I have an Nikon LS20 scanner.

I'm in the process of trying to work out some development times and ISO ratings for films and would like to try and use it as a densitometer.

I have a series of negatives which I have measured using an actual densitometer (taking into account the density of the film base) - so I have a means of translating the scan values into a densometric value.

I tried it last night, essentially scanning the film base followed by each of the negatives then used the following formula:

X*(DN-DB)=D

where:

X = the constant which I was hoping to use to translate scan values into a densometric value

DN= the density of the negative according to the scanner (as a number between 0 and 255)

DB=the density of the film base according to the scanner (as a number between 0 and 255)

D=the densitometer reading for the negative in question.

Of course, it didn't work. The value of X - which should be a constant - ranged between around .001 and .003 - that is, it varied by a factor of up to 3.

It occurs to me that the scanner might be auto-exposing which would account for the variation. Does anyone know? Does anyone know how to switch it off?

Any idea what else I might be doing wrong?

Is there any other software out there (Silverfast perhaps?) which allows densometric readings of negatives.

Thanks in advance

Gareth Jolly

-- Gareth Jolly (sgfj@minters.com.au), March 25, 1999

Answers

I have done some experiments with flat-bed scanners. On the (very cheap) scanners I tested, with a gamma setting of 1.0, the pixel value (0..255) was proportional to the exposure from the paper, i.e. proportional to the percentage of light reflected from the print. The density, as you may know, is a log function, which would explain why your linear equitation wouldn't work.

-- Alan Gibson (Alan.Gibson@technologist.com), April 06, 1999.