Digital BW, The Print

Frank,

I am using the new Minolta Dimage Scan Multi Pro.  I primarily scan 
Tri-X 6by7 negative and some 35mm.  Having just upgraded from a 
2400ppi scanner, I was seeing some grain I had not seen before.  

Research lead me to a phenomenon know as grain aliasing.  As you 
probably know, although I have not noticed it mentioned in this 
thread thus far, Grain aliasing isn't real grain but an interference 
pattern between film grain (dye clusters in color film) and the CCD 
pixel size.

Check out the SilverFast grain reduction tutorial at 
http://www.computer-darkroom.com/sf5-negafix/grain_reduction.htm and 
the rather technical but illuminating explanation at 
http://www.photoscientia.co.uk/Grain.htm.

I finally took to undertaking a detailed empirical analysis of my 
own, using a 200x microscope and comparing the negative, to the 
pixels in Photoshop, to the Piezography print (again under the 
microscope), not because I am technician rather than an artist, but 
because I wanted to be aware if the technologies within my workflow 
were interjecting something into my final product.

The answer, and this is my own subjective analysis using some pretty 
in depth research and hard facts, that I am actually awed by how well 
this scanner interprets the grain to render a digital facsimile. 

My only adjustment in my process was to switch from D76 to Microdol 
and to lighten up on my agitation a little.  I liken this to when 
(excuse the metaphor if it does not make sense) I traded in my BMW 
for a Ducati... all the nuances of my riding style that did not even 
show up on the older, slightly sloppier motorcycle, appeared as 
explicit instructions to my new motorcycle.  We are in the 4000-
5000ppi threshold where the CCD pixel size is in the same space as 
the grain (dye cloud), and as such, this phenomenon may raise it's 
ugly head and create some pretty strange results.

There are all sorts of solutions proposed:

1) slightly tweak the manual focus so that it does not react in such 
an extreme manner to each grain.  This also will impact the overall 
sharpness of the image, as landscape photographer, this did not work 
for me at all.

2) GEM seems to introduce other artifacts, plus it only works on 
color scans, so my 255mb 16bit grayscale scan will triple in size.  
This takes a lot of processing cycles as well.  Not the solution for 
me.

3) Get past the zone where the sampling and the grain are at 
the 'same frequency'.  This would mean spending more than the $3,500 
I just spent on my scanner or sending it out for drum scanning.  I 
shoot, develop and print in the same day.... so forget this.

4) Use other software plug-ins or post processing tools to manage 
grain.  This seemed to be a cure for the symptom and did not really 
do what I want, which is to faithfully recreate my image throughout 
the entire workflow all the way to the paper.

5) Shift your grain size and scanning PPI so that they are farther 
apart.  I have done this some.  I am scanning at 3200PPI and like I 
mentioned, my development is tuned for finer grain, without forgoing 
acutance (sharpness).  

The last option, and there are many more I am sure, seemed to work 
best for my and what my criteria is.

A simple search in yahoo, looksmart, alta vista et al on 'grain 
aliasing' will return a lot of information.

Hope I am offering useful information and not being redundant nor 
pedantic.

Regards,

James
www.blueskytech.info

--- In DigitalBlackandWhiteThePrint@y..., "david_nancy_bogart" 
<david.g.harris@r...> wrote:
> Frank,
> 
> Since the Multi Pro is the only scanner I have ever worked with, 
> you can take this post for what it is worth. 
> 
> I too have found the same grain problem. In queries to Minolta 
> and a myriad of other experts who should know, I have received 
> no real explanation of why this is, and more importantly, no 
> workable solution. I just attribute it to, as you say `it is
> sharper than the Nikon' so maybe this is why the grain is more 
> apparent. If you are able to find the answer, please let me know.
> 
> As far as the scanner USM-ing on its own as someone 
> suggested, as you know this doesn't happen unless you've set 
> it. That being said, I find I have to be extra careful with these 
> scans in PhotoShop as normal sharpening will exacerbate the 
> problem. As to the vivid colours, I don't put much seasoning into 
> my image correction recipe. I find if I turn out a fairly bland 
scan,
> it opens with realistic colours.
> 
> But what do I know. The point of this reply is to let you know you 
> are not alone.
> 
> David