Re: Which scanners REALLY provide 36 bit output? HP?

From: Warren Young (
Date: Sun Dec 10 2000 - 19:59:11 PST

  • Next message: Daniel Kerry Hiltgen: "Re: seg-faults for network access"

    Steffen Kluge wrote:
    > On Sat, Dec 09, 2000 at 11:00:06AM +0800, Steve Underwood wrote:
    > > > represent all the information possible. For 19th century photography I
    > > > believe 1200 dpi comes very close to this.
    > >
    > > This is about the resolution of most modern emulsions (unless you are using
    > > super slow specialist films). I doubt the 19th centruy emulsions have
    > > anything like that resolution. 19th century lenses are incapable of
    > > resolving this well, anyway. Such a high resolution would only make sense
    > > when scanning documents, or maybe paintings.
    > >
    > > The resolution of most film images are much poorer than most people
    > > realise.
    > Well, all decent 100ASA slide or negative emulsions are capable
    > of recording in excess of 100 lines/mm, at a contrast of
    > 1:1000. That would be 2500lpi. Good lenses can get across maybe
    > about 1200~2000lpi. Keep in mind that the pixels in a film
    > emulsion are scattered (and different in size), so you need a
    > lot more pixels in an orderly rectangular grid to get the same
    > visual resolution.

    It's worse than that. First assume a perfect photo: stable camera,
    perfect focus, proper exposure, no push/pull processing, etc. Given
    that, the maximum resolution you quote would only apply to objects
    exactly at the focus plane. Old photos especially have low depth of
    field since they couldn't afford to stop down, since the exposure time
    for the films and photo plates of the time was so low. Therefore, not
    everything in the photo is technically "in focus", and therefore will
    not have the best theoretical resolution.

    Let's also not forget that these photos have doubless deteriorated.

    1200dpi should be plenty of resolution for this work. Higher bit depths
    will be a plus, though, since that will capture delicate tonal
    gradations that will be useful when you begin fiddling with stain
    removal, histogram stretching, etc.

    = Warren Young -- See my Palm pages at:
    = ICBM Address: 36.8274040 N, 108.0204086 W, alt. 1714m

    -- Source code, list archive, and docs: To unsubscribe: echo unsubscribe sane-devel | mail

    This archive was generated by hypermail 2b29 : Sun Dec 10 2000 - 19:58:09 PST