Best film/develper combo for scans?

[cmox]

Unfortunately I do not have enough space for an enlarger and trays... just enough for my PC and an Imacon 646 scanner (cost me more than all my cameras and lenses, in fact).

I shoot 35mm and 6x7cm, edit the scans in Photoshop and print VERY large B/W prints on an Epson printer.

My favourite films are Tmax 100 and Tri-X, they are souped in XTol 1:1. These are good films and a good developer, but there is a disadvantage.

My Imacon works like a condensor enlarger and emphasizes the grain. As the grain becomes very visible I can not sharpen my landscape images as much as I would like to.

I heard some reports that it might be a better idea to use a developer that creates sharper negatives from the beginning so that there is not so much sharpening necessary... or staining developers... other friends tried to convince me using C-41 films like XP2... but I prefer silver films.

According to your own experience, what is a better combination for very high quality scans for extremely large prints?

 

[Maxapple88]

I guess Rodinal would give very sharp images. maybe too strong though. RHS gives surprisingly sharp images I think.

 

[jlw]

My Imacon works like a condensor enlarger and emphasizes the grain. As the grain becomes very visible I can not sharpen my landscape images as much as I would like to.

I heard some reports that it might be a better idea to use a developer that creates sharper negatives from the beginning so that there is not so much sharpening necessary... or staining developers... other friends tried to convince me using C-41 films like XP2... but I prefer silver films.

According to your own experience, what is a better combination for very high quality scans for extremely large prints?

Welcome to the unpleasant world of grain aliasing (click here for an older link that remains the most scientific article I have found on the subject.)

The short summary is that grain aliasing isn't a quirk of your particular scanner, film, or process -- it's an inevitable phenomenon that involves a "sampling" interaction between the random structure of the grains and the non-random structure of your scanner's imager. The result of this interaction is that the random grain pattern appears magnified.

Chromogenic films produce less of the effect because their grain structure consists of soft-edged dye clouds rather than the sharp-edged silver halide granules of conventional b&w films; the soft-edged dye clouds don't interact as strongly.

If you don't want to restrict yourself to chromogenic films, various tricks have been suggested (and undoubtedly will be suggested in this thread) involving defocusing the scanner, or applying this or that Photoshop plug-in to the scanned image.

However, because all these approaches involve throwing away part of the scan data (rather than preventing the aliasing before it occurs) there's no way they can work without removing image information from your scans.

You may decide one or another of these approaches produces a compromise you can live with, in which case there's your answer... but remember that it's still a compromise. (For example, you might find that the trick or filter or plug-in that produces acceptable results for your landscapes won't work for portraits, and you need to find a different tactic for those.)

One trick I've used myself with some success is to make two copies of your scan layer in Photoshop, then offset one layer slightly with respect to the other. You have to choose the direction to suit your taste, and the amount to correspond to slightly less than the average "diameter" of the grain structure. The effect of this is to blur the grain edges slightly in a way that makes the aliasing less apparent. Of course, then you'll have to go back and make level adjustments to restore the correct overall density of the image.​

But basically there's no way you can completely avoid the problem, because it's based on the fundamental principles of how scanning works. To be honest, the inability to get fully satisfactory scans from b&w negatives coupled with my need to distribute digital images was what eventually drove me to stop using film (mostly) and switch to a digital camera!

 

[Rhoyle]

Although I'm not using a drum scanner, I've had some limited success by setting the scanner software to color positive instead of bw negative. Then I flip it in Photoshop, either keeping it as an RGB or converting it to Greyscale. It seems that I get less obvious grain that way. I was talking to a friend who works in a pro lab and he laments the same issue. Sometimes, he'll print a neg as a wet print, then do a high quality scan of the print to get a file. He showed me some examples of 4x5 tri-x negs printed at about 4x5 feet. There was a definite difference in grain between the two and apparent sharpness and contrast were fairly equal. I"m not saying that this is the definitive solution, just pointing out that you are not the only one with the problem. However, see if scanning as a color positive helps.