[ImageJ 1.3.1_13] image calculator question

[Marcin Kowalski, Smarttech]

Hi all,

what actually gives "32-bit Result" in Image Calculator window? Does it scale floating point result of division?? It seems to me that this option scales every R/G/B channel separately because I get result as grayscale. I'm interested in exact mathematical meaning of this option in ImageJ.

I use ImageJ 1.3.1_13

Thanx for answer
Marcin

[Wayne Rasband]

> what actually gives "32-bit Result" in Image Calculator window? Does
> it scale floating point result of division??

What it does is convert both source images to 32-bit floating point and
then it does the division. No scaling is done.

>  It seems to me that this option scales every R/G/B channel separately
> because I get result as grayscale. I'm interested in exact
> mathematical meaning of this option in ImageJ.

You get a grayscale result because the RGB source images are converted
to 32-bit grayscale when "32-bit Result" is checked. Convert the source
images to 3-slice stacks using Image>Type>RGB Stack and the image
calculator will do the calculations separately on the 3 channels.

-wayne