see luminous landscape: http://www.luminous-landscape.com/tutorials/dq.shtml
Great link. I browsed until I got here:
http://www.normankoren.com/Tutorials/MTF7.html
and I especially looked at the Russian newspaper example.
What would be difficult to put into numbers is the effect of "grain" on image quality. Yes, the image at the right looks "grainier", but if you ask me to pick the "better" one, I'll pick it for sure. In the example at left, it doesn't matter that the image is "smoother" as long as I can't read the vertical line. On the right one I can definitely read "Gheneralnaia litzenzia" (sorry for the poor transcription), but try that on the left one.
There's much to debate there in the choice of lens and comparison method. First, the lens. I had once a "bow wow" 28-80 zoom that was a great equalizer for film quality. No matter what, all images looked the same. I.e. soft ;-) I wonder why they didn't use one of Canon's excellent macro lenses instead. Also, looking at the photodo diagrams, it seems that the sample they tested was defective ;-)
The next objection would be this: why not do the test the other way round. I.e. feed the digital image through a slide printer and then compare the films with a loupe. Why do they always have to scan the film image and compare the result with the raw digital one ? How about printing the digital one with an Epson inkjet and then scanning *that* ? I submit that at least for B&W photography, ancient darkroom techniques, when properly done, will beat any current digital ones.
Then there's the choice of film. What about some Fuji Acros, for a change ?
And so many other questions....
cheers, caveman
