Nikon 8000 scanner, banding

Discussion in 'Nikon' started by nobody nowhere, Oct 13, 2003.

  1. Thank you Bill, once again. Here is the latest, on the assumption that
    your patience has not run out:-

    I looked at the slide again, the relevant area is really very opaque,
    nearly black. But I could not detect any "bands" when, as advised by
    someone else, I put the slide on a light box, and looked at it with a 22
    times magnifying glass. I did the adjustments in 14 bit mode, but I was
    still getting the "bands" on print. However, when I magnified the same
    picture up to 300 per cent, and 200 per cent, there was no sign of
    banding on the monitor. I interpreted this to mean that the manipulated
    image in Photoshop caused the printer to "see" bands where there were

    I did however solved the problem by resampling to 300 dpi (you may
    recall that the "banding" pictures had an original dpi of nearly 800, no
    less). When I resampled to 300 dpi, there was no longer "banding" on
    the print, and the relevant part became less unacceptable than before.
    I took this to mean that when the Epson received a 300dpi file, it was
    able to see that there were no "bands" for it to reproduce (rightly or
    wrongly). Perhaps not good enough for our friend Rafe, whose standards
    are higher than mine, but acceptable to me. A copy of this latest
    picture could be seen at

    and the file name is *kenwoodhouse-9*

    Thanks again for your interest.

    PS. I don't know how to make a mask for 16 bit, and should be obliged
    for your advice.

    nobody nowhere, Oct 15, 2003
    1. Advertisements

  2. Sorry, apparently I exceeded my web limit, and was unable to upload the
    relevant file to my web site.

    nobody nowhere, Oct 15, 2003
    1. Advertisements

  3. nobody nowhere

    Rafe B. Guest

    Talk about a stretch. In what way do you imagine 16 bit processing
    could "fix" this image problem?

    rafe b.
    Rafe B., Oct 16, 2003
  4. Thank you very much Bill, I am very grateful. The only thing I can
    teach in return is how to waste large amounts of Super A3 genuine Epson
    paper... :)

    nobody nowhere, Oct 16, 2003
  5. Conclusion. I have fiddled a little more, and my feeling is that Bill
    has intuitively identified the problem. I don' t remember exactly what
    I did (sorry), but I think that I converted from 14 bit to 8 bit,
    selected the offending area, saved it on the alpha channel, loaded it
    back on the 14/16 bit picture, and tried to lightened up with levels (as
    suggested by Bill), in my heavy handed style. As I was moving to
    extremes, I could clearly see the bands appearing on the selected part
    of the picture *on my monitor* (previously I could not see the bands on
    the monitor, but only on the print). As I was being less "radical"
    (in levels), the bands gradually disappeared, as if, this time, levels
    were telling me where the limits were. The moral of the story, for
    those who followed this post: the bands were due neither to the scanner
    (Nikon 8000), nor the printer (Epson 1290), but to excessive
    manipulation of "levels" in Photoshop.

    nobody nowhere, Oct 16, 2003
  6. nobody nowhere

    Rafe B. Guest

    I had a good look at the images in question and looked closely
    at the trouble spots. The area in question was extremely dark and
    grainy and had a few other weird artifacts as well. From past
    experience I guessed that the extreme grain in particular resulted
    from an extreme tonal move -- ie., a very steep curve to induce
    some detail where before there had been almost none.

    That increase in contrast (from the steep curve) also is what
    brought out the streaks/scratches -- what the OP refered to as

    There is simply no way that more bit depth in the processing would
    have prevented those streaks or scratches from appearing.

    The OP's tonal move did exactly what he intended. It brought out
    details that had formerly been hidden in shadow. Unforutnately,
    it also brought out extreme grain and some streaks/scratches
    that had been lurking in those shadows. There's no free lunch.

    Chromes can be a bitch to scan, or to expose properly in the first

    rafe b.
    Rafe B., Oct 17, 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.