color fidelity of digital cameras and a way to test for it

Discussion in 'Digital Photography' started by michaele@ando.pair.com, Jan 18, 2006.

  1. Guest

    I am interested in how good current digital cameras are as regards
    color fidelity. I think the image examples of different cameras which
    are to be found on enthousiasts' Web sites are poorly thought out. For
    example, one might see various versions of the same landscape or a
    face. What is the point of that without the original face illuminated
    by the same light to compare the reproductions to? What are they
    thinking of?

    I propose the following experiment. Take a picture of the cover of this
    week's Time magazine in color corrected light and put the file online
    for downloading. The downloader prints the file at the default setting,
    without "correcting" the colors, on high quality glossy photo paper
    using a good quality printer in good working order. She then compares
    the print to HER copy of that issue of Time magazine, viewing the print
    and the magazine with color corrected light.

    Different folks with different digital cameras could post their images
    of the cover of the same issue of a popular magazine. If effected, this
    test would be a great help in choosing a camera.
    by demonstrating comparative color fidelity among the various models.
    Time and Newsweek often have color photos on their cover. Perhaps there
    is some popular magazine with the same cover every week and which
    features a useful variety of colors, but I can't think of any.

    Thanks in advance for your consideration of my proposal.

    Mike Eisenstadt
    Austin, Texas
     
    , Jan 18, 2006
    #1
    1. Advertising

  2. <> wrote in message
    news:...
    >I am interested in how good current digital cameras are as regards
    > color fidelity. I think the image examples of different cameras which
    > are to be found on enthousiasts' Web sites are poorly thought out. For
    > example, one might see various versions of the same landscape or a
    > face. What is the point of that without the original face illuminated
    > by the same light to compare the reproductions to? What are they
    > thinking of?
    >
    > I propose the following experiment. Take a picture of the cover of this
    > week's Time magazine in color corrected light and put the file online
    > for downloading. The downloader prints the file at the default setting,
    > without "correcting" the colors, on high quality glossy photo paper
    > using a good quality printer in good working order. She then compares
    > the print to HER copy of that issue of Time magazine, viewing the print
    > and the magazine with color corrected light.
    >
    > Different folks with different digital cameras could post their images
    > of the cover of the same issue of a popular magazine. If effected, this
    > test would be a great help in choosing a camera.
    > by demonstrating comparative color fidelity among the various models.
    > Time and Newsweek often have color photos on their cover. Perhaps there
    > is some popular magazine with the same cover every week and which
    > features a useful variety of colors, but I can't think of any.


    What makes you think that the cover of a magazine can act as a standard?
    Kodak, and others, offer standards for tests such as these (but they ain't
    cheap).

    Not trying to trash your idea, by the way, as it is an important issue.
     
    Charles Schuler, Jan 18, 2006
    #2
    1. Advertising

  3. Chip Gallo Guest

    Charles Schuler wrote:
    > <> wrote in message
    > news:...
    >
    >>I am interested in how good current digital cameras are as regards
    >>color fidelity. I think the image examples of different cameras which
    >>are to be found on enthousiasts' Web sites are poorly thought out. For
    >>example, one might see various versions of the same landscape or a
    >>face. What is the point of that without the original face illuminated
    >>by the same light to compare the reproductions to? What are they
    >>thinking of?
    >>
    >>I propose the following experiment. Take a picture of the cover of this
    >>week's Time magazine in color corrected light and put the file online
    >>for downloading. The downloader prints the file at the default setting,
    >>without "correcting" the colors, on high quality glossy photo paper
    >>using a good quality printer in good working order. She then compares
    >>the print to HER copy of that issue of Time magazine, viewing the print
    >>and the magazine with color corrected light.
    >>
    >>Different folks with different digital cameras could post their images
    >>of the cover of the same issue of a popular magazine. If effected, this
    >>test would be a great help in choosing a camera.
    >>by demonstrating comparative color fidelity among the various models.
    >>Time and Newsweek often have color photos on their cover. Perhaps there
    >>is some popular magazine with the same cover every week and which
    >>features a useful variety of colors, but I can't think of any.

    >
    >
    > What makes you think that the cover of a magazine can act as a standard?
    > Kodak, and others, offer standards for tests such as these (but they ain't
    > cheap).
    >
    > Not trying to trash your idea, by the way, as it is an important issue.
    >
    >


    Even in a camera such as the Canon 20D, color fidelity with the auto
    white balance is problematic. I have been using Rawshooter to select an
    appropriate color temperature for critical shots. This software also
    allows one to apply the same settings to multiple photos. You can
    download an eval at http://www.pixmantec.com/

    Aside from online viewing, various color printers will handle your
    camera file differently. White House Custom Color, a digital photography
    service, assists the photographer in getting consistent print output. Go
    to www.whcc.com for info. Dry Creek Photo has information on getting
    good results with various commercial print machines including the Fuji
    Frontier, used by Wal-Mart photo departments.

    I don't work for these companies but have found their information useful.

    Chip Gallo
    www.flickr.com/photos/chipgallo
     
    Chip Gallo, Jan 18, 2006
    #3
  4. Lorem Ipsum Guest

    <> wrote in message
    news:...
    >I am interested in how good current digital cameras are as regards
    > color fidelity. I think the image examples of different cameras which
    > are to be found on enthousiasts' Web sites are poorly thought out. For
    > example, one might see various versions of the same landscape or a
    > face. What is the point of that without the original face illuminated
    > by the same light to compare the reproductions to? What are they
    > thinking of?


    There are online comparisons that include standard color charts.

    > I propose the following experiment. Take a picture of the cover of this
    > week's Time magazine in color corrected light and put the file online
    > for downloading.


    The magazine is printed at many different locations all over the country at
    the same time, so it can appear different. Further, fidelity can change
    during a run. And how do you expect the novice to know if his lighting is
    color corrected, or even what color the daylight might be?
     
    Lorem Ipsum, Jan 18, 2006
    #4
  5. Scott W Guest

    wrote:
    > I am interested in how good current digital cameras are as regards
    > color fidelity. I think the image examples of different cameras which
    > are to be found on enthousiasts' Web sites are poorly thought out. For
    > example, one might see various versions of the same landscape or a
    > face. What is the point of that without the original face illuminated
    > by the same light to compare the reproductions to? What are they
    > thinking of?
    >
    > I propose the following experiment. Take a picture of the cover of this
    > week's Time magazine in color corrected light and put the file online
    > for downloading. The downloader prints the file at the default setting,
    > without "correcting" the colors, on high quality glossy photo paper
    > using a good quality printer in good working order. She then compares
    > the print to HER copy of that issue of Time magazine, viewing the print
    > and the magazine with color corrected light.
    >
    > Different folks with different digital cameras could post their images
    > of the cover of the same issue of a popular magazine. If effected, this
    > test would be a great help in choosing a camera.
    > by demonstrating comparative color fidelity among the various models.
    > Time and Newsweek often have color photos on their cover. Perhaps there
    > is some popular magazine with the same cover every week and which
    > features a useful variety of colors, but I can't think of any.


    The problem is that most cameras are way more accurate then most
    printers.

    Scott
     
    Scott W, Jan 18, 2006
    #5
  6. Guest

    Why do you need a standard? The notion is to compare the color
    fidelity of a reproduction of something to the colors of the original
    it is a reproduction of. The quality, good or bad, of the original
    photo or how it appears on the cover is besides the point.

    I recommended using the cover of a popular magazine so that anyone
    almost anywhere can procure a copy of it with the same colors as
    everyone else's copy to compare to her printout of the image taken by
    this or that camera. I assume that the covers are probably printed in
    one run on the same presses so every copy's color values should be more
    or less identical.

    Mike Eisenstadt
    Austin, Texas
    Austin, Texas
     
    , Jan 18, 2006
    #6
  7. On 18 Jan 2006 13:07:49 -0800, wrote:

    >I am interested in how good current digital cameras are as regards
    >color fidelity. I think the image examples of different cameras which
    >are to be found on enthousiasts' Web sites are poorly thought out. For
    >example, one might see various versions of the same landscape or a
    >face. What is the point of that without the original face illuminated
    >by the same light to compare the reproductions to? What are they
    >thinking of?


    Just photograph nothing but Gretagmacbeth color checkers.

    http://www.gretagmacbeth.com/index/...orchecker-charts/products_colorchecker-dc.htm

    http://www.gretagmacbeth.com/index/...colorchecker-charts/products_colorchecker.htm

    You do know what color management is and who to do it, don't you?


    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #7

  8. > The problem is that most cameras are way more accurate then most
    > printers.


    And I'll add that some recent printers, which try to "help" us with features
    such as "photo enhance" only make matters worse. I am specifically
    referring to an Epson R300, by the way. A good printer, for the price, but
    "photo enhance" sucks.
     
    Charles Schuler, Jan 18, 2006
    #8
  9. On 18 Jan 2006 14:02:36 -0800, wrote:

    >Why do you need a standard? The notion is to compare the color
    >fidelity of a reproduction of something to the colors of the original
    >it is a reproduction of. The quality, good or bad, of the original
    >photo or how it appears on the cover is besides the point.
    >
    >I recommended using the cover of a popular magazine so that anyone
    >almost anywhere can procure a copy of it with the same colors as
    >everyone else's copy to compare to her printout of the image taken by
    >this or that camera. I assume that the covers are probably printed in
    >one run on the same presses so every copy's color values should be more
    >or less identical.


    Anyone any where can buy a correct color swatch chart.

    Print colors will vary during the course of a run.


    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #9
  10. On Wed, 18 Jan 2006 17:04:31 -0500, "Charles Schuler"
    <> wrote:

    >
    >> The problem is that most cameras are way more accurate then most
    >> printers.

    >
    >And I'll add that some recent printers, which try to "help" us with features
    >such as "photo enhance" only make matters worse. I am specifically
    >referring to an Epson R300, by the way. A good printer, for the price, but
    >"photo enhance" sucks.
    >


    If you are color managing your work flow that's not an issue as you
    cut off the print management.


    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #10
  11. Guest

    I don't think correctly set up printers in good working order are all
    that different. Even if they are, significantly, the downloader will be
    keeping her printer. So she need only choose the camera model which has
    the best color fidelity to the original (the magazine issue which she
    has acquired to compare to) when printed on her printer.

    Your remark about the auto-white balance control on the Canon 20D
    suggests that you are expecting the camera to correct to any color
    temperature of light you happen to be standing it. It would be
    remarkable if it could do it right. I was proposing that the picture is
    taken in color-corrected light provided by the photographer either with
    5500 degrees Kelvin, with a strobe, or 3200 degrees Kelvin, with a
    color-corrected quartz halide. Or, failing which, one can take the
    picture in direct sunlight in mid-day.

    Consistent print output is not pertinent to my proposal although it is
    a real issue in its own right. I am merely interested in which model of
    digital camera reproduces colors with the most fidelity. When compared
    to the original. The colored side of American paper money is perhaps
    too monochromatic. A 20 Euro bill which I have at hand is slightly more
    colorful but not by much.

    Mike Eisenstadt
    Austin, Texas
     
    , Jan 18, 2006
    #11
  12. Guest

    Your remark makes no sense at all. How could you tell that the camera
    is more accurate?
    How do you propose to view the image?

    Mike Eisenstadt
    Austin, Texas
     
    , Jan 18, 2006
    #12
  13. On 18 Jan 2006 14:24:09 -0800, wrote:

    >I don't think correctly set up printers in good working order are all
    >that different. Even if they are, significantly, the downloader will be
    >keeping her printer. So she need only choose the camera model which has
    >the best color fidelity to the original (the magazine issue which she
    >has acquired to compare to) when printed on her printer.
    >
    >Your remark about the auto-white balance control on the Canon 20D
    >suggests that you are expecting the camera to correct to any color
    >temperature of light you happen to be standing it. It would be
    >remarkable if it could do it right. I was proposing that the picture is
    >taken in color-corrected light provided by the photographer either with
    >5500 degrees Kelvin, with a strobe, or 3200 degrees Kelvin, with a
    >color-corrected quartz halide. Or, failing which, one can take the
    >picture in direct sunlight in mid-day.
    >
    >Consistent print output is not pertinent to my proposal although it is
    >a real issue in its own right. I am merely interested in which model of
    >digital camera reproduces colors with the most fidelity. When compared
    >to the original. The colored side of American paper money is perhaps
    >too monochromatic. A 20 Euro bill which I have at hand is slightly more
    >colorful but not by much.


    Have you profiled your camera?
    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #13
  14. Guest

    How do you know which magazines are printed where? Most I believe are
    printed somewhere in the mid-west to reduce postage costs? Even if they
    are printed locally, which I very much doubt, they are presumably
    printed on the same presses with the same ink and with a techie present
    to maintain consistency.

    As for novices, why do we have to depend on novices? Not everone owns a
    Minolta color temperature meter but some do. All we need is one
    competent photographer for each model of camera. Perhaps some competent
    photographers own more than one SLR.

    You write that there are online comparisons that include standard color
    charts.

    That might work if one owned the color chart and the site provided
    files to be downloaded, printed and compared with the original. Do you
    know of any?

    Mike Eisenstadt
    Austin, Texas
     
    , Jan 18, 2006
    #14
  15. Scott W Guest

    wrote:
    > Your remark makes no sense at all. How could you tell that the camera
    > is more accurate?
    > How do you propose to view the image?
    >
    > Mike Eisenstadt
    > Austin, Texas

    The best way is to photograph a known color chart using a known color
    space and then check the colors in the digital image. The camera's
    job is to reduce the color down to a number using a very well
    documented color space, normally with Adobe RGB or sRGB. It is easy to
    look at the values R G B values the camera came up with and from that
    tell what the error is. This is in fact done all the time in reviews.
    The better one show you both the errors when using sRGB and Adobe RGB.
    (some cameras are know to over saturate in Adobe RGB BTW)

    The point is it is really hard to get good accuracy in a printer, and
    so if you are looking at the output from a camera by looking at a print
    you will be see more the errors in the printer then errors in the
    camera.

    Scott
     
    Scott W, Jan 18, 2006
    #15
  16. On 18 Jan 2006 14:38:54 -0800, wrote:

    >How do you know which magazines are printed where? Most I believe are
    >printed somewhere in the mid-west to reduce postage costs? Even if they
    >are printed locally, which I very much doubt, they are presumably
    >printed on the same presses with the same ink and with a techie present
    >to maintain consistency.
    >
    >As for novices, why do we have to depend on novices? Not everone owns a
    >Minolta color temperature meter but some do. All we need is one
    >competent photographer for each model of camera. Perhaps some competent
    >photographers own more than one SLR.
    >
    >You write that there are online comparisons that include standard color
    >charts.
    >
    >That might work if one owned the color chart and the site provided
    >files to be downloaded, printed and compared with the original. Do you
    >know of any?


    Mike,

    Get back us when you understand how color management works.

    You use a colorimeter to check the printed output..

    I suggest before you make a bigger fool of yourself than you have
    already done go and read:

    http://www.colorremedies.com/realworldcolor/


    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #16
  17. On 18 Jan 2006 14:40:03 -0800, "Scott W" <> wrote:

    > wrote:
    >> Your remark makes no sense at all. How could you tell that the camera
    >> is more accurate?
    >> How do you propose to view the image?
    >>
    >> Mike Eisenstadt
    >> Austin, Texas

    >The best way is to photograph a known color chart using a known color
    >space and then check the colors in the digital image. The camera's
    >job is to reduce the color down to a number using a very well
    >documented color space, normally with Adobe RGB or sRGB. It is easy to
    >look at the values R G B values the camera came up with and from that
    >tell what the error is. This is in fact done all the time in reviews.
    >The better one show you both the errors when using sRGB and Adobe RGB.
    >(some cameras are know to over saturate in Adobe RGB BTW)


    Only if you don't shot RAW. Shoot RAW and Process it in PhotoPro RGB
    color space at 16 bits.


    ****************************************************

    "The booksellers are generous liberal-minded men."

    Samuel Johnson
    "Life of Johnson" (J. Boswell), Vol. I, 1756
     
    John A. Stovall, Jan 18, 2006
    #17
  18. Guest

    What do you mean by checking the colors in the digital image? With what
    device? A computer monitor? How can the R G B values the camera came up
    with tell you with what fidelity it reproduced the colors of the
    original? The original is not digital (a face, a landscape, a magazine
    cover - only the last is everywhere available for the comparison).

    My point is that judging color fidelity of a camera requires having the
    original thing photographed in hand and comparing it to how the camera
    reproduced it. How do you propose to make visible what the camera
    captured in order to do the comparison? Even if one loses the magazine
    idea and instead procures a standard color chart, one will still need
    the output of the camera which took its picture. How could that be done
    without printing it out? I suggested a magazine cover because it is
    cheap and available everywhere, unlike pricey color charts.

    Your last remark makes no sense. How do you see the errors in the
    camera to compare with the errors when printed out? In the viewfinder?

    Mike Eisenstadt
    Austin, Texas
     
    , Jan 18, 2006
    #18
  19. Dave Cohen Guest

    "John A. Stovall" <> wrote in message
    news:...
    > On 18 Jan 2006 13:07:49 -0800, wrote:
    >
    >>I am interested in how good current digital cameras are as regards
    >>color fidelity. I think the image examples of different cameras which
    >>are to be found on enthousiasts' Web sites are poorly thought out. For
    >>example, one might see various versions of the same landscape or a
    >>face. What is the point of that without the original face illuminated
    >>by the same light to compare the reproductions to? What are they
    >>thinking of?

    >
    > Just photograph nothing but Gretagmacbeth color checkers.
    >
    > http://www.gretagmacbeth.com/index/...orchecker-charts/products_colorchecker-dc.htm
    >
    > http://www.gretagmacbeth.com/index/...colorchecker-charts/products_colorchecker.htm
    >
    > You do know what color management is and who to do it, don't you?
    >
    >
    > ****************************************************
    >
    > "The booksellers are generous liberal-minded men."
    >
    > Samuel Johnson
    > "Life of Johnson" (J. Boswell), Vol. I, 1756


    I'm easy to please. If people coming out looking like people I'm happy. I
    thought custom white balance was for situations demanding more fidelity,
    although I think you have to fill the frame for the test shot so I'm not
    sure how you handle that outdoors, the card would have to be fairly big.
    Does anyone know if you can use a smaller card if you use spot exposure?
    Dave Cohen
     
    Dave Cohen, Jan 18, 2006
    #19
  20. Scott W Guest

    wrote:
    > What do you mean by checking the colors in the digital image? With what
    > device? A computer monitor? How can the R G B values the camera came up
    > with tell you with what fidelity it reproduced the colors of the
    > original? The original is not digital (a face, a landscape, a magazine
    > cover - only the last is everywhere available for the comparison).
    >
    > My point is that judging color fidelity of a camera requires having the
    > original thing photographed in hand and comparing it to how the camera
    > reproduced it. How do you propose to make visible what the camera
    > captured in order to do the comparison? Even if one loses the magazine
    > idea and instead procures a standard color chart, one will still need
    > the output of the camera which took its picture. How could that be done
    > without printing it out? I suggested a magazine cover because it is
    > cheap and available everywhere, unlike pricey color charts.
    >
    > Your last remark makes no sense. How do you see the errors in the
    > camera to compare with the errors when printed out? In the viewfinder?

    You might want to read up a bit on color spaces.
    A camera will produce a coordinate in a color space for the colors it
    images. You can check that coordinate by looking at the R G B values
    in the image, these are numbers. If you are using a calibrated color
    chart the values for the different patches of color are documented, so
    you can look at how far off the number are. If this seems like a lot of
    work it is but that is why it is nice to have a reviewer with the
    equipment, software and time to do this test. The people who make the
    color charts also sell the software to calibrate your camera if you
    really want to get the colors right.

    The point is that viewing the output of the camera visually requires
    some kind of output device and that is going to add a lot of error to
    what you see.

    Scott
     
    Scott W, Jan 18, 2006
    #20
    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. Silverstrand

    VR-Fidelity USB Speakerphone @ ThinkComputers.org

    Silverstrand, Jan 2, 2006, in forum: Front Page News
    Replies:
    0
    Views:
    648
    Silverstrand
    Jan 2, 2006
  2. Guest

    test test test test test test test

    Guest, Jul 2, 2003, in forum: Computer Support
    Replies:
    2
    Views:
    999
    halfalifer
    Jul 2, 2003
  3. Keith Wiley

    printed color way off from screen color

    Keith Wiley, Aug 21, 2003, in forum: Digital Photography
    Replies:
    5
    Views:
    446
    Marvin Margoshes
    Aug 25, 2003
  4. nickra
    Replies:
    0
    Views:
    456
    nickra
    Jun 7, 2007
  5. Giuen
    Replies:
    0
    Views:
    1,160
    Giuen
    Sep 12, 2008
Loading...

Share This Page