35mm Film Res vs Digital Capture Res
 

The resolution of film.....

If one treats the output from film as any other signal processing problem then the point where noise becomes objectionable relative to the signal is the resolution limit. Considering ONLY 35mm film below (larger formats give different results), I've run the experiments described below myself.

For 400 ISO film, the equiv film resolution necessary to roughly equal a digital camera signal to noise ratio is, for a normal observer of an 11x14 image, a 3-4MP digital image from a camera with a good lens (Canon for example). I choose 400 ISO because really 100 ISO is too slow to use outside of a studio or a bright sunny day on the beach.

It is easy to check this. Shoot some Caucasian skin tones in natural light with a good 35 mm SLR. Obtain a good Nikon film scanner (a cheap scanner will add noise). Begin scanning at the Nikon's highest resolution. Examine the output on screen at a blow-up equivalent to 11x14 (takes a few seconds). Note the ugly film grain noise.

Then, set the resolution of the scanner down 15% and re-scan. Repeat the above.

Repeat until you arrive at the resolution where the noise becomes small relative to the (now blurred, lower res) signal. Take the file size of the film scanner output and divide by 3. This is the digital camera equivalent necessary to match the film speed signal to noise ratio. Now take a picture with digital cameras from 2 MP, 3MP, 5MP, and 6MP. Compare the results. 3MP will, for most observers, match the signal to noise of 400 ISO film at 11x14 size. If you do trilinear or bilinear interpolation the 3.2MP cameras are often better than ISO 400 film for 35mm cameras.

For 400 ISO color film (Kodak or Fuji....but Fuji will have less noise by quite a bit) a mildly discerning person can mostly get equivalent or better results with a 3-4MP camera.

For ISO 100 speed film a 6-8MP digital camera will give the same resolution results.

Backing up modern digital images to film introduces large noise, color shifts, could result in degrdation with time, etc.

The only reason to continue to use film is the following: If there is ever a widespread lack of electricity a few pictures of your kids, in print albums, will be valuable in the extreme. Photography was invented and perfected BEFORE widespread electricity availability.

We have only had reliable electricity for about 70 years. I continue to use film for key photographs of my children that I place in albums in fireboxes.

For everything else I burn DVD's and cross my fingers that electricity will be available in my old age.

I don't know what the highest resolution of the scanner is but the highest resolution for the film is limited by grain. If you aren't scanning the grain then you aren't scanning at films resolution.

Your assumption about 100 asa film is incorrect. I use 50/100/200 asa exclusively and many professionals use 50asa.

Saw another new book at the library just yesterday. It states 35mm is 18MP. It might have said at 100asa but I don't remember. Once again coroborating what I've seen before as being 12 mp to 25mp depending on who you're reading.

One of the wonderful things about the web is the amazing amount of info available at our fingertips. Norm Koren has done an amazing amount of work on this issue and you can read his findings at his site here. For a complete understanding of the issues and the underlying math to back up his findings I would allow a day or two for reading his entire site.

Mr. Koren's conclusions are, that 6 MP dSLR images meet or exceed film up to a print size of about 11x14 to 13x19. This is due to the relative lack of noise from these cameras. If you need larger prints, an 11 MP camera like the Canon 1Ds is required. He comments that the Canon 1Ds exceeds medium format film (6x7) in larger size prints, again because of the lower noise levels.

Yes. It's amazing the things you find on the 'net.

When moive studios digitize film for FX processing, they scan at 4k x 4k.