Russel Kwan
While I remain committed to working entirely with traditional materials, I’ve always had an interest in all things digital. In an earlier career, I helped bring about the imaging digital revolution, by developing software for Adobe, Microsoft and Apple Computer – yup, I’m also a professional software engineer. Now, I’m doing the exact opposite – trying to preserve the glories of film, chemistry, and silver gelatin prints.
For the longest time, my decision to work in film was very simple: film-and-projection-printing methods produced prints containing far more detail than anything digital. Recently, that advantage has begun to change, but since I work mostly in medium format and larger, I’ve not really thought about it – until today. Sensors in DSLRs are getting into the 20+ MP range, and so I’ve become curious as to just how much detail my materials actually record.
For this experiment, I selected a photograph made with my medium format Mamiya 7II 6x7cm camera. There is nothing special about this picture, other that it features some bottles with tiny printing on their labels – a reasonably objective way to measure clarity. I used Ilford HP5 film (ISO 400, rated at 160, not a particularly fine-grained film), and the camera was on a tripod for the 2 second exposure. A regular 8x10 RC print was made from the negative, and scanned with my Microtek 9800XL at 1200ppi. The unaltered (except for downsizing and jpeg conversion) scan is shown below.
Below is a full-resolution detail from the scan, showing the 3 bottles in the center of the frame above. This image is unaltered except for jpeg conversion. The white speckles are the texture of the semi-matt paper I used. The film grain is visible (somewhat fuzzily) in the bright label backgrounds.
If I use a loupe on the original print, the box at the bottom of the label of the middle bottle reasonably clearly reads:
Clearly, the original print has detail that my scanner at maximum optical resolution cannot resolve.
So what’s 14MP look like anyways? Well, I don’t have access to such a camera, but we can simulate it by downsizing my original 83MP scan to 14MP. In truth, the quality of a downsized image should be a little better than an actual image shot at 14MP.
Wow. Pretty good! Not as good as traditional medium format, but pretty close. I must admit that I’m surprised. Will I go digital? Nope. But, I’ll keep following developments.
Finally, here’s some interesting facts: to make a digital 8x10 that has the clarity of my 1200 ppi scan, we will need to capture an 83MP (yup, 83 – not 8.3) source image, which works out to a 250MB file. We’ll also need an output device that can print 1200 discrete, non-overlapped dpi. And we’ll need to be able to print any of the 256 gray values onto any dot. And we still won’t be able to read what’s in that box at the bottom of the label of the middle bottle.
Google Analytics