Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: film vs digital test

on 7/14/04 8:48 PM, Norm Carver at nfcarver@iserv.net wrote:

> I am in the midst of doing a basic comparison between my Hasselblad and the
> new Kodak SLR Pro (14mb, full frame). I don't need a super accurate test,
> just reasonably fair. My work is half color, half b&w with the end product
> in books and large exhibition prints 20 to 40".
> I invite suggestions and/or critiques of my approach as outlined here:
> I gave up the dark room several years ago after too many decades. So I must
> compare scanned film against digital RAW. Also, though it may invite scorn
> from some purists, I am comparing the actual tools I work with most of the
> time, not the ultimate options in lenses. These are:
> Hblad 203fe with 60-120 zoom
> Kodak SLRpro with Tamron 28-300.
> I take the test images from the same position and adjust the Tamron zoom
> factor to match the approximate vertical coverage of Hblad zoom. To avoid
> any focus hocus pocus I am measuring distances. I care mainly about the
> clarity issue --ie details and sharpness and less about color accuracy as
> this is more easily adjusted.
> The digital raw is 16 bit, 4500x3000 @ 300 rez which equals about 10" ht
> image.
> The 220 films, (CN400) and Ektachrome VS 120 are scanned on a Minolta
> MultiPro (a Nikon 8000 is also available). But here is where I need some
> advice. I believe I should scan to end up with the same 300 rez but to what
> file size?
> Here are two I have tried and the thinking behind each:
> 1. Scan the 2.10 x 2.10 area at maximum of 4800 dpi which gives an image
> size of 34" sq and a file size of 604 mb which is simply too unwieldy.
> 2. Scan at the nearest even dpi to approximately double the image size
> since the 220 film is a little over 2x the ht of the Kodak orig of 1" which
> means 3200 dpi and image size 22.7, file size 268.
> I stand ready for any further ideas. At the end, if any one is interested,
> I shall try to post the results on my web site, normancarver.com
> Thanks for any help, Norm Carver

An idea:  Choose a test which would let each system shine to the moximum,
giving you the best product it can within your constraints, in the format of
your choice, which appears to be large prints.  The main point is that the
test should focus on the final print, and not be a mere equipment test done
independently of your needs.

Since you print up to 40", you might want to use that size print as the
standard.  Or maybe it would be more reasonable to pick a smaller size print
that you know you want the 300dpi resolution on, say 16".  Perhaps the
larger prints can hold up at lower res since they might be viewed from a
greater distance.  This 16" square print would require a file size of 69MB
at 24-bit depth.  The scanner resolution must be 2285 ppi.

Then make the best print you are able to make, on the paper you prefer, with
that film scan file.

Then make the best, highest resolution digital Kodak camera image, of the
same subject(s) under similar conditions.  The camera actually produces a 14
megapixel, 42 MB file, about 60% of the scanned file size.  Now if you
demand the square format, then you have to throw away 33% of that file,
leaving you with 3024 pixel in each dimension.  At 16", that produces 189
ppi image resolution at the print, so it is substantially below the 300 ppi
from the 'Blad.  Anyway, this is what it gives you, so print at the same 16"
print size, or whatever was used before.

It seems like the lenses are not comparable zoom ranges for a more general
test, but I guess you are testing your own setup.  Others might be
interested in a similar (equivalent) zoom range for the two cameras.


Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 


Copyright © Lexa Software, 1996-2009.