Digital v Film

Discussion in 'Digital Photography' started by Simon Marchini, Jul 13, 2003.

  1. Simon Marchini

    Mxsmanic Guest

    With a 100-lpi screen, that's true, and 100-lpi is typical for many
    types of offset printing. I wasn't talking about photographic prints,
    which don't use halftone screens.
     
    Mxsmanic, Jul 19, 2003
    1. Advertisements

  2. How does it look at that size?
     
    Tony Whitaker, Jul 19, 2003
    1. Advertisements

  3. That depends on the purpose, define what you intend, Inkjet Lambda etc?
    I think it depends on the original, given a Tack sharp original? I only
    waste my time on good originals.
    Ok do you have a visual confirmation, as in an example print which
    looks better than a typical PS printed file?
    Define artifacts, I am interested in that description as I have a theory about them.

    --
    Check out My Homepage at
    http://members.bellatlantic.net/~gblank

    Support bacteria - they're the only culture
    some people have." -Stephen Wright
     
    Gregory W. Blank, Jul 19, 2003
  4. So then your scanning at 2700 dpi at the full size scaling
    for 8x10. Thats a big file, does the printer not resample the
    file down anyway, to say 400 DPI?
    How do you establish the evidence of this?

    --
    Check out My Homepage at
    http://members.bellatlantic.net/~gblank

    Support bacteria - they're the only culture
    some people have." -Stephen Wright
     
    Gregory W. Blank, Jul 19, 2003
  5. The issue with the 2450 and similar non-dedicated filmscanners is that the
    result is inferior in a direct comparison with e.g. an LS-8000 (at let's say
    8x10inch and up). The good thing however, is that lacking a direct
    comparison, a bit of effort can come along way in mimicing quality (the 3200
    model is probably a notch better).
    Nevertheless, the 2450 just lacks overall detail contrast due to internal
    relections, two additional air/dirty glass surfaces, and a fixed focus lens
    (and the film is likely to curl during the scan because it is fully lit
    instead of by a slit light source). It limits the film+scanner resolution
    (in terms of 10% MTF response) to something in the 30 cycles/mm range, so
    that could produce an uncropped (length wise) 8x10 inch of 8 line pairs/mm
    or 400 ppi (e.g. matching Durst Lambda input), which is generally considered
    as good resolution at normal viewing distance. A comparison with an LS-8000
    scan is still obvious, especially above 8x10 inch.
    Yes I do, but that is not a very quantifyable comparison, because Qimage
    even makes something nice from worse input due to its interpolation methods
    like Lanczos and "Vector" and my corrected vision is better than average. My
    general impression is, it's sharper, and that's what I hear from other users
    as well (e.g. in the Yahoo support group for QI). You can try it yourself on
    your prints, it's mostly functional shareware for 30 days. The theory behind
    it can be found on this already somewhat dated (the program is improved in
    the mean time) page http://www.ddisoftware.com/qimage/quality/ .
    about them.

    Not very scientific, but clearly visible haloing instead of the perception
    of edge definition, and subject surface structure/texture. It can also
    manifest itself as interference/moiré instead of subtle detail (try printing
    something like this http://www.worldserver.com/turk/opensource/z128.gif at
    various printer driver resolutions). The image content itself is also
    important, because the Human Visual System (HVS) makes up detail where it
    expects it. The HVS is complex, and according to the literature not yet
    fully understood. But many known factors, such as ambient light level, are
    known to have its effect on what we think that we see.

    Bart
     
    Bart van der Wolf, Jul 20, 2003
  6. Looks very very good to me. Better than prints from 35mm negatives
    that size I've seen -- but I haven't seen that many prints from 35mm
    negs that big, have made only a few myself, and the ones I made myself
    were years ago and hence from less good negatives (and B&W rather than
    color). So I can't of my own knowledge say it looks better than a
    print from 35mm film would have looked if shot with *today's* films.
     
    David Dyer-Bennet, Jul 20, 2003
  7. Simon Marchini

    Mxsmanic Guest

    But you can. That's one reason why digicam colors look smooth and
    candy-like: they are simply smeared over several pixels. And if you
    look a bit closer, you see that digicams "color outside the lines" a
    lot.
    The smearing of colors in most video recording formats far exceeds that
    produced by matrix filters on single-CCD video cameras, so that's not
    it.
     
    Mxsmanic, Jul 21, 2003
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.