1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

A method for monitor brightness callibration

Discussion in 'Digital Image Editing & Printing' started by Andrew Ranshaw, Apr 26, 2019.

  1. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    I use a Spyder 4 pro system to calibrate my monitor and while colour balance has matched prints from good online photo printers I frequently have problems with the density not being as expected from looking at the monitor image ,so clearly the monitor calibration has produced too bright a screen. Some people online suggest setting a brightness of 120 cd/m2, but my monitor is not sophisticated enough to quantify brightness levels in such units.
    Therefore I came up with this objective method which others who have the same problem may find useful, it worked for me.

    I placed an 18% Kodak grey card on a bookshelf in the same lighting environment where I hang prints, set the camera on spot metering and with the lens de-focused made an exposure of the card using the meters recommended settings. I was careful to not allow any shadows to fall on the card as this was done in daylight and the card was facing a large window.
    I then turned the card around and made an exposure of the white side keeping the same settings.
    Next I loaded the grey card image into Lightroom cropped the central portion of the card and pressed F so it filled the monitor screen. Standing back at a good viewing distance I took a spot meter reading from the centre of the image and adjusted the monitor brightness until the camera meter indicated an exposure which was the same as that used to make the actual image . Then to check I was on the right track I loaded the white image on the monitor to see if the camera meter indicated the same exposure as that used to make the grey card image, which it did, I have to say that the white card looked a bit greyish but I trusted the objective power of the metering and went with what it was telling me, which was that the monitor was now emitting the same amount of light which the card had been reflecting when placed on the downstairs bookshelf.
    I just got some prints back from the online photo printer and the subjective result is that what looks right on the monitor screen in a darkened room also looks right in the print in bright daylight. And about bloody time too.
     
  2. PeteRob

    PeteRob Well-Known Member

    There was a similarly trick doing the rounds a while back. I forget the detail but filling the camera frame and adjusting the monitor brightness to give a specific exposure was a way to get something similar to 120 cd/m2.

    The main issues are three-fold. Many monitors are quite bright out of the box - they need to be for office work - so without calibration they can be brighter than is best for image work. The second is that the room lighting also matters. The 120 cd/m2 is for work in subdued light so that the monitor can be seen clearly. In my office it is no good me trying to edit any photographs before 15:00 if the sun is out, even with curtains drawn it is simply too bright. The third, with respect to prints, is that the light under which they are viewed vastly affects the appearance so you should judge the print in good daylight in order to compare with the monitor version.

    I haven't used my Spyder software for years but I thought it did enable some kind of brightness measurement. The software I use now, with my ancient Spyder 3, allows for both white-point and black-point setting. The latter is important if matching different monitors to ensure they both have the same contrast after calibration.
     
  3. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    My home computer lives in a dingy room with a tiny north facing window creatively described as 'bedroom' by the estate agent who sold us the house. Not much good for anything but storage and photo editing. The Spyder 4 system does include a brightness calibration but using it results in a screen which is way too bright. Following the method described above resulted in lowering the monitor brightness down to it's lowest possible setting.
     
  4. GlennH

    GlennH Well-Known Member

    The Spyder4 Pro software should be able to measure luminance in candela per square metre. In the past, it was only the Express package that didn't allow this. It's the software rather than the monitor that decides it. You can also pair a Spyder4 with DisplayCal (open-source software). Monitor calibration doesn't do anything to luminance typically unless you interactively adjust it via the OSD control using feedback from the software.
     
    Last edited: Apr 27, 2019
  5. PeteRob

    PeteRob Well-Known Member

    I think this is generally true. I swapped to Basiccolor Display 5 which can hardware calibrate* my NEC monitor - so it programs the monitor and does not use lookup tables for the graphic card output. It can set the monitor brightness. It seems also to be able to set the brightness on my imac for which it describes the calibration as part hardware part software.

    *although I am stuck with an old software version because NEC changed their API and support for monitors using the old API was dropped.
     
  6. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    My ColorMunki Display measures the brightness of the screen and guides you to set it very accurately with the normal screen controls. It certainly comes extreemly close to matching the screen and print. Though it does not have the ability to iterate the result like their top models do.
    My previous Huey Pro had no such ability at all.

    It seems surprising to me, that not all devices have the ability to measure and set screen brightness...
     
  7. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    My Spyder pro 4 does calibrate screen brightness, it asks the user to adjust brightness till a target value is met, it's just that the target value is completely wrong.
     
  8. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    Is that a fault with all spyder fours or just yours?
    It can not be a difficult task to make an accurate measurement. Any more than it is to calibrate an exposure meter accurately. However any device can develop a fault.
    But like an exposure meter it should be possible to calibrate at a different point on the scale to compensate. For a consistent error.
     
  9. Chester AP

    Chester AP Well-Known Member

    I have a shot of my wife's face taken outside on a bright overcast day using the camera body's 'daylight' white balance, and use this for colour adjustment of my PC monitor. Also, of course, the monitor is used in a dim room with no light source falling onto the screen, and the brightness and contrast are set just high enough for the image to look realistic (it helps to have an image of something you know and can easily spot if the colour balance is wrong when displayed). When I am preparing an image to be printed by a lab, I know from experience that I need to increase its brightness slightly specifically for this purpose. Also, I always turn off any 'automatic enhancement' offered by the lab because this appears to boost all the things that I have been careful to adjust.

    Adjusting a large flat screen TV screen the same way, and viewing it under the same lighting conditions, works well too. I suspect that most display screens (TV and PC monitor) leave the factory with the brightness and contrast set far too high for the viewing of images, because it is expected that users have bright lights on in the room. Lowering ambient light and then using the minimum brightness and contrast also appears to improve the perceived sharpness of the image, especially on large TV screens. Why do so many people spend all that money, adjust them so badly, and watch them in the worst possible conditions?
     
  10. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    I don't know if the error is with my Spyder 4 or all of them as I only have one of them and I have no idea what brightness level the Spyder 4 is trying to calibrate the monitor to.Perhaps it's just doing what it was programmed to do by someone who did not understand the reality of photo processing, who knows?
    It would serve no useful purpose to calibrate at a different point on the scale as you suggest because first I would need to quantify the degree of error and to do that I would use the method described in my first post which then makes the Spyder 4 brightness calibration function un-necessary and redundant so why bother using it?
     
  11. PeteRob

    PeteRob Well-Known Member

    I doubt the instrument is at fault. It wouldn't calibrate properly if it wasn't measuring correctly. I was trying to remember how the Spyder brightness setting worked. I found an on-line guide for Spyder 4 elite that rang a faint bell. After setting desired option for brightness ( e.g. 120 cd/m2) there is an iteration sequence where you measure-adust-measure screen brightness until you get within 4% of target. This will certainly use all three colour channels in the sensor. I wonder if it assumes the monitor is white - the guide says run full-cal then re-cal so it might need a near accurate profile loaded in order to measure brightness correctly. Without profiles my imac screen is blue and my NEC is brown. It is quite odd watching them snap together when the profiles load on computer start up. My NEC also has a warm-up period. It takes about 30 mins to reach full working brightness from off, and a good 5-10 minutes from stand by. Calibrating too soon after turning it on will result in too bright a screen. Having two monitors side by side with matched calibrations (as close as possible, the imac has a smaller gamut) makes it very obvious. I also calibrate to a pre-press setting D5000 for print rather than D6500 for sRGB following guidance with my software but I've not compared the two to see if it makes a difference as to how I woukd process an image for printing.
     
  12. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    So for those of you who are happy with the screen brightness levels set by whatever system you use, do you then find that once calibrated your screens are significantly dimmer than screens which are generally used for computer use such as web browsing, word processing etc?
     
  13. PeteRob

    PeteRob Well-Known Member

    Yes. I have a desktop and laptop for "work" and both are brighter, higher contrast for office duties. Neither is colour accurate nor needs to be. The laptop would be totally useless anyway. When travelling I use it to store image files off the camera. It is so sensitive to viewing angle as to be worthless as a processing platform, even if calibrated you'd never see the same thing twice.

    I do web-browse on my imac but I don't do office work on it. It is dedicated to photography.
     
  14. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    Yes indeed. Mostly calibration is done to 120 cd/m2 ad a gamma of 2.2 which is significantly far less bright than most monitors straight out of the box. But it should give a far closer match to the resulting prints to the screen. Though on the ColorMunki at least, you can choose any standard brightness and gamma that you want.
     
  15. Chester AP

    Chester AP Well-Known Member

    No, because there is no bright light falling directly onto the screen. My wife also uses the PC for word processing and emails, and has never complained it is to dim. If fact, she acknowledges that text is clearer on it (a 13 year old 20" flat panel) than on a similar sized newer screen she had to use at work that was 'adjusted' for use under bright fluorescent strip lights.
     
  16. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    Well that's interesting to learn, as the Spyder 4 brightness calibration results in a screen which seems easily as bright as anything I come across in general use whereas calibrating using the cameras light meter results in a brightness level set to it's absolute possible minimum.
    I don't think there is anything wrong with the hardware because when the system measures the ambient light falling on the screen it tells me that the room is dim but that light levels are good for photo editing which is subjectively true so it must be measuring incident light correctly at least. Any way with the gray card photo now kept in it's own folder in Light Room it's an easy matter to open it up and take a meter reading from the screen.
    |Thanks for all your interesting comments.
     
  17. GlennH

    GlennH Well-Known Member

    Luminance of 120 cd/m2 is an arbitrary figure unless you have reason to use it. The ISO 12646 standard for graphic display soft-proofing used to recommend 160 cd/m2, which would be matched to a print-viewing area. Other people use 100 cd/m2 or even less (typically those that hold the print up next to the monitor for a match). But if you use software to achieve these ultra-low luminance levels, you compress the gamut and increase the likelihood of banding. Consumer-level monitors are also said to perform better at slightly higher luminance (greater stability, more accurate color). Just points to consider.
     
  18. Terrywoodenpic

    Terrywoodenpic Well-Known Member

    Calibration hardware does not use software to turn down the brightness level or by using a lookup table. It simply guides you to how much to turn down the brightness of the monitor physical controls.
    More sophisticated software used for matching printers, paper, inks and monitors uses an iteration process that measures/compares both the print and the screen and applies the result with a LUT. But it requires the brightness of the monitor to be in the right ball park for accuracy.
     
  19. Andrew Ranshaw

    Andrew Ranshaw Well-Known Member

    It seems to me that on the principle of employing Occam's razor the method I have described must be more likely to produce the correct calibration than other systems commonly employed because not only does it require calibrating the image on a monitor directly with the subject of that image which at the the end of the day all systems have to do somehow, but it does so by equalizing two numerical values which are the ev value of the subject and the ev value of the image of the subject . How could it be more simple than that? Of course if you are happy with with the results produced by whatever system you use then there is no reason to change .
    Holding the print up to the monitor is worse than a waste of time.
     
  20. PeteRob

    PeteRob Well-Known Member

    That is mainly a matter of implementation. These days the monitors are all electronically controlled so even the physical controls on a monitor alter the internal program settings rather than make physical changes. That said, there are so many different monitors that it would not be practical for calibration software to address all of them and it would need appropriate digital connection too. Laptops usually lack independent controls and rely on software settings, hence the brightness control function keys.

    Colour management should reduce the need for iteration but there will always be a need for test prints. I agree that the matter of monitor brightness is not just getting prints too light or too dark but getting the tonal representation correct. If monitors are very much brighter than 120 cd/m2 then the contrast on screen goes bigger than 500:1 which makes it increasingly difficult to anticipate what the print, contrast say 300:1 max, will look like.
     

Share This Page