We use NIH-Image instead of Photoshop. The advantage of NIH-Image is that it
has a curve fitting function to find out the best formula to fit the
calibration parameters into the curve for reference. If your curve turns out
to be linear anyway, you might not need anything else. Alternately, to
satisfy the reviewer, you might like to analyze the data with curve fitting
In article <rdlab.62.001184AA at gandalf.bio.uci.edu> rdlab at gandalf.bio.uci.edu
(rlab davis) writes:>From: rdlab at gandalf.bio.uci.edu (rlab davis)
>Subject: Using Photoshop for densitometry?
>Date: Thu, 14 Dec 1995 17:30:58
>We recently began using Adobe Photoshop to quantitate signals on Northern
>blots after we lost access to a densitometer. We determined a range of total
>RNA loaded in which the signals from our mRNAs of interest were linear, and
>the data we obtained using the scanner + Photoshop agreed with our earlier
>densitometry data. A colleague who used similar quantitation methods recently
>had problems with a reviewer who said this quantitation technique was not
>My question is: Can Photoshop validly be used to quantitate signals on
>autoradiographs? If so, what conditions/guidelines must be followed? Also,
>any relevent references would be appreciated.
>Mol. Biol. and Biochem.
>Univ. Calif., Irvine