white and black point algorithm. If you have the option in the software to get a high bit 'raw' scan out of the scanner, use that.
A lot of scanners use 12 bits internally, and map them to an 8 bit output, after calibarating themselves to the brightest and densest parts of the image. This system is no good for even relative density values, since the RGB levels for a given density will change, dependent on the density range of the scan.
I'd suggest getting some ND filter gels of known density, and scanning them in various combinations to see if the pixel values stay consistent. If not..... nice idea, but give up.
-- Pete Andrews (email@example.com), February 15, 2001.
I am curious about the suggestion you have proposed. I have heard about "Beyond the Zone System's" proposal for using a spotmeter as a densitometer. Do you thing that the light source used would have to have a constant voltage and/or current supply? Does anybody have any additional information in this regard? Has anybody here actually converted their spotmeter? If so, how did you like the results you obtained and would you recommend this to others?
-- GreyWolf (firstname.lastname@example.org), February 15, 2001.
This is some really intereting stuff, unfortunately it was way over my head. Could any of you put this in more layman's terms and give a little more info about how you would do this. I have a Epson Perfecta 1200U scanner. It will scan a 4x5 neg or pos. I shoot 4x5. I also have Photoshop. What I don't have is the knowledge to use this in this combination. Any help would be great. Thanks for all you help.
-- Doug Theall (email@example.com), February 16, 2001.
The concept described seems related to how some vendors of color management software use a standard scanner for building printer device profiles (e.g. Monaco EzColor). I guess in this case, the vendor is using the scanner as a poor man's colorimeter. If I remember, the process does require you to scan the test chart with all scanner controls turned off.
-- Larry Huppert (Larry.Huppert@mail.com), February 16, 2001.
The other day, I was doing some idle reading and found a fairly technical discussion about using a scanner as a densitometer at
Scroll down to "The Zone System and Digital Prints". He gives a formula (providing you know the scanner gamma) to convert the RGB value provided by Photoshop directly to a density similar to that provided by a densitometer. He has a comparison curve between a real densitometer and a scanner and the results are really good.
If you can work through the technical details, this might be what you are looking for. Good luck!
-- Phil Erickson (firstname.lastname@example.org), February 17, 2001.