Bayer Color Mask For CCDs

Discussion in 'Digital Photography' started by Robert E. Williams, Nov 29, 2003.

  1. I just got to wondering how the RGB filters are applied to a CCD in a
    Bayer pattern.
    A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    microns on a side.
    That is VERRRY small. Less than half the size of a red blood cell!!
    About 1/25 the diameter of a human hair!!
    How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    registration with no overlap or leakage between dots on a sensor 5.4mm x
    7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.

    Does anyone know how this is done?
    Bob Williams
    Robert E. Williams, Nov 29, 2003
    #1
    1. Advertising

  2. On Fri, 28 Nov 2003 16:45:31 -0800, "Robert E. Williams"
    <> wrote:

    >I just got to wondering how the RGB filters are applied to a CCD in a
    >Bayer pattern.
    >A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    >microns on a side.
    >That is VERRRY small. Less than half the size of a red blood cell!!
    >About 1/25 the diameter of a human hair!!
    >How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    >registration with no overlap or leakage between dots on a sensor 5.4mm x
    >7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    >
    >Does anyone know how this is done?
    >Bob Williams


    In terms of chip fabrication (photolithography), the size of an
    individual CCD sensor of 3.1um is large by modern standards. To put
    it into context, the details of individual components on a modern CPU
    are about 0.9um across, hundreds of times less area.

    "Britney Spears" tells you all about it (and other semiconductor
    principles) in fairly simple terms here:

    http://britneyspears.ac/physics/fabrication/photolithography.htm

    Or you can Google for many, many, MANY more...

    Andy
    Andy Blanchard, Nov 29, 2003
    #2
    1. Advertising

  3. Robert E. Williams

    Tony Spadaro Guest

    They hire little tiny children with very small hands.

    --
    http://www.chapelhillnoir.com
    home of The Camera-ist's Manifesto
    The Improved Links Pages are at
    http://www.chapelhillnoir.com/links/mlinks00.html
    A sample chapter from my novel "Haight-Ashbury" is at
    http://www.chapelhillnoir.com/writ/hait/hatitl.html
    "Robert E. Williams" <> wrote in message
    news:...
    > I just got to wondering how the RGB filters are applied to a CCD in a
    > Bayer pattern.
    > A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    > microns on a side.
    > That is VERRRY small. Less than half the size of a red blood cell!!
    > About 1/25 the diameter of a human hair!!
    > How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    > registration with no overlap or leakage between dots on a sensor 5.4mm x
    > 7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    >
    > Does anyone know how this is done?
    > Bob Williams
    >
    >
    Tony Spadaro, Nov 29, 2003
    #3
  4. Andy Blanchard wrote:

    > On Fri, 28 Nov 2003 16:45:31 -0800, "Robert E. Williams"
    > <> wrote:
    >
    > >I just got to wondering how the RGB filters are applied to a CCD in a
    > >Bayer pattern.
    > >A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    > >microns on a side.
    > >That is VERRRY small. Less than half the size of a red blood cell!!
    > >About 1/25 the diameter of a human hair!!
    > >How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    > >registration with no overlap or leakage between dots on a sensor 5.4mm x
    > >7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    > >
    > >Does anyone know how this is done?
    > >Bob Williams

    >
    > In terms of chip fabrication (photolithography), the size of an
    > individual CCD sensor of 3.1um is large by modern standards. To put
    > it into context, the details of individual components on a modern CPU
    > are about 0.9um across, hundreds of times less area.
    > Andy


    Actually, a 0.9 um spot is more like 12x less area than a 3.1 um spot, but
    your point is well taken.
    I wonder if they apply a SINGLE resist, shoot the RED positions, wash away
    that, then apply the RED plastic to the resulting holes. Then repeat the
    process for the GREEN positions and then for the Blue positions.
    Or do they use 3 separate masks "predrilled" with holes in the right place
    and apply consecutive treatments with RGB plastic?
    Or, do they use something like slide film with all the colors in place and
    then drop it on the CCD?
    Also, I wonder if those dyes fade with time like most dyes?
    Man! I'd love to tour a factory where this technology is used.
    Bob Williams
    Robert E. Williams, Nov 29, 2003
    #4
  5. Robert E. Williams

    Steve Young Guest

    Mr. Plastic Laminate, "Tony Spadaro" <> wrote

    > They hire little tiny children with very small hands.


    *very*, *very* useful information
    and funny as all get out
    :(
    Steve Young, Nov 29, 2003
    #5
  6. Robert E. Williams

    mmcgr Guest

    Robert E. Williams wrote:

    > I just got to wondering how the RGB filters are applied to a CCD in a
    > Bayer pattern.
    > A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    > microns on a side.
    > That is VERRRY small. Less than half the size of a red blood cell!!
    > About 1/25 the diameter of a human hair!!
    > How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    > registration with no overlap or leakage between dots on a sensor 5.4mm x
    > 7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    >
    > Does anyone know how this is done?
    > Bob Williams
    >



    Registration isn't necessarily perfect and overlap and leakage can
    occur, manifesting itself as slightly different spectral response of
    green filters in alternate rows. Some of the more sophisticated
    demosaicing arrangements take account of this.

    Mike
    mmcgr, Nov 29, 2003
    #6
  7. Robert E. Williams

    jpc Guest

    On Fri, 28 Nov 2003 16:45:31 -0800, "Robert E. Williams"
    <> wrote:

    >I just got to wondering how the RGB filters are applied to a CCD in a
    >Bayer pattern.
    >A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    >microns on a side.
    >That is VERRRY small. Less than half the size of a red blood cell!!
    >About 1/25 the diameter of a human hair!!
    >How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    >registration with no overlap or leakage between dots on a sensor 5.4mm x
    >7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    >
    >Does anyone know how this is done?
    >Bob Williams
    >


    I believe bayer patterns are created by interference filters where the
    color transmittance in determined by the thickness of the filter over
    each pixel.

    jpc
    jpc, Nov 29, 2003
    #7
  8. jpc wrote:

    > On Fri, 28 Nov 2003 16:45:31 -0800, "Robert E. Williams"
    > <> wrote:
    >
    > >I just got to wondering how the RGB filters are applied to a CCD in a
    > >Bayer pattern.
    > >A typical 4 MP 1/1.8" CCD has pixel sites that are only about 3.1
    > >microns on a side.
    > >That is VERRRY small. Less than half the size of a red blood cell!!
    > >About 1/25 the diameter of a human hair!!
    > >How do manufacturers apply 4+ million colored plastic(?) dots in perfect
    > >registration with no overlap or leakage between dots on a sensor 5.4mm x
    > >7.2mm. In printer terms, that's about 77,000 dots/inch. That is Awesome.
    > >
    > >Does anyone know how this is done?
    > >Bob Williams
    > >

    >
    > I believe bayer patterns are created by interference filters where the
    > color transmittance in determined by the thickness of the filter over
    > each pixel.
    >
    > jpc


    That sounds reasonable and a very clever way to do it.
    It gets around using dyes that could fade.
    Still, it would probably require 2-3 masks to get the right thickness of
    interference filter over the correct pixels.
    If can get my hands on a junked camera, I'll recover the sensor and examine
    it under a 200X microscope to see just how accurately the pattern is applied,
    and what it looks like.
    If any one has done this, please give us a heads up on what you found
    Bob Williams.
    Robert E. Williams, Nov 29, 2003
    #8
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Tony Whitaker

    Canon CMOS Sensors vs CCDs - A No-Brainer?

    Tony Whitaker, Sep 4, 2003, in forum: Digital Photography
    Replies:
    9
    Views:
    542
    eddie.m
    Sep 11, 2003
  2. DH

    35 mm size CCDs in Digital consumer SLRs?

    DH, Sep 9, 2003, in forum: Digital Photography
    Replies:
    12
    Views:
    472
    J. B. Dalton
    Sep 13, 2003
  3. BG250

    four color Bayer sensors?

    BG250, Nov 20, 2003, in forum: Digital Photography
    Replies:
    1
    Views:
    376
    Martin Brown
    Nov 21, 2003
  4. George Preddy

    The chances Bayer gets 1 pixel color right

    George Preddy, Feb 10, 2004, in forum: Digital Photography
    Replies:
    46
    Views:
    928
  5. Giuen
    Replies:
    0
    Views:
    591
    Giuen
    Sep 12, 2008
Loading...

Share This Page