DSLR v Consumer Image quality

Discussion in 'Digital Photography' started by oink, Feb 22, 2005.

  1. oink

    paul Guest


    Is that the same thing or is there a special digital purple flavor
    fringe? I understand that CA is due to mis-scaling of different colors
    from the center and can be corrected with software that separates &
    matches those mis-scaled colors. The PS RAW plugin does this.
     
    paul, Feb 24, 2005
    1. Advertisements

  2. oink

    JPS Guest

    In message <>,
    Yeah, I guess with the zoom lens you need optimal aperture and the
    sweetest part of its range, perhaps, plus a shake-free exposure to get
    the last bits of sharpness out of the lens.

    The concept I was endorsing probably applies more to the 3MP/6MP models.
    --
     
    JPS, Feb 24, 2005
    1. Advertisements

  3. oink

    JPS Guest

    In message <>,
    I would not call CA "purple fringing".
    --
     
    JPS, Feb 24, 2005
  4. oink

    Larry Guest

    Niether would I, but from the way I see it discussed from time to time, a
    large amount of people cant tell the difference.

    I've seen the "fringing" on a Sony F-828 called C.A. (repeatedly) and it isnt
    C.A. its sensor "bloom".
     
    Larry, Feb 24, 2005
  5. oink

    Larry Guest


    I have that one two, and I use that one in 6mp mode because the difference is
    MUCH more noticable, and usually for the better. That one is my "outdoor
    nature shooter" because of the 10x lens.
     
    Larry, Feb 24, 2005
  6. You've missed the point: lots of people use their cameras only for email and
    web display. They always downsample. Presumably many people do it in the
    camera.

    Most of such people don't need a dSLR.

    David J. Littleboy
    Tokyo, Japan
     
    David J. Littleboy, Feb 24, 2005
  7. Because the eye is not a single frame snapshot camera. It is
    more like a video stream. The eye moves rapidly in small
    angular amounts and continually updates the image in one's
    brain to "paint" the detail. We also have two eyes, and our
    brains combine the signals to increase the resolution further.
    The we typically move our eyes around the scene to gather
    more information.

    Roger
     
    Roger N. Clark (change username to rnclark), Feb 24, 2005
  8. The eye has a different upper photon limit simply
    because the analog chemical system can tolerate more photons
    before becoming saturated. It is not an electronic sensor
    like a CCD or CMOS so does not hold photons in a "well" in
    the same way.

    What amazes me is the marvelous nature of our eyes.
    All the technology humans have developed pales in
    comparison the the human eye and brain processing
    that high resolution video stream real time.

    Roger
     
    Roger N. Clark (change username to rnclark), Feb 24, 2005
  9. Great point!
     
    Roger N. Clark (change username to rnclark), Feb 24, 2005
  10. Scott,
    Your post makes a lot of sense. The only thing I would add,
    is that given a higher resolution printer and higher
    resolution image, you likely could tell the difference
    at even higher DPI. The problem we consumers face is that
    technology is limiting what we can do, and that limit
    hasn't changed much in a few years because the manufacturers
    say "it is good enough" and they want to keep prices down.

    Here is a test I did using a 4x5 (650 mbyte image) and
    blind tests show the 600 ppi prints look sharper
    than 300 ppi prints:

    http://www.clarkvision.com/imagedetail/printer-ppi

    I must add, that perception is light level dependent.
    In poor light, it gets harder to see any difference
    between the prints.

    When I've had 4x5 film 8x10-inch prints done, either digitally
    or traditional chemical, the sharpness is astounding,
    and clearly beating 35mm film. But with many digital
    printers you may have difficulty seeing a difference
    due to printer limitations. (This doesn't mean that
    the 35mm or digital 8x10 prints are not
    outstanding too, just that it can get better.)

    Roger
     
    Roger N. Clark (change username to rnclark), Feb 24, 2005
  11. oink

    JPS Guest

    In message <>,
    I believe that the term "purple fringing" started with over-exposed
    sensors (ones covered with green filters) spilling their charge over
    into the neighboring sites, which have blue and red filters, causing the
    demosaicing algorithms to assume that this was red and blue light,
    coming through their own filters. Red plus blue equals purple.

    Then, people started (incorrectly) applying the term to CA.

    That's how I understand it.
    --
     
    JPS, Feb 24, 2005
  12. oink

    JPS Guest

    In message <>,
    They are quite different phenomena. CA is an optics issue. PF is
    partly a logical issue, as it only has color because of what software
    *expects* from those pixels.
    --
     
    JPS, Feb 24, 2005
  13. oink

    Mark² Guest

    That's not true at all.
    If all you wanted to begin with was an all-around lens (as is built into
    point-and-shoots), then you can certainly use a single lens with a DSLR and NEVER fiddle
    with lenses at all.
    Even with a single lens, you still have superior image quality, speed, focus, and control.
    There is absolutely no requirement of multiple lenses, or a fancy bag full of stuff, just
    because it is a DSLR.
    If one solution doesn't fit all, then why on earth are you singing the ANTI-praise of
    DSLRs?? If anything, that is an argument against point-and-shoots...not DSLRs.
     
    Mark², Feb 24, 2005
  14. oink

    Stacey Guest

    ???

    What has that got to do with judging the image quality of a camera using
    prints? Are you sugesting we using 75kb web size images to decide on which
    one is better?

    Rafe flips back and forth between saying we should be looking at email size
    images, then saying we should be thinking about huge size prints? 95%+ of
    camera users will never print anything larger than 8X10 and of the
    remaining 5% most won't print past 11X14's. Why worry about image quality
    issues that go beyond what you'll see in the print sizes you are going to
    make? People start looking at 200% crops instead of actually looking at the
    output they plan to use. Guess if you're a techo "nerd" that's what you
    would enjoy about digital photography? All the math theory, specs and
    looking at the pixel level? A photograpgher is going to look at the end
    product on their chosen output..

    The OP was asking about dSLR vs consumer image quality and how to best see
    it or at least that's what =WE= were talking about. What are -you- talking
    about?

    This looks more like you're just trying to find something negative to reply
    to one of my posts again? You and rafe still playing tag team? :)
     
    Stacey, Feb 24, 2005
  15. Nothing. You missed the point. Which was that "a lot of folks aren't
    interested in prints".
    Only if that's your only use...
    My experience here is that my printer (Epson R800) can render more detail at
    A4 than a 6 or 8MP camera can capture. So if I compare images at 100%
    pixels, print both at 300 dpi, and show the prints to someone, the one they
    pick as better invariably corresponds to what I saw at 100% pixels.

    So at the print size I actually make (A4), looking at 100% pixels is quite
    valid.

    David J. Littleboy
    Tokyo, Japan
     
    David J. Littleboy, Feb 24, 2005
  16. oink

    Scott W Guest

    This has not been my experience. I have found that it is what looked
    like it has the most detail on the screen will not necessarily produce
    the best print. In particular a scan of a 35mm negative will show more
    detail then the same shot from a digital camera but when both are
    printed at 8 x 10 the digital will look sharper to most people. And by
    the same token noise that you can see on the screen will often be low
    enough in contrast that when printed at 300 dpi it disappears.

    Scott
     
    Scott W, Feb 24, 2005
  17. Ah, that's a different question<g>. I was assuming rather similar images
    that are subtly different on the screen. Like two different digital cameras
    or two different edits of the same image.
    That's certainly true. Noise isn't a problem unless you try to do something
    extreme, like boost shadows, sharpen excessively, or shoot at too high an
    ISO.

    David J. Littleboy
    Tokyo, Japan
     
    David J. Littleboy, Feb 24, 2005
  18. oink

    Scott W Guest

    I used Roger's test pattern, with the ppi set at 600, and printed out
    the image, the link below is a scan of this print scanned at 3200 dpi.
    I will say that this pushes both my scanner and my printer.

    http://www.sewcon.com/300dpi_test/res-chart.jpg

    What you can see is that my printer can resolve 300 lppi in the
    horizontal direction and can easily resolve 150 lppi in the vertical
    direction.

    Roger's pattern can be found at this link
    http://www.clarkvision.com/imagedetail/printer-ppi

    I would suggest that people print this chart out to get an idea of
    just how small even 150 lppi is. Keep in mind that the printout is very
    high contrast and that lower contrast content will be much less
    visible.

    With my glasses off I can just resolve the vertical lines at 300 lppi.
    With my glasses on I can just resolve the 150 lppi.

    Now my guess is that all of us feel we have good eyes and can see a lot
    of detail, so show the prints to a number of other people
    non-photographers and ask then what the smallest lines that they can
    make out are, I would be interested to know what most people can see.
    My optics book (Modern Optical Engineering by Warren J. Smith) would
    say that a normal person should be able to see around 200 lppi when
    viewed at 12 inches but only when the contrast is high. The visible
    lppi will go up a small amount under brighter illumination.

    Scott
     
    Scott W, Feb 24, 2005
  19. oink

    andrew29 Guest

    You may argue that a 500% blowup on screen is too revealing, and shows
    defects that don't matter. Fair enough.

    However, it is not necessary to to make a print to "properly see" an
    image. Unless you're arguing that properly seeing actually *requires*
    the losses inherent in printing, which would be an opinion, albeit a
    rather preposterous one.

    Andrew.
     
    andrew29, Feb 24, 2005
  20. Paint Shop Pro has chromatic aberration reduction as a built-in function,
    which also works well on purple-fringing. An excellent reason to upgrade
    from PSP8 to PSP9.

    David
     
    David J Taylor, Feb 24, 2005
    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.