This posting is to correct information I posted on APUG about color film. I'm posting it here, because even though it involves film, a few folks on APUG probably don't want to see it.
Correction:
It's been almost 25 years since I last worked with the PhotoCD Format so I was mixing the figures up.
In PCD, the highest standard resolution was called 16 Base and was 2048x 3072 (6.291 Mpx) with an uncompressed file size of 18 MB. See https://en.wikipedia.org/wiki/Photo_CD
The file size was established in the Research Labs who looked at lots of images and decided that the 16 Base size adequately represented consumer 35 mm images due to issues like flair, camera shake, misfocus, etc. It was not designed to represent the maximum image quality possible.
Addition:
From a presentation I used to make in the early 1990's while at Kodak:
There is no generally accepted way to convert an AgX image into its 'PIXEL' equivalent, but we can approximate the pixel equivalent in the follow way:
Resolving power of a typical 100 speed color negative film at medium contrast is around 65 line pairs/mm.
A line pair represents a BLACK and a WHITE line so a line pair represents two pixels.
This means the film has around 130 pixels/mm.
For a 24 x 36 35mm frame this gives 3120 x 4680 = 14.6 million pixels.
The Nyquest Criterion (Sampling Theorem) says that to fully capture all the information, we must sample the information at a rate at least twice as high as the highest frequency detail we want to capture.
Thus, to fully capture all the information on this film, we must use a scanner that can capture at least 6240 x 9360 = 58.4 million pixels.
Note this is the maximum information content. Things like flair, camera shake, misfocus, etc, will significantly reduce it.
Correction:
It's been almost 25 years since I last worked with the PhotoCD Format so I was mixing the figures up.
In PCD, the highest standard resolution was called 16 Base and was 2048x 3072 (6.291 Mpx) with an uncompressed file size of 18 MB. See https://en.wikipedia.org/wiki/Photo_CD
The file size was established in the Research Labs who looked at lots of images and decided that the 16 Base size adequately represented consumer 35 mm images due to issues like flair, camera shake, misfocus, etc. It was not designed to represent the maximum image quality possible.
Addition:
From a presentation I used to make in the early 1990's while at Kodak:
There is no generally accepted way to convert an AgX image into its 'PIXEL' equivalent, but we can approximate the pixel equivalent in the follow way:
Resolving power of a typical 100 speed color negative film at medium contrast is around 65 line pairs/mm.
A line pair represents a BLACK and a WHITE line so a line pair represents two pixels.
This means the film has around 130 pixels/mm.
For a 24 x 36 35mm frame this gives 3120 x 4680 = 14.6 million pixels.
The Nyquest Criterion (Sampling Theorem) says that to fully capture all the information, we must sample the information at a rate at least twice as high as the highest frequency detail we want to capture.
Thus, to fully capture all the information on this film, we must use a scanner that can capture at least 6240 x 9360 = 58.4 million pixels.
Note this is the maximum information content. Things like flair, camera shake, misfocus, etc, will significantly reduce it.