Datacolor User to User Support Group.

In a message dated 3/14/06 10:23:19 PM, lowlife_inc@... writes:


> 
> Thanks for taking your time discussing this issue. So - is it OK using
> RGB sliders for adjusting white point in my situation?
> 
This may be more than you want to hear, but the last issue has left me 
feeling like I need to cover all the bases, so that it doesn't become a political 
football:

The only two ways you can adjust the RGB balance (how the color of white is 
formed from RG&B) is with hardware adjustments (physically adjusted 
whitepoint), or with videocard LUT adjustments ,(VLUT corrected whitepoint). There is no 
such thing as a profile corrected whitepoint (though there could be a profile 
miscorrected one, by lying about what it is), so it has do be done in one of 
the first two methods. Physically corrected whitepoint (typically using front 
panel controls) is prefered for CRTs, where there are physical corrections 
possible. For LCDs, any correction you are making is always one of adjusting the 
polarization filters, be it with the VLUTs, or some other setting. But with 
high bit monitor (greater than 8 bits per channel) there is some hope that the 
front panel control adjustments of RGB Gains are not losing you levels from the 
available levels on screen, so you will not get banding and levels loss from 
them. So, for high bit LCDs (more than 256 levels per channel AT THE SCREEN, 
not just at the board inside the monitor), front panel RGB Gains adjustments are 
preferable to VLUT ones as well. Tests show that this works for some models, 
and doesn't (thus costing you levels) for others, based on how many levels at 
the screen, and how they are wired.

As for whether you should be white point adjusting an LCD at all: lower cost 
LCD screens take a real hit from whitepoint adjustment. Users may mindlessly 
choose 5000k for a screen with a native whitepoint of 6500k or higher, and lose 
  A LOT of levels on screen to the correction, not to mention making other 
adjustments more difficult. We are writing software to be used by a wide range 
of users, in nine languages, and often by people who are not using it in their 
native language. Even native English speakers, with a fair understanding of 
the topic often barely read what we write on screen, and none of what we add in 
the help. 

So we need to have basic settings and workflows that will work best for most 
users, on most monitors, without expert adjustment. For that, using native 
whitepoint is more foolproof. As is using the default settings on your LCD 
monitor, instead of trying to improve on the black luminance, for instance, that the 
engineer at the factory chose as best. These settings are easiest, most 
foolproof, and for many monitors can't really be improved on by expert adjustment.

That does not mean that users with a reasonable level of knowledge, and 
special goals, should not adjust the monitor to other settings. You may want to 
balance the luminace of two screens to match. You may want to adjust the 
whitepoint of two monitors to match. Makes perfect sense to try it, and if they are 
reasonable quality monitors, you should get good results. But if they are five 
year old LCD   Z-Mart specials, then you may find that the only way to get 
reasonable results from them is at native whitepoint, default settings, no special 
matching.

Hope I didn't beat that innocent question too far into the ground...

C. David Tobie
Product Technology Manager
ColorVision Business Unit
Datacolor Inc.
CDTobie@...
www.colorvision.com