Monday, February 04, 2013
Scanning old film files is an interesting exercise. I'm finding photos that I never printed and simply forgot about. Unless you make at least contact prints, film images are easy to forget about and never see again. They are shut away in file pages inside boxes. Looking at them requires pawing through and holding each page up to the light while mentally converting the negative image to a positive one. To complicate things even further, unless they are badly out of focus they all look sharp in small negatives. It isn't until you use a magnifying glass or put the negative into an enlarger that you can see for sure how sharp it is and whether the DOF is all you intended.
Aside from that though there is a subtle quality to to a film image that digital doesn't quite match in my opinion. When digital cameras first came out I was skeptical because film images are made up of random grains of silver whereas digital images are made up of square pixels. The reason that makes a difference is that we instinctively recognize things by shape and pattern. The random grains of silver have no inherent pattern so we immediately recognize the image they form. With early digital and its low resolution we had a grid of squares which competed with the image for recognition, a lot like a tiny mosaic. Once resolution passed the 11-12 megapixel range though the pixels were so small that (theoretically at least) we can't see individual pixels so it shouldn't make any difference whether it is made up of random grains or a grid of squares. Yet somehow there is a difference.
Back when CDs were a new technology an audio technician in a recording studio told me that they did all their recording and mastering on analog tape. I asked why they didn't record digitally and he explained that digital recording lost subtle resonances. In any musical performance there are sounds above and below our threshold of hearing which add to the overall quality of the sound. In order to create a digital file however, they needed a constraint on the size of the file, much the same as the way that pixels are all described by the same number of bits, 8 bits/channel in an 8 bit image, 16 per channel in a 16 bit image. To do that in a reasonable sized file the audio engineers designing the recording system had to chop off the frequencies above and below human hearing and in the process lost the subtle resonances. Thus (according to him at least) the best recordings were AAD, analog recorded, analog mastered and digitally reproduced.
The corollary in photography is that the tonal range is 256 shades of brightness from pure black to pure white. The number of tones was determined by what a monitor could display. Besides that, the steps between tones had to be even because they are mathematically derived. To make it work you can't have pixels in one area described by 1 or 2 bits and others described by 80-100 or more. They are all described by the same number of bits with even jumps from one to the next. Consequently continuous tone with digital is an illusion created by making the steps so small that the eye can't see the jump from one to the next. But light doesn't work like that. It really is continuously variable without steps.
Arguably film isn't continuous either. The silver particles may be random but they are opaque and cast a shadow on the printing paper resulting in white areas on the print. The resulting tones are determined by the density of the grains in the same way that a newspaper half-tone screen does, albeit much smaller. There is a difference though because the halftone screen, like a digital image, is a grid, not a random pattern like silver grains on film. Also the printing ink is always black while a silver emulsion print can create shades of grey.
Scanning negatives has gotten me thinking about all this because I have recently read a couple of articles questioning whether one could tell from a print whether a photo was originally made with film or a digital camera and that question was apparently lurking in the back of my mind. Of course I know with my own photos which is which and my gut tells me that there is an undefinable difference something like those tones in music that are above and below the range of hearing but never the less add to the overall performance. Can I tell when looking at someone else's prints? I'm not ready to place any bets.
Today's new/old photo shot on 120 B&W negative film with a 645 camera. I made it on a daily exercise walk past a beaver pond that was about a mile and a half from our home. Like the Snowy Trees image from a few days ago this is the first time anyone other than me has seen it and before I scanned it, I had only seen the negative version. The roll it was on was not contact printed. In the audio technician's terms this would be an FDD image, Film recorded, Digital scan mastered and Digitally presented.