View Full Version : Gamma for scanning

Ron Marshall
18-Jun-2005, 17:37
I have just finished doing my testing to determine normal development and expansions and contractions in my new Jobo 3010.

Now I must decide what gamma to aim for. I am developing FP4+ for scanning on a flatbed.

My guess is about 0.62, but advice would be appreciated.

ronald moravec
18-Jun-2005, 18:35
I never measured a gamma in 45 years. Use the neg that prints on #2 paper as the standard and work from there.

Or if you don`t print, make the standard the one that looks good on a calibrated monitor. A visual calibration is good enough so long as you can separate the darkest two steps and lightest two on a grey scale.

My copy of Photo Lab Index says .65 should be the standard. That is the slope of the straight line portion only of the characteristic curve. Other standards have been developed to overcome the deficiencies in gamma definition such as Contrast Index and the is one Ilford uses but the name escapes me now.

ronald moravec
18-Jun-2005, 18:44
I guess I didn`t read your question well. My negs developed to print on #2 paper with a condenser enlarger scan and look good without manipulation on a decent monitor. Kodak`s contrast index for this is .42, where as a diffusion enlarger requires .62. Keep in mind this is not gamma.

I can`t find a definition of contrast index except it is similar to gamma and takes into consideration the toe of the curve.

I`ll try a google.

Brian Ellis
19-Jun-2005, 05:28
I guess I'll have to reveal my ignorance here but I don't understand the question. If you've tested for your development times and you've tested for your film speed, why (and how) would you run a separate test for gamma?

FWIW, use of "gamma" as a number to describe a film curve fell out of favor some years ago because it deals only with the straight line portion of the film curve (i.e. it doesn't include the toe). The more current definition of a curve slope is "average gradient" (Ilford more or less) or "contrast index" (Kodak). Since each of these three methods deal with different portions of the curve the numbers will be different for any given film. However, all three numbers are based on development (time, temperature, and agitation), which you've already tested for, which leads me back to my original question.

Leonard Evens
19-Jun-2005, 08:40
It would probably be better to think about the density range from fog plus base to whitest white. Modern scanners can easily handle ranges up to at least 3.0. For scanning it makes a certain amount of sense to make the negative more constrasty with a higher maximum density. The reason is that the values need to spread out into the 0..255 range, and the higher the original dmax the less likely you will encounter gaps in the histogram on later processing in a photoeditor. On the other hand, overdeveloping can produce some adverse effects in the negative, so you don't want to overdo it. Many people find that they do perfectly well for scanning by developing normally for printing on #2 paper. The upshot is that you should experiment and find out what fits your needs best. Because of the great latitude of b/w film and the fact that you are unlikely to tax the dmax of the scanner, it is not hypercritical. You can always make adjustments in your photoeditor, so you are looking at subtle issues rather than whether or not you will get a usuble scan.

Henry Ambrose
19-Jun-2005, 16:40
I suggest the following approach:

Scan your film in 16bit after manually setting the black and white points, then look at the resulting histogram. If its full edge to edge with nothing clipped you are where you want to be. If its very compacted you should spread its range and if its clipped you would need to lessen the range of the film densities. You want to fill the bucket full but not to overflowing.

The idea being that you want to get the full information of the scene onto the film in a way that it can be read by your scanner. Consider your scanner's capabilities to be equivalent to the capabilities of your printing paper/enlarger/chemistry.

I suggest you not overdo development -- until you test your scanner you don't know what it's range is or the maximum density from which it will gather information.