Digital BW, The Print

Ernst Dinkla wrote:
> Roger wrote:
>   
>>> Or for example multisampling/long exposure with a Nikon 8000 
>>> on wet mounted color negative film like I do. Similar 
>>> artefacts seen in digital HDR can show though. In this case 
>>> higher noise in the highlights for color negative.
>>>
>>> Exposure bracketing on a DSLR isn't bad either if the 
>>> subject allows it, goes faster between shots than on an 
>>> analogue MF camera.
>>>       
>> I'm surprised you see a reduction in scanner noise in *color* negative highlights using multiexposure scans.  I usually find that if I set the overall scanner exposure correctly the entire dynamic range of the color negative is well within the scanner's range and nowhere near the scanner's Dmax.  Some B&W negs and slides do benefit from this technique.  I use Photomatix's combine exposures feature to blend bracketed scans (and DSLR shots).  I like it better than HDR and tonemapping.
>>     
>
> It isn't the dynamic range or Dmax not captured by the 
> scanner but its problem with reproducing the tone values 
> accurately from the compressed range of the color negative 
> that can be solved with multisampling and spreading the 
> result on 16 bit. The noise in the highlights is what I 
> still see though so it just works not as good in dense parts 
> of the color negative. The long exposure is what I use for 
> B&W negative and there I see the same (not similar) 
> artefacts in the shadows observed in digital HDR. Nothing 
> special just Vuescan's tools. Few times I scan slides. I 
> agree that it could have been written more clearly.
>   
It's not a failure of color negative materials. People routinely blame 
poor scanner performance on color negative tonal compression. Yet, I 
routinely scan color negatives on my drum scanner with excellent 
results. I never experience of these so-called compression problems. And 
if any scanner / software / operator can get excellent results from 
color negative materials, the fault is not with the film, but elsewhere.

CCD scanners, OTOH, may well be subject to this "compression" failing. I 
have too little experience to venture forth an opinion on CCD scanner / 
software performance.

All that said, the nature of film is that graininess is not a constant. 
It varies directly with density. Low density regions of the film show 
very low graininess, while high density regions can show orders of 
magnitude higher graininess. With a tranny, greatest film density is in 
the print's shadows where the graininess is hardest to see. With 
negatives, greatest film density is in the print's highlights were it's 
much easier to see. This makes negative film "seem" grainier (noisier) 
when in fact modern negative materials nearly always have lower 
graininess indices then their same ISO tranny counterparts.

Since this is a B&W group, I should add that the same thing applies to 
B&W negative films; that graininess is directly related to density. But 
there's an added twist. Callier Effect is also directly related to 
density. The effect is that higher density regions of the film cause 
more light scatter (Callier Effect) which in turn lowers the local 
contrast for that region of the film. The result in the print is 
highlight compression, and again many people blame this on film 
shouldering. Both scanners (CCD and PMT) and darkroom enlargers 
experience this, and it's one reason why optimizing B&W film processing 
for scanning means decreasing development to decrease Dmax which in turn 
decreases the amount of metallic silver in the dense regions of the 
film, which in turn reduces both graininess and Callier Effect.
--
Bruce Watson