View Full Version : How many dpi (cpi) is film?

paul stimac
14-Apr-2005, 12:22
Since film is just silver crystals in a suspension, average crystals per inch should be measurable. Has anyone done this? Can it be determined from the RMS number? Obviously different films and film speeds would give different numbers. Iím just curious.

Bruce Watson
14-Apr-2005, 14:14
My understanding is that this is difficult. The reason being that the size of grain clumps or dye clouds depends in part on the film itself, and in part on the exposure, and in part on the processing.

The film itself determines the smallest size grain you will get. The image determines to some extent the size and spacing of the grain - a large dynamic range in the image means a larger range in sizes and spacing of film grain, for example. Developer and processing can act on that to alter grain size and shape, often in conjuction with the image.

In other words, the dpi for a given film is not fixed, but is instead variable.

Oren Grad
14-Apr-2005, 15:44
The relationship between apparent graininess and the actual clumps of metallic silver is complicated and doesn't map directly into a "dpi" equivalent. Here are a couple of interesting references:

www.kodak.com/US/en/motion/students/handbook/sensitometric6.jhtml (http://www.kodak.com/US/en/motion/students/handbook/sensitometric6.jhtml)

www.kodak.com/global/en/professional/support/techPubs/e58/e58.pdf (http://www.kodak.com/global/en/professional/support/techPubs/e58/e58.pdf)

More generally, as with lenses, the performance of film in rendering detail is most accurately characterized by a modulation transfer function (MTF), which shows the extent to which contrast is preserved at different spatial frequencies. If you wanted to compare the performance of film (or photographic paper) with that of inkjet, dye sub or other "digital" media in a quantitative way, you'd really want to see an MTF for the particular ink/printer/paper combination as well. In both cases, no single number (whether it's lp/mm or dpi) can accurately characterize the performance of the medium in rendering detail.

John D Gerndt
14-Apr-2005, 16:37
Isn't this a bummer? People can talk all they like about perceptions, the look of a lens or the sound of an amplifier and there is very little one can do to quantify it. I know ther are some statistics gathered but not in a consistant manner. We are left with, "Try it, you'l like it!"


Neal Shields
14-Apr-2005, 20:26
Logically it takes two rows of pixels or dots to reproduce a line pair. Emperically, it takes at least three.


Also if you don't want to do the math, the FBI has already done it.


Many sites compare digital to low res scans which has no meaning at all unless you intend to submit your film to low res scans.

Roger Scott
15-Apr-2005, 00:02
For a rough estimate of pixels per inch you can multiply the lp/mm figure by 50.8. Something like Velvia for example is on average somewhere between 4000 and 8000 ppi depending upon contrast. Tri-X is on average somewhere between 2500 and 5000 ppi. It's a bit like comparing apples and oranges as you can't place a single number on it. I have bits of film which resolve absolutely nothing - whoops. I guess I could always scan it into photoshop - I've heard you can do anything with photoshop. :-)


15-Apr-2005, 05:11
I read somewhere on this forum that 35mm film has a resolution of about 22 to 40 megapixels depending on film, lens, etc. Is this true?

If it is why are people so excited about silly 6 - 8 megapixel cameras that are no better thanh 110 film??

Brian Ellis
15-Apr-2005, 06:19
Serious photographers can get pretty excited about "silly" 6 or 8mp cameras because the quality of prints up to at least 8x10 in the case of 6mps and 11x14 in the case of 8mps is better than those size enlargements made from 35mm film , assuming a good lens on the digital camera and an operator who has a basic knowledge of the medium. If 35mm film really was the equivalent of 22 to 40 mps, and if there was nothing more to photography than the number of mps and their equivalent in film, then outstanding enlargements of 20x24 inches or larger could be made from 35mm negatives as they can from digital backs in the 16mps range. But we all know that 35mm stinks with prints larger than 8x10 at best. There's much more to digital photography than the number of mps and IMHO trying to compare film and digital on the basis of equivalent mps for film is a pointless exercise., digital and film are totally different media.