I would disagree with this. If one wants to be able to do work in the darkroom, then one is stuck with that. However, I believe the best scanning neg is twice as dense as the darkroom neg. Top end is at roughly 2.0 vs 1.0 in transmission density. I had been making negs like this for printing in platinum and I found that the same negs scanned wonderfully (I have a drum scanner).
I have great respect for both Don and Bruce's opinions but my experience varies on this one.
Lenny
EigerStudios
I'm not saying Mr. Eiger is wrong. Different films, developers, workflows, scanners and software give different results. I've run experiments with 5x4 Tri-X where I ran my Dmax from around 1.0 up to around 3.0. Since I'm also drum scanning, it wasn't difficult to scan any of these. But the results varied quite a bit.
The problem is two fold I think. First, graininess is directly related to density. The more density you build, the more graininess you build, because it's the creating and overlapping of the silver that gives you the grain clumps that form the density. The physics is the physics; it's just how it works. If you are using really fine grained film, or you aren't going for big enlargements, perhaps this isn't such a big deal.
The second thing is Callier Effect. Callier Effect is also directly related to density, for the same reason. The more silver, the more light scatter. What this means to scanning is a decrease in local contrast in the most dense areas -- that is, highlight compression. One can compensate for this with a curve in Photoshop. To some degree anyway. And this, I think, is a big deal. It's a big deal to anyone scanning or enlarging in the darkroom. This is why the darkroom guys preach "only as much density as you need to print easily and no more." It applies to scanning too.
There is a third problem and that is that some films will start to show artifacts if "abused" the way I was abusing Tri-X for awhile. I was seeing "density halos" which looked sort of like ripples in a pond. They formed dense areas of the film that were adjacent to thin areas of the film. Perhaps a developer exhaustion problem, perhaps a development products problem. Since I was using continuous agitation (Jobo CPP-2, 3010 tank) I was surprised either way. And I'm sure that some films will do this and others won't. I'm just sayin'...
Once I got a handle on what the heck was actually going on with the film, particularly Callier Effect, I started testing film going the other way -- thinner than normal. What I found with Tri-X, and have repeated with TMY-2, is that less density is good for scanning. Especially with TMY-2, graininess drops to just about invisible at an 11x enlargement level (normal for my personal work, sufficient for a nice big 125 x 100 cm print). More importantly, I no longer have to do any decompression of my highlights in Photoshop, because at lower densities Callier Effect is much less noticeable.
Clearly, you have to optimize for your film (which will undoubtedly vary widely in how it responds), your developer, your workflow, your scanner and software, etc. This will require anyone who does it to personally do the testing. Just like doing Zone System testing to find your personal EI and "N +/-" development times. Gotta do the testing to know what works for you.
And just as clearly, if you are *ever* going to try to print in the darkroom, *only* optimize for the darkroom. It will scan just fine. You'll just make yourself miserable if you optimize for scanning then try to print in the darkroom.