Am I oversharpening

Discussion in 'Digital Cameras' started by Celtic Boar, May 30, 2005.

  1. Celtic Boar

    Skip M Guest

    Monochrome does not mean greyscale, greyscale means greyscale. Monochrome
    means one color, whether it be red, blue, green, cyan, magenta, yellow, or
    shades of grey. Thus, Bayer sensors are monochrome, since they only sense
    one color on each pixel.
     
    Skip M, Jun 25, 2005
    1. Advertisements

  2. Celtic Boar

    Skip M Guest

    If he made a half million on the crop of that image, he'll hear from my
    lawyer, because we just sent in the CD of that shoot to the copyright office
    for registration.
     
    Skip M, Jun 25, 2005
    1. Advertisements

  3. Celtic Boar

    Bill Funk Guest

    Right.
    Canon and Nikon and all the other digital camera manufacturers are
    buying their own products to make it look like sales are great, then
    just dumping the product in the ocean because no one will use them.
    And Sports Illustrated lies.
    And Canon just offered me $950US for my Digital Rebel/300D, because
    B&H screwed up and sold it to me.

    Where do you come up with this stuff?
     
    Bill Funk, Jun 25, 2005
  4. Celtic Boar

    Bill Funk Guest


    ".. as a woman..."??!!
    Your credibility is running away from you.
     
    Bill Funk, Jun 25, 2005
  5. Celtic Boar

    Bill Funk Guest

    Hint:
    So are Foveon sensors. The color data is added later, depending on
    which sensor data is being looked at.
     
    Bill Funk, Jun 25, 2005
  6. Celtic Boar

    Bill Funk Guest

    Show a few.
    Done. Which ones say he invented the CCD?
    Not even his various bios on sits that extol his work while at Bell
    mention this, so far as I can see.
    Maybe I didn't read enough of the entries, but since you say they are
    there, a link would help to give you some shred of credibility.
     
    Bill Funk, Jun 25, 2005
  7. SNIP
    Credibility?
     
    Bart van der Wolf, Jun 25, 2005
  8. Sure. Why should you contradict yourself?
    Yawn.

    Bart
     
    Bart van der Wolf, Jun 25, 2005
  9. Celtic Boar

    Ray Fischer Guest

    It's easy to create a sockpuppet to agree with your bullshit.
     
    Ray Fischer, Jun 25, 2005
  10. Celtic Boar

    Ray Fischer Guest

    That's because you're nuts. A preddy sockpuppet.
    LOL! Kook!
     
    Ray Fischer, Jun 25, 2005
  11. Here's a link from Lucent, the current incarnation of Bell Labs. It
    talks about Boyle and Smith, the inventors of the CCD. No mention of
    Carver Mead:
    http://www.lucent.com/press/0699/990630.blb.html
    I did. The only thing in the first couple of pages linking the two was
    a neural network chip done at Bell Labs that credited Carver Mead's
    previous design. This chip is not an image sensor; it has nothing
    to do with CCDs or CMOS image sensors.

    So far, I've seen nothing to indicate Carver Mead was ever at Bell Labs.
    Are you simply making this up out of thin air?

    Dave
     
    Dave Martindale, Jun 26, 2005
  12. Celtic Boar

    Ken Tough Guest

    It's ridiculous. Neither the foveon and the Bayer sensors detect
    colour. They both detect the brightness of components in Red,
    Green, Blue. The only difference is one detects all three at
    nearly the same physical point.

    To truly detect colour, you would have to detect the light spectrum
    across the pixel area of interest.
     
    Ken Tough, Jun 26, 2005
  13. Celtic Boar

    Bill Funk Guest


    Well, let's try that again, without the creative cutting that allows
    you to redefine what I said...

    "So what?
    Sigma claims the SDx are 10.3MP cameras; what do they know?
    Trying to redefine industry-wide standards to meet the desires of
    Sigma/Foveon is your best trick, but it's all smoke. Not even mirrors,
    just smoke. "

    Are you trying to say that each sensor site in an X3 sensor is a
    "pixel"?
    Or, what is it you're trying to say?
     
    Bill Funk, Jun 26, 2005
  14. Celtic Boar

    JPS Guest

    In message <>,
    That sounds like all 8M sense the same monocolor, which is not true.
    There are 4M green, and 2M each red and blue, all in distinct (8M
    different) 2D locations.
    .... or over twice as many 2D locations.
    No, that would be stupid, as it means shifting recorded data spatially,
    causing geometric shifts between the color channels. That is not what
    happens in bayer demosaicing, by any measure. Combining arbitrary 2*2
    bayer-filtered pixels only makes sense for binning greyscale data.
    And none of them (that are correct) say that each 2*2 tile is one
    optical unit, or anything else so ridiculous. It's more a case of "Hi,
    I'm a red-filtered sensor, and I witnessed X amount of light at my
    location. My green- and blue-filtered neighbors experienced different
    values, not just because they have different filters, but also because
    they are in different 2D locations".
    --
     
    JPS, Jun 26, 2005
  15. Celtic Boar

    Ken Tough Guest

    Yes, that's where he's wrong. Each sensor (three of which are
    stacked vertically) is not a "picture element" [pixel]. The three
    together are a picture element, because they are the same point in
    2-D space. You can't win both ways. Either they are stacked
    vertically and sample the point in the image (3.3M pixels) or
    they don't (10.5M monochrome pixels).

    There are 3.5M pixels. There are 10.5M monochrome sensors.

    You say there are 10.5M monochrome pixels in a foveon, but there
    aren't. That is true of the Bayer pattern sensor cameras.
     
    Ken Tough, Jun 26, 2005
  16. Celtic Boar

    JPS Guest

    In message <>,
    Well, actually, you can't *see* the dynamic range as presented; dark
    shadows can be noisy and/or posterized; it's not until you boos the
    shadows that you see what lurks there. Luminance-wise, I agree, the
    dynamic range is excellent here, and the deepest shadows (when boosted
    in software) are posterized (as is the case with all 12-bit capture) but
    almost noiseless.

    The problem is, the green and blue levels in the shadow rocks vary, in
    large clumps, so the color is a bit off. If the shadows were of
    blue/green subjects, they would be much worse.

    Great greyscale dynamic range, though, for 12-bit capture.
    Well, actually, the snow is blown (255,255,255) in many areas.
    The snow is blown in many areas. The bright snow areas just happen to
    be thin, so you don't see large expanses without detail. *that* is the
    trick.

    --
     
    JPS, Jun 27, 2005
  17. Celtic Boar

    Ken Tough Guest

    The objective tests at dpreview show that at high ISO the SD10 loses
    saturation and hue accuracy. I suppose this shows up as performance
    in the shadows too.

    http://www.dpreview.com/reviews/sigmasd10/page13.asp
     
    Ken Tough, Jun 27, 2005
  18. Celtic Boar

    JPS Guest

    In message <>,
    One ISO's highlights are another ISO's shadows. They're just the same
    analog sensor signals digitized into different RAW number ranges (and
    precisions).
    --
     
    JPS, Jun 27, 2005
  19. Celtic Boar

    JPS Guest

    In message <>,
    No, that is what is needed for a three-channel *input* pixel, and it is
    no guarantee of accurate color.
    No, they do not, as they have unique spatial information for each of the
    advertised pixels.
    --
     
    JPS, Jul 2, 2005
  20. Celtic Boar

    Paul Heslop Guest

    can I ask what articles please? I'm only seeing the last two posts in
    this thread and it's a subject of interest to me
     
    Paul Heslop, Jul 4, 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.
Similar Threads
There are no similar threads yet.
Loading...