Re: Dynamic Range testing
Re: Dynamic Range testing
- Subject: Re: Dynamic Range testing
- From: "Terence L. Wyse" <email@hidden>
- Date: Tue, 4 May 2004 16:03:13 -0400
On May 4, 2004, at 3:18 PM, Douglas Rhiner wrote:
I wish to test the Dynamic Range of my two scanners, a TANGO and an
LS-8000.
What do I need to accomplish this? I.e. what target would I scan?
How should the scanner be set up for the test?
Any procedural aspects to look for so I get a "true" reading?
Also, what would be the best way to test how large of a color
space/gamut the scanner can capture?
Possibly create profiles and compare the results with ProfileEditor or
similar application?
Any info on how to do this in a scientific method compatible way would
be of great use!
Well, I don't know how about "scientific method" but here's how I'd
test it. I'd get myself a good dye/carbon-based transmission step wedge
(Kodak T-34 is a good one) and possibly even have on hand a 1.0 or 2.0
ND filter material.
Read the step wedge using a calibrated transmission densitometer and
record the values for each step.
Mount the step wedge and cover HALF of all the steps with the ND filter
(the 1.0 ND should extend the density well past 4.0 which is probably
enough). Mask it off real well to avoid light scattering (credit to Don
Hutcheson for mounting a ND filter over the step wedge).
Scan this combination and then plot a graph comparing RGB values vs.
original density. When scanning, you'll need to scan "linear" and make
sure you've got the exposure up high enough that the first few
highlight steps are blown-out or reading 255/255/255. To extend past
about 3.0 density, simply add 1.0 (or 2.0) to the steps you covered
with the ND filter material. Once plotted, you should pick up a
definite toe and shoulder to the curve. Draw a line on the straightline
portion of the curve and you should be able to tell where the dynamic
range starts falling off. Subtract the minimum density of the toe from
the Dmax of the of the curve's shoulder and you should have the density
or dynamic range of the scanner. It's hard to pick an "absolute"
dynamic range number as you have to decide how far into the shoulder of
the curve you're going to say that it's still recording/seeing detail.
The numbers might still be changing but you could be well into the
"noise" level of the scanner at that point.
Maybe you could pick Jim Rich's brain on this as I believe he's done
exactly what you're wanting to do. I've done my method for drum
scanners but never for a flatbed. Come to think of it, the T-34 step
wedge probably cannot even be mounted/scanned on the LS-8000. And
you'll probably wince at the cost of the T-34 step wedge. :o
As far as the color space/gamut, the only thing i can think of is to
scan a target with sufficient color/dynamic range (Don Hutcheson's Fuji
Velvia HCT targets come to mind) and profile it.
hope this helps,
Terry
_____________________________
WyseConsul
Color Management Consulting
email@hidden
704.843.0858
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.