Display Calibration Alternatives

  • 2
  • Question
  • Updated 1 week ago
I would estimate that ~50% of LR/PS user base don't own a monitor calibrator and are NOT interested in becoming techno geeks. As the new Lightroom Desktop, Web, and Mobile user database increases that percentage will increase as well.

Another alternative is to use TFT Central's database with over 400 calibration display profiles created by them or submitted by users. They also provide the calibration control settings used for creating each of the profiles and full instructions. Granted there will be differences between the same model display due to manufacturing tolerances and age of the display compared to the end user's display. However, it will get the user a lot closer to a fully calibrated display than simply using sRGB or Adobe RGB profiles. Just another suggestion for users with display profile issues.

https://www.tftcentral.co.uk/articles/icc_profiles.htm

Note: This conversation was created from a reply on: Lightroom: Bad advice about display profiles in Adobe help article.
Photo of Todd Shaner

Todd Shaner, Champion

  • 1452 Posts
  • 486 Reply Likes

Posted 2 weeks ago

  • 2
Photo of Todd Shaner

Todd Shaner, Champion

  • 1452 Posts
  • 486 Reply Likes
Worthless expect for examining say the differing color gamuts etc. Displays are HIGHLY device dependent,
TFT Central's display profiles are created using the specific model monitor to create the profile, which is "device specific." I did mention the potential issues in using these profiles, which are far fewer than using the system sRGB or Adobe RGB profiles.

Andrew, I respect that you are an "expert" in the field of color management, but to quote John R. Ellis "Let's not make the perfect be the enemy of the good."
Do you hardware calibrate and profile your display?
I calibrate my NEC PA272W monitor using NEC Spectraview with an X-Rite i1 Display Pro device. This monitor as well as many other "professional grade" displays use LED backlight stabilization circuitry to minimize White Point and Luminance drift over time.
Displays change behavior over time and need recalibration on a regular basis (once a month is reasonable and takes little time).
I was initially calibrating my monitor monthly, but started doing validation checks to measure the actual calibration drift. I now calibrate every six-months, which maintains a very low Delta-E. This is with an average daily non-standby usage of ~4 hours daily or about 720 hours between calibrations. Here's a validation done today after about one month. The irony here is that "professional" users with more expensive monitors that incorporate backlight stabilization circuitry have much less need for recalibration compared to "consumer grade" displays.

Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
First, I'm still waiting on your data about the number of users of LR who don't calibrate and profile their display. Not that it changes the fact that Adobe's recommendation mentioned above NEED to be updated. Or that if you care about color consistency visually, you profile and calibrate your display. 

2nd. The screen capture you provide does NOT show trending (drift) over time. That information window: The Color Tracking tab shows graphically how well the grayscale (colors from black to white) of the display is conforming to the currently selected Target. Ideally all points along a grayscale measured on the display should have the same white point (chromaticity) and should follow the luminance curve of the Target.

The graph shows the displays actual measured grayscale color tracking using measurements taken whilegenerating an ICC/ColorSync monitor profile of the display.

EVEN if you were spot on about calibrating every 6 months, the device is changing. Because that's the nature of displays and based on the fact NEC picks their panels from the manufacturer from the production runs, you'd be hard pressed to find a lesser quality display system that doesn't drift far more than those panels (along with all kinds of electronics like GammaComp, ColorComp (on displays that still use it) etc. 

Next"Let's not make the perfect be the enemy of the good." 
Let's properly inform the user NOT to keep the canned, RGB working space profiles assigned as the display profile AS the Adobe white paper clearly states. That's been my point from the 2nd post and it doesn't matter how many or few LR (or all Adobe) customers calibrate or don't calibrate their displays. It's utterly irresponsible to tell people to use an RGB working space for a display for anymore than a few hours at best! Let's stick to the Original Post. 

As for TFT's profiles, they are again utterly useless for defining nearly all their displays (the exception would be something like OUR SpectraViews, when the target calibrations are totally defined and match). Because it's a color reference display system. IF you and I both calibrate identically, we should get a pretty close match in profiles although there's still the graphic system that can, and does change per computer system. But otherwise, the concept of a canned display profile for use defining the profile behavior of any one user is so iffy, it's not worth considering. Might as well stick with EDID. 
Photo of Todd Shaner

Todd Shaner, Champion

  • 1452 Posts
  • 486 Reply Likes
So NEC is providing false information in the validation report? Then how would one do an actual verification of the "current state" of the monitor's calibration and why does NEC even provide this capability?

Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
It's not false; you don't understand the report. It's not trending!
Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
FYI, this is what a real trending report provides:
Photo of Todd Shaner

Todd Shaner, Champion

  • 1452 Posts
  • 486 Reply Likes
Comparing my screenshot of the validation at time of the calibration and the validation check 33 days later it appears the "trend" is to lower Delta-E for all values. So the backlight stabilization circuitry is keeping the trend at 0 or better. Below are the RGB curves using 52 steps setting, which also looks virtually identical (<1.0 Delta-E) after 33 days. So what is the benefit of recalibrating the monitor at this time?

Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
I'm sorry, Todd, you just don't seem to understand the concept of trending here, with respect to how the display differs over time and how it is reported. Despite my copy and paste from the SpectraView manual and from how a differing product actually DOES show trending. 

I'll try again but this is going pointless and off topic while you refuse to answer my questions and concentrate on the topic of Adobe's bad advise that needs fixing textually (the point of this conversation). 

Trending takes ALL the measurement data used to build a profile, not just some of it, and compares that from day 1 to measurements made on day 30, then 60, 90 etc (or any dates between calibrations you desire), and provides a trending of device DRIFT from the ALL sets of measurements. This allows the user to see the AVERAGE deltaE between all of the calibration sessions. The idea IS to give the user an idea, over many calibration sessions AND over time, how much the device drifts over that period of times. This does give the user a better idea how often they should recalibrate. But you don't have to go there, you an simply calibrate on a regular basis AND assign that number of days between calibrations to remind you it's time to recalibrate. It's a feature in the SpectraView software. 

What you think you're showing us above doesn't do that. The dE report isn't a report about trending and the manual makes this clear! Which is why I bolded "currently" above from the text directly from the manual. I guess I should have underlined that text too, you missed it apparently. It's also not using the ALL the data measured for the report again outlined in the manual you should read. 

Now to finish, I assume, suspect and speculate (full disclosure) that 41% of LR users face North when editing their images. And no, I have no data to back that silly assumption up. Nor is it important or have anything to do with the text from Adobe, about using an RGB working space assigned to the display: IT SHOULD ONLY be used temporally and then the actual profile that DEFINES that display's condition should be selected. For ALL ICC aware applications to use for previewing image data. 

You wrote: 

I was initially calibrating my monitor monthly, but started doing validation checks to measure the actual calibration drift

You're not doing that! Do you now see that fact? 
Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
BTW, if you actually get a zero dE report comparing two actual sets of measurements, you should immediately run the tests over again or suspect either user error (or misunderstanding as we see here) or a software bug. NOT going to happen with any averaged measured spectral data. 

I can run the same target though a $5000 Spectrophotometer a minutes apart, produce a dE report and it will never result in dE 0.00! Never. There's noise present in all measuring devices such as spectrophotometers and colorimeters but unless they are junk, they will be below dE 1 (and the formula is key too). 
Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
Below are the RGB curves using 52 steps setting, which also looks virtually identical (<1.0 Delta-E) after 33 days. So what is the benefit of recalibrating the monitor at this time?


FTFM ;-) Page 44 and keep in mind, this is specific to NEC SpectraView's! 

Question 1: How often do I need to calibrate my monitor?
Answer: The answer will depend on how critical a color accuracy is necessary, how long the monitor is used per day, environmental conditions and the characteristics of each individual monitor. Some users working in highly critical color applications may calibrate their displays every couple of days. Others may only need to do it once a month. The recommended and default re-calibration period is every 2 weeks. The Validate Calibration feature can always be used to verify the display’s white point and Intensity and determine if the monitor needs to be re-calibrated. Always remember to allow the monitor sufficient time to warm up and reach optimum color performancebefore performing any color critical work or calibrating the display.


Of course, nowhere in the manual does it state it can do what you suggest the from the Color Tracking tab. 

Photo of Brian Pierce

Brian Pierce

  • 14 Posts
  • 7 Reply Likes
Calibration is also dependent on environmental lighting so even if manufacturers were to provide accurate data and build within tolerances  anything other than a 'proper calibration device' is not going to provide inaccurate results.  Nice idea but its a non-starter.

Photo of Andrew Rodney

Andrew Rodney

  • 781 Posts
  • 152 Reply Likes
Indeed, Brain, good point. 
A useful Adobe White Paper on that subject:
http://lumita.com/site_media/work/whitepapers/files/calibrating_digital_darkroom.pdf