Re: digital negative possibilities for gum
On Oct 20, 2006, at 5:44 AM, stwang@bellsouth.net wrote: I think this last point is especially well taken, and should provide a warning against reading too much into comparisons where the printing method is held constant across different negative types or conditions. We think we're making our results more interpretable or generalizable by holding everything constant, but in many cases we're doing the opposite. For example, I've been puzzled in the past by demonstrations that compare the DMax or contrast for different concentrations of dichromate, keeping the exposure time and development time constant. Of course, if you're going to go that, you can show anything you want by choosing an exposure time that favors the conclusion you want to draw. But the only reasonable and meaningful way to compare the contrast and DMax of different concentrations of dichromate, is to print each concentration at its natural exposure time, which will be different for each dichromate concentration (since speed varies directly with concentration) and compare the contrast and DMax of the resulting prints that are properly printed for the dichromate concentration. By the same token, if you compare different negative types using the protocol for one negative type as a standard protocol applied to all the different negatives, then the negative type whose protocol you used will of course come out looking best, because the protocol is calibrated for that material. But it would be misleading to say that means that material is better just because it looks better in a side by side comparison of this kind. If you'd calibrated the protocol on one of the other types of material, then that's the kind that would come out looking best. The only way to really compare different negative types in a meaningful way is to print each using the protocol which optimizes its performance, which as Sam says will be different for each of them, and then compare those prints. These *seems* less standardized, but it's actually more standardized. You're optimizing the print, rather than standardizing the protocol, and optimizing the print after all is what we're after, isn't it, or have we got so obsessed with standardization that we've forgotten what the goal is? By standardizing the protocol, you're actually confounding the results, not clarifying them. Perhaps an example would help clarify my meaning. It's like the time I printed a small image on a lot of different paper samples, to use in a demonstration to show how gum looks on different paper surfaces and textures and so forth. I certainly wasn't going to calibrate the printing time for 20 different papers (and yes, different papers do require different methods too) so I just used the printing time I always used for my standard paper, which at the time was Arches Aquarelle. But they drew a completely different conclusion from the demonstration than I intended them to. I just wanted them to see that different papers give a different look, in terms of detail and texture, to a gum print. But instead, they concluded that because the print on Arches was perfect in tonal scale and contrast and all that, that Arches must be the best paper to print gum on. I'm not sure I ever managed to convince them, though I tried very hard, that the print on Arches looked best because my whole printing protocol was calibrated for Arches, and that if I had calibrated for a different paper, any of the other 19, then the print on that paper would look best. Sorry, I'll get off my soapbox now, but this is something that I've been thinking about for quite a while around different issues. Katharine
|