[argyllcms] Re: problems hitting white point (color temperature, RGB gain/contrast) while maintaining brightness: dE not stable

  • From: David Heinrich <dh003i@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 10 Dec 2009 09:02:34 -0500

On Thu, Dec 10, 2009 at 12:50 AM, Graeme Gill <graeme@xxxxxxxxxxxxx> wrote:

> David Heinrich wrote:
>
>> Why are the readings bouncing around? Troy Sobotka tells me he gets dE of
>> 0.2 on a TN panel LCD he uses and that the readings are stable. So what's
>> wrong here? Either I'm doing something wrong, or something's wrong with my
>> CRT (it looks fine), or something is wrong with the Spyder2 colorimeter
>> I'm
>> using.
>>
>
> I'd be very suspicious of a mode named "dynamic". For color profiling
> one wants the opposite of this, stable and reproducible. So my guess
> would be that you can't get consistent readings because "dynamic"
> mode is varying things dynamically.
>

Yeah, I switched it from Dynamic to "Standard" mode (can't get up to
brightness/contrast in "Professional" mode). Results were maybe a little
better, but difficult to tell. Here is the result of running the testing on
all values several times after finishing with the adjustments.

One thing I find puzzling is the percentages? Are they natural log
percentages? Take the first set of values, target brightness: ln(96) -
ln(93.32) = -2.8%. For target 50% level, it would seem to be multiplied by
2: 2 x [ln(17.08)-ln(20.64)] = -3.8%. For target near-black: ln(0.93)-ln(1)
= 0.1%.

  Target Brightness = 96.00, Current = 93.32, error = -2.8%
  Target 50% Level  = 20.64, Current = 17.08, error = -3.7%
  Target Near Black =  0.93, Current =  1.00, error =  0.1%
  Target white = x 0.3135, y 0.3236, Current = x 0.3128, y 0.3228, error =
0.48 DE
  Target black = x 0.3135, y 0.3236, Current = x 0.3199, y 0.3318, error =
4.10 DE

  Target Brightness = 96.00, Current = 93.39, error = -2.7%
  Target 50% Level  = 20.64, Current = 17.09, error = -3.7%
  Target Near Black =  0.93, Current =  1.00, error =  0.1%
  Target white = x 0.3135, y 0.3236, Current = x 0.3127, y 0.3227, error =
0.52 DE
  Target black = x 0.3135, y 0.3236, Current = x 0.3213, y 0.3326, error =
4.46 DE

  Target Brightness = 96.00, Current = 92.99, error = -3.1%
  Target 50% Level  = 20.55, Current = 17.20, error = -3.5%
  Target Near Black =  0.93, Current =  0.97, error =  0.0%
  Target white = x 0.3135, y 0.3236, Current = x 0.3135, y 0.3221, error =
1.28 DE
  Target black = x 0.3135, y 0.3236, Current = x 0.3199, y 0.3300, error =
3.39 DE

So even with the CRT in Standard Mode, doing fine-adjustments I still can't
get too great. The problem is still similar to what I quoted in the original
e-mail: when adjusting R, G, and B channel gains individually, the dE's
bounce around as does the recommendation of which values to change. Is there
a way to change how long the Spyder2 takes readings before updating? It
would seem that if there was a way to get the program to average the results
of 5-10 updates or 5-10 seconds (I'm not sure how this works: is argyll
getting a reading from spyder2 every several seconds, or is it averaging the
readings of spyder2 over several seconds?), it would smooth out this stuff
and allow for me to get a better idea as to what to do on my display.


> Displays and the electronics driving them will have unit to unit
> variations, and will change as they age. If you can't hit your targets,
> then you only have the options of changing your targets to something
> you can hit, or moving to a different display.
>

Other related posts: