I have read a lot of comments comparing traditional (35 mm) film scanning against direct digital pictures from a DSLR, D100 included... and a lot of comments disturb me, mainly comparisons like “I can scan 35 mm at 4000 (or more) ppi”.
I have experience with high performance scanners, from Crosfield, Hell, and Heidelberg (the successors from Hell). I can scan films at 19000 ppi in a Heidelberg 3900 (it is real, no interpolation), 11000 ppi in a more recent Primescan family, or a “mere” 5080 ppi in a Heidelberg Nexscan family. I used all of them.
1) It is possible to scan at very high resolution 35 mm traditional film;
2) If you are using no sharp or blur you will get significant noise from 2000 ppi from a ISO 100 negative film (sometimes lower), and you will get the same noise from 1300 ppi if you are using ISO 400 negative film (the variation is not so huge if you test different brands);
3) Some people get apparent noise free images at higher resolution, it is due to the poor quality of the sensors of the employed scanners that “blur” the images (please examine your negatives in the microscope, if you have doubts);
4) You can get 0%-70% better results if you are using slide films, and/or very slow films;
5) Summary: the current state-of-art DSLR (like D100) can offer comparable resolution quality to negative films and in some cases slide films too.
BUT theoretically the Bayer mosaic and the algorithm used by the DSLR to “demosaic” it (this is the procedure to convert the RAW data to a workable image file, like interpolation) could be a main issue offering a lower resolution than the expected one, but I cannot find any problems (yet).
Of course we have the latitude subject. I have confirmed that my D100 have a latitude higher than a slide film, but slight lower than a negative print film - and this can be a main issue in some dedicated applications, BUT I am convinced resolution is not the question.
O. Cristo, An amateur photo clicker
O.Cristo - An Amateur Photographer
Opinions of men are almost as various as their faces - so many men so many minds. Franklin
#1. "RE: Resolution: a misconception." | In response to Reply # 0guyb Registered since 11th Apr 2002Wed 16-Oct-02 11:41 PM
I would add some things.
To 3. above I would say it isn't all poor sensors; you can get different noise with different drivers.
Also if you scan at too high of a resolution you may add other information you do not want; you may add the scan of dust, minor scratches etc., not just noise.
Finally there may be something that is not immediately apparent. The comparison of pro/pre-press scanners may lead to wrong scan settings on Nikonian equipment. I do not mean the statements made are wrong. They are just not about Nikon scanners for example. A 4000 dpi Nikon scan will give a higher apparent and real resolution then 1100 dpi. Basically I am saying to get the same overall image quality you may have to set the Nikon scanner at 4000 dpi and a Heidelberg at a lower setting.