How accurate are the colors on the monitor for a DSLR?

Discussion in 'Digital SLR' started by Steve, Feb 24, 2006.

  1. Steve

    Steve Guest

    For example, the D200 has a nice big ,bright monitor. Are the colors
    accurate? Or do I have to view the picture (maybe with a Gretag Macbeth
    colorchecker card) on a calibrated monitor to see the true colors?

    To get even more complicated, I notice that the same picture opened by
    Irfanview v3.98 and Photoshop 7.0 look (slightly) different. Maybe this is a
    subject for a different newsgroup.


    Thanks for all the helpful answers, past and future,


    Steve
     
    Steve, Feb 24, 2006
    #1
    1. Advertisements

  2. Today Steve commented courteously on the subject at hand
    What is your definitions of "accurate"? Lab-correct
    wavelengths of the 16.7 million colors possible, how your
    camera, multiple apps, your PC's video card and monitor
    settings? Yes, they will /always/ be different, and they will
    vary according to subject, lighting, brightness/contrast,
    photometric settings in the camera, exposure variations, not
    to mention lighting brightness and color temperature in the
    room(s) or outside as you view these. And, are you using any
    color calibration profiles anywhere in your workflow? e.g., a
    color profile in PS7 but not Irfanview? And, if you stray into
    the tall weeds of printing, all bets are off again, as what is
    "accurate" and "right" now is complicated by the concept of
    reflected vs. transmitted light.

    In the final analysis, though, what is "accurate" or "right"
    is a rather personal thing. I've long thought there are at
    least 3 defs: 1) what is technically correct including color
    temperature, 2) what looks "right" to you, and, most
    important, 3) simply what you prefer. 1) has one and only one
    answer, but 2) and 3) are infinite and variable from picture
    to picture

    So, please help us by telling us more of what you're trying to
    do, the subject(s) you shoot, whether daylight or indoors
    available light or with flash, how you're camera is set-up,
    etc., and whether you view yourself as a creative, artistic,
    documentary, or some other kind of photographer. All of those
    things, and many more, drastically affect any human's
    perception of "accuracy" and "right".

    Then, there's "precision", which most folks mistakenly think
    is a synonym for "accuracy", but is not.
     
    All Things Mopar, Feb 24, 2006
    #2
    1. Advertisements

  3. Steve

    Paul Furman Guest

    I don't know about color but brightness & contrast is generally suspect
    on most LCDs because the angle of view effects that. Brightness &
    contrast are also dependant on camera settings and irrelevant when
    shooting raw.. and the LCD & histograms don't show raw, just the
    processed jpeg. I suppose cranking down contrast in the camera settings
    would be the more realistic way of working even if shooting raw but if
    you do raw plus jpeg that spoils the effect of applying 'normal' contrast.
    Probably you are shooting in AdobeRGB and Irfanview is only able to
    assume sRGB. Irfan is great for quick & dirty browsing though.
     
    Paul Furman, Feb 24, 2006
    #3
  4. Steve

    babalooixnay Guest

    I'm assuming the D200 is like earlier Nikons and you are seeing a JPEG
    preview in your LCD. What you see on the LCD will be the result of
    your JPEG settings inside the camera. What you see on your monitor
    (assuming that your camera and monitor are using the same settings,
    colorspace and your monitor is calibrated should be the same for JPEGs
    only. Your camera will probably have at least sRGB and Adobe RGB 1998
    settings, you'll have to check your monitor for it's settings) Raw
    files will be different on your monitor as no settings have been
    applied. I'm fairly new to digital and spent the past few months
    dealing with color issues. I did end up with a Spyder 2 to calibrate
    my monitor and have sorted out the colorspace isssues. I use Adobe RGB
    1998 from camera through software. The colorspace issue will come up
    again when sending things to be printed.

    There are good tutorials around the internet. Try

    http://luminous-landscape.com/

    and

    http://www.earthboundlight.com/

    and

    http://www.cambridgeincolour.com/tutorials.htm

    All three of these websites are essentially chronicles of professional
    photographers who've made the transition from film to digital and have
    documented that transition with tips and tutorials in well indexed
    archives.

    Good luck!
     
    babalooixnay, Feb 24, 2006
    #4
  5. Steve

    Steve Guest


    That's what I was looking for, thanks!


    I'm off to see the wizard.
     
    Steve, Feb 24, 2006
    #5
  6. Steve

    Steve Guest

    ----- Original Message -----
    From: All Things Mopar
    Newsgroups: rec.photo.digital.slr-systems
    Sent: Friday, February 24, 2006 12:17 PM
    Subject: Re: How accurate are the colors on the monitor for a DSLR?




    Wow...maybe I was a little vague. I'm just getting started printing
    photos. So far I'm printing scanned slides (Acadia National Park is my home)
    and I've been lucky in that by just doing some adjustments to my monitor
    what I see there usually comes out of the printer, and it is very close to
    the original slide. No spyder, no calibration software, no verb:) Monitor,
    printer, scanner, and printer(little ole Epson 925) line up pretty well.

    So after reading your post and focusing my thoughts, I guess what I'm
    wondering is will my DSLR monitor be useful in judging if I got the white
    balance ,etc. right in the shot, or do I need to haul the flash card back to
    my desk top to have a look see? I haven't heard of any way to bring the DSLR
    monitor into the workflow.

    Steve
     
    Steve, Feb 24, 2006
    #6
  7. Steve

    Steve Guest

    I didn't realize that the LCD only displayed a JPEG, huh! It seems like the
    LCD, along with the histogram will tell you if the exposure is close, but
    you have to get to your work station to see if the shot has the right white
    balance, etc. and skin tones aren't too blue or yellow.
    Those are both true.


    Steve
     
    Steve, Feb 24, 2006
    #7
  8. Steve

    Paul Furman Guest

    I can tell if the auto WB went weird on the camera LCD but contrast is
    hard to judge depending on viewing conditions. But if you are concerned,
    definitely shoot raw, it makes adjusting WB a breeze. Much easier than
    tweaking a jpeg in PS!
     
    Paul Furman, Feb 24, 2006
    #8
  9. Steve

    Paul Furman Guest

    But at least the D200 shows separate color channels in the histogram (I
    think).
    You can tell pretty well on the LCD. Not precisely but decent.
     
    Paul Furman, Feb 24, 2006
    #9
  10. Steve

    Steve Guest


    It does. If all three histograms aren't clipped doesn't that mean the
    exposure is correct? Is there any other info you can get from the
    histograms?

    I knew there would be a steep learning curve when I bought this camera,
    no matter how much I've read already. It still has been a little
    overwhelming, trying to get decent prints. I miss my slides and scanner
    sometimes.
     
    Steve, Feb 24, 2006
    #10
  11. Steve

    Paul Furman Guest

    Yes but it's not the raw histogram so a little clipping is usually OK.
    Turning down contrast & saturation should make it closer. Of course
    sometimes you might want to clip highlights to get more shadow detail.
     
    Paul Furman, Feb 24, 2006
    #11
  12. Irfanview isn't color managed, so it can't use a monitor profile either.
     
    Ed Ruf (REPLY to E-MAIL IN SIG!), Feb 24, 2006
    #12
  13. Steve

    babalooixnay Guest

    Check your D200 manual carefully for what shows up in your LCD. I use
    a D50 and there is very little info in the manual. I had to sort out
    myself what is showing on the LCD. A few weeks ago I realized how far
    the JPEG settings will go in camera.

    I was shooting moonlit scenes using tables for my exposure as my
    lightmeter was somewhere else. It was real cold out and I was
    distracted and went into the tables wrong. Things didn't feel right as
    I had done similar shots before and the times, in the minutes, seemed
    off. After the first shot I looked at the LCD and the exposure was
    roughly correct so didn't pay it any more mind and continued on using
    the tables incorrectly. When I downloaded everything later I was three
    stops off on everything and quickly found the error I made in the
    tables. A quick look at the histogram probably would have showed me
    that the LCD was showing me a JPEG of the raw file bumped up three
    stops. It made a viewable picture in the LCD but there was nothing
    really there in the raw file to work with.
     
    babalooixnay, Feb 24, 2006
    #13
  14. First don't worry much about the camera display but calibration will
    help.

    But you must calibrate your monitor or you are wasting your time if
    you want get the best out of your photos.

    Also each application must be able to support your calibrated monitor
    profile.

    You need to start learning about color management in your digital work
    flow. Get Fraser's book on the subject and start there.


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

    "A combat photographer should be able to make you see the
    color of blood in black and white"


    David Douglas Duncan
    Speaking on why in Vietnam
    he worked only in black and white
    http://www.hrc.utexas.edu/exhibitions/online/ddd/
     
    John A. Stovall, Feb 24, 2006
    #14
  15. Steve

    JPS Guest

    In message <dcILf.13129$>,
    Some cameras do, but even that can still be a bit deceiving, if you're
    looking for optimal RAW capture. RAW conversion has to change hue,
    saturation, and brightness based on RAW hue and saturation. The red,
    green, and blue filters in the camera do not correspond to red, blue,
    and green in the displays we use. Red is often orange in the RAW data,
    blue is purple, etc., even after white-balancing it. Real-world
    saturated reds are boosted in luminance in the output, etc,

    One case in which taking a JPEG-based RGB histogram literally over a
    luminance histogram can result in an even worse exposure decision is
    with some red flowers; they clip the red channel in the JPEG, while the
    luminance histogram looks OK. The red channel in the RAW data can still
    be 1 to 2 stops away from clipping, but an RGB histogram of the JPEG
    would show the red as clipped, suggesting taking the picture again at a
    lower exposure, making the entire image under-exposed.

    What RAW shooters really need is a RAW RGB histogram, unpolluted by
    white balance, color profiles, or any other JPEG settings. Either the
    manufacturers are too ignorant (or too stratified) to understand the
    value of this, or they simply don't care, as it has very little market
    influence. All they'd have to do is sample a pseudo-random subsample
    (so as not to favor any geometric bias) of the raw pixels of each color,
    to create a histogram without leaving all of the RAW data in memory.
    There are many things that the manufacturers could do to make using
    their cameras better, that cost almost nothing to implement, but they
    just don't seem to care or understand.
    --
     
    JPS, Feb 24, 2006
    #15
  16. Steve

    JPS Guest

    In message <>,
    Well, you should be able to emulate the LCD somehow on the monitor, but
    the thing is, your brain sees a small LCD, sees that it has no real
    black; sees that the contrast is very low, and expects very little from
    it (and the lack of black hides noise). On the monitor, your
    expectations are higher, and using the Curves that would simulate the
    LCD would be unacceptable to you.

    To do the simulation, you might do something like raise the output
    blackpoint in the PS "Levels" tool; and increase the gamma value (move
    the middle triangle on the input line to the left).
    --
     
    JPS, Feb 24, 2006
    #16
  17. SNIP
    Either that, or a per channel clipping indicator.

    The indicators could be as simple as flashing dots, and could be
    triggered by a simple comparison between the bottom and the top part
    of the highest histogram bin. Given the small LCD histograms there is
    only a limited number of bins, so dividing the highest one (e.g. in
    two equal halfs) wouldn't cost too much.

    Based on that, more accurate Auto-ISO and clipping calculations could
    easily be implemented, with an accuracy depending on available
    processing power.

    I'll have to search existing patents, but otherwise I'd claim it as
    public domain prior art so it can be freely available for all.

    Bart
     
    Bart van der Wolf, Feb 25, 2006
    #17
  18. Steve

    Paul Furman Guest

    Wow, three stops, that seems too much. My guess is the illusion of
    viewing the LCD at night with your eyes adjusted for darkness could
    explain that?
     
    Paul Furman, Feb 25, 2006
    #18
  19. Steve

    babalooixnay Guest

    Surprised the hell out of me, too. I've always wondered how the in
    camera settings work, I read stuff about pre-programmed shots built in
    to the firmware but I suspect it's a simple working with the histogram.
    In my case I think it bumped up what exposure was there and spread out
    what it could find. Back on the monitor it was posterized to beat the
    band but in the LCD it showed a reasonable picture. You're probably
    right about part of it being an illusion, a night shot, long exposure
    and wondering what will show up, eyes atuned to darkness. I've also
    bracketed +2 and -2 with that camera playing with HDR stuff and had all
    three show exactly the same brightness in the cameras monitor.
     
    babalooixnay, Feb 25, 2006
    #19
  20. Steve

    Alan Browne Guest

    I suggest that you use the following "meterless" technique and not ever
    rely on your monitor for low light shooting.

    Cheers,
    Alan

    http://www.fredparker.com/ultexp1.htm
    http://www.fredparker.com/ultexp1.htm#Light Intensity Chart
    http://www.fredparker.com/ultexp1.htm#EXPOSURE FACTOR RELATIONSHIP CHART B
     
    Alan Browne, Feb 25, 2006
    #20
    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.