Discussion of 16 bit image editing vs 8 bit

Discussion in 'Photoshop' started by Robert Feinman, Nov 4, 2003.

  1. Since the new version of photoshop adds improved 16 bit image processing I
    decided to see if there is much difference between editing 16 bit and 8 bit images.
    So I've added a tip highlighting the differences between using 8 and 16 bit
    depth files for capture and editing to my tips page.
    I did the same edits on both versions of the image using photoshop 7.

    I leave it to you to decide as to the results, I'm still pondering the
    meaning..


    Follow the link to the tips section on my home page and then scroll down
    to the New! tips.
     
    Robert Feinman, Nov 4, 2003
    #1
    1. Advertisements

  2. Robert Feinman

    Mike Russell Guest

    Good page.

    One thing to keep in mind is that PS adds 1/2 bit of noise when it converts
    from 16 bit to 8 bit. I've seen discussions of 8 vs 16 bit that ignore this
    and rely only on histograms to illustrate the "point" that 16 bit is
    superior - yours is a refreshing change from this because you actually
    compare the two images..

    Much of the difference between the two images, including the histogram
    combing, may be due to that factor alone.

    --

    Mike Russell
    http://www.curvemeister.com
    http://www.zocalo.net/~mgr
    http://geigy.2y.net
     
    Mike Russell, Nov 5, 2003
    #2
    1. Advertisements

  3. Robert Feinman

    Anonymous Guest

    Well......The 16 bit, at that enlargement is what I would call a "tad"
    sharper. Not a very technical term, I know. The question would be,
    would the difference be perceived in a normal print by the average
    viewer? Probably not. At an extreme enlargement, pixalization becomes
    the overiding factor, I think.
    My ideal print is one that is "perceived" to be as sharp an image as one
    could get. Going back to my photo printing days, I remember the endless
    lens testing, shooting and printing resolution charts to achieve the
    sharpest possible image only to be defeated by the grain of the film.
    Then there was the diffuse light sources vs. condensors.
    To add to that, if you are old enough, there was the old debate over
    Panatomic X vs. Plus X, and Microdol develpor vs. D 76 and a myriad of
    other sharpness issues! The quest for sharpness never ends, but that's
    what make this field interesting and challenging.
    Robert
     
    Anonymous, Nov 5, 2003
    #3
  4. Robert Feinman

    Flycaster Guest

    I'm unclear here. At what point did Photoshop "convert" this image? I
    thought the page indicated that they were two discrete scans.
     
    Flycaster, Nov 5, 2003
    #4
  5. Robert Feinman

    Warren Sarle Guest

    A grainy, film-based image is not the place you are likely to see any
    difference between 8 and 16 bit editing. Posterization is more likely
    to show up in a digital photo of a very smooth subject with a shallow
    gradient that you want to change to a steep gradient, such as a light
    blue sky that is burned in for dramatic effect. Also, stacking up lots
    of adjustment layers and using Lab mode make posterization worse.

    Although I routinely use 16-bit mode for any major adjustments, I
    doubt that anybody is likely to see any difference between 8 and 16
    bit editing for ordinary photographs that are reasonably well
    exposed. But I do some weird things where it makes a big difference.
     
    Warren Sarle, Nov 5, 2003
    #5
  6. Robert Feinman

    Mike Russell Guest

    Flycaster wrote:

    [re Photshop's treatment of 16 bit vs 8 bit images, c.f.
    http://robertdfeinman.com ]
    There are two different scans in Robert's example.

    Photoshop converts from 16 to 8 bit when it displays the image, and when
    calculating the histogram.

    In both cases, the 1/2 bit of noise involved in that 16->8 bit conversion
    can cause a change in appearance that would smooth out posterization and
    banding, and also smooth out the histogram. This is just one of the issues
    that make comparison of actual 16 vs 8 bit images problematical.

    --

    Mike Russell
    http://www.curvemeister.com
    http://www.zocalo.net/~mgr
    http://geigy.2y.net
     
    Mike Russell, Nov 5, 2003
    #6
  7. Robert Feinman

    Flycaster Guest

    Huh, I did not know this about the histogram; I'd always thought it was
    based on the actual underlying pixel data unless it was cached. It's odd
    that Fraser/Blatner have never mentioned this quirk before, and it would
    indeed make a true comparison problematic, or at the least somewhat
    inaccurate.
     
    Flycaster, Nov 5, 2003
    #7
  8. Robert Feinman

    Hecate Guest

    And sharpness is not the whole issue either. I still use, and will
    gone on doing so until they either stop making it or death overtakes
    me (whichever comes first <g>) Kodak's Tri-X as my standard B&W film.
    Why? Because I love the grain - i positively *want* to see the grain
    ;-)
     
    Hecate, Nov 6, 2003
    #8
  9. Robert Feinman

    Anonymous Guest

    To which I will add, that I agree. Were it not for my advanced age which
    limits the time I can stand in a darkroom, I would still be doing the
    same. I first and only preference was for Tri-X developed 1:1 in D76.
    Nothing better!
    Robert
     
    Anonymous, Nov 6, 2003
    #9
  10. Robert Feinman

    Hecate Guest

    I'm glad there's still a few of us left :)
     
    Hecate, Nov 7, 2003
    #10
    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.