SD9s now selling for 1/3rd the price of the 10D

Discussion in 'Digital Photography' started by George Preddy, Dec 14, 2003.

  1. The standard Foveon donkey would have 4 legs, but George would count it
    as 12 using some obscure reasoning.

    On the other hand, George would declare that the standard Bayer donkey
    really has only one "full colour" leg. Despite that, the donkey would
    not fall over, and would walk normally.

    Dave
     
    Dave Martindale, Dec 18, 2003
    1. Advertisements

  2. George Preddy

    Alan Browne Guest

    An infinite number. Just choose an infinitely fine division.
     
    Alan Browne, Dec 18, 2003
    1. Advertisements

  3. George Preddy

    Alan Browne Guest

    Again: There are 6M pixels at "positions" on the 6Mpix sensor the
    resolution of information is real. The color is borrowed from adjacent
    pixels.

    Again: real iamges show the difference, and the SD-9 is just not as
    sharp as the 10D. Face it.
     
    Alan Browne, Dec 18, 2003
  4. For practical applications, 5 or 10 nm steps (about 64 or 32 bins to
    cover the visible range) seems to be enough.

    Dave
     
    Dave Martindale, Dec 18, 2003
  5. Bayer 6MP-interpolated cameras borrow 75% of their output pixel's color data
    from adcent sensors. There are only 1.5MP of discrete full color data in
    any image. The 10.3MP SD9 never borrows any colors from any other pixels.
     
    George Preddy, Dec 18, 2003
  6. Borrowing colors from adjacent pixels = blurry.
    The 10D has about half the optical resolution of color 35mm film, the SD9
    enjoys rough parity with 56mm medium format color film. As a result, 10D
    images are blurry even at relatively small print sizes like 8" x 10", while
    SD9 can create extrememly sharp photographs up to 40". In fact, the 10D
    can't even create a non-interpolated color photograph. No Bayer camera can
    do such a thing. To take advantage of the extreme high end, 68 billion
    non-interpolated color photographic emulsion printers like Lighjets and
    Dursts, you can't start with color interpolation and expect any benefit.
     
    George Preddy, Dec 18, 2003
  7. George Preddy

    Junque Guest

    Except the human eye is only sensitive to three colours, providing our
    photographic image cause the signal in the optic nerve to be identical
    to the original scene we have achieved the objective for normal
    photography, there will be specialist exceptions, i.e. false colour. We
    can leave the hundreds or thousand for metrology.
     
    Junque, Dec 18, 2003
  8. George Preddy

    Mark Herring Guest

    Just when I thought I had seen George's stupidest statement, he goes
    one better.

    Let's see now: Film is normally good for around 80lp/mm. By that
    measure:

    35mm: 24x36 = 1920x2880. Now we need 2 samples per lp (Nyquist sez)
    So we have 3840x5760 samples = 22Mp

    You cant match the best 35mm film with 10Mp, much less the 3.4Mp in
    the SD9. Medium format takes mabye 2x more depending on which exact
    format you are using.

    George---you are a really bad joke
    **************************
    Mark Herring, Pasadena, Calif.
    Private e-mail: Just say no to "No".
     
    Mark Herring, Dec 18, 2003
  9. George Preddy

    Junque Guest

    Se you acknowledge that an SD9 has only 3.43 million pixels.
     
    Junque, Dec 18, 2003
  10. At which point Greg spontaneously burst out laughing. The kids outside
    start laughing back at him....

    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
    <"56mm">
    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
    <Ooops, nearly fell out of the chair. Can't breathe so well...>
    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
    <"Well, maybe if you scanned it at 500 dpi and 8 bit color.">
    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
    <"Only on Planet Preddy.">
    LOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOLLOLOLOLOLOLOLOLOL!
     
    Greg Campbell, Dec 19, 2003
  11. George Preddy

    Alan Browne Guest


    My stomach hurts. From laughing.
     
    Alan Browne, Dec 19, 2003
  12. George Preddy

    Lionel Guest

    "Two legs good, four legs bad! Baaaaaa!"

    (Wondering how many people will get this reference...)
     
    Lionel, Dec 19, 2003
  13. George Preddy

    JPS Guest

    In message <QVu3RiFmUP4$>,
    It sounds more practical to me to have bands than to try to capture the
    exact frequency of every individual photon.
    --
     
    JPS, Dec 19, 2003
  14. George Preddy

    JPS Guest

    In message <brt9vs$nul$>,
    Nope. Not at all, unless all the detail is chroma and the luminance is
    constant.

    In most cases a heavy blur to the chroma while maintaining the luminance
    is barely, if at all, perceptible.
    --
     
    JPS, Dec 19, 2003
  15. George Preddy

    JPS Guest

    In message <brrosi$kkn$>,
    No, you pathetic liar. *YOU* brought up the 68M figure. I mentioned
    how it is derived; it is 4096 to the 3rd power.
    I said it can't see anywhere near 68 million, in reality. Do you have
    any ethics, whatsoever? For someone who tries to come across as being
    an ethical person, you have no ethics at all. You are a moral
    basketcase.
    That would mean that all your posts shouldn't have been posted.
    --
     
    JPS, Dec 19, 2003
  16. http://foveon.com/faq_technology.html#FAQ_tech_13
     
    George Preddy, Dec 19, 2003
  17. Right, just re-use the same data in over and over, it doesn't hurt a bit,
    trust us. In fact, a 4 sensor Bayer interpolated up to 14M output pixels
    has better resolution that an SD9, or even a lowly 1.5MP 10D. Just
    interpolate more, I mean moire, I mean more, it's a free lunch.
     
    George Preddy, Dec 19, 2003
  18. Housing 10.3M sensors in 3 layers, same as any 3 chip camera. The 10D is
    only a 6MP-interpolated 1 layer camera, that means almost double the
    bandwith to carry less than half the optical color resolution.
    http://www.outbackphoto.com/artofraw/raw_05/essay.html

    Any digital camera that interpolates shouldn't be considered professional,
    film doesn't interpolate, and that is the pro standard.
     
    George Preddy, Dec 19, 2003
  19. Already done, the SD9 has 2.3X the optical resolution of the 1Ds...
    http://www.outbackphoto.com/artofraw/raw_05/essay.html
     
    George Preddy, Dec 19, 2003
  20. George Preddy

    imbsysop Guest

    go re-read yr own credo at the foveon site and find out for yrself
    what sort of stupid claims you defend ..
    http://foveon.com/X3_tech.html

    or are you to dumb to understand a tech ad ?
    10.3sensorsper location/3layers = 3.4 pixel positions .. period
    even if you would sense all the colors of the rainbow at the given
    location it would still be a 3.4 Mpx chip

    or maybe you just don't understand the drawing at
    http://foveon.com/X3_better.html
    where they've drawn the color arrows *next" to each other instead of
    *over* each other ..

    funny you keep on defending things that you seemingly don't even
    understand ..
     
    imbsysop, Dec 19, 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.