More about monitors and gun control
More about monitors and gun control
- Subject: More about monitors and gun control
- From: "Nick Milley" <email@hidden>
- Date: Wed, 13 Jun 2001 15:24:10 -0400
In a message dated Tue, 12 Jun 2001 16:13:54 EDT email@hidden writes:
>
Hmmm... mathematically half, perceptually maybe 80% to my eye, as the gain
>
effects lighter colors that are much more readable by the eye than the
near
>
black colors most effected by bias. No doubt adjusting both would be
ideal,
>
but as colorimeters, such as the Sequel device used by LaCie, are weakest
at
>
reading the color value of tones near black, adjusting these tones is less
>
accurate as well as less important.
>
>
C. David Tobie
>
Design Cooperative
>
email@hidden
Two points of interest here:
1. adjusting black level gains is "less important"
2. colorimeters, such as the Sequel device, are weakest at measurements
near black
1. WRT black level adjustment, it is confusing how adjusting white balance
using RGB gain control can be critical but adjusting gray balance is not.
Black level gains affect luminance between (about) 3 cd/m2 to 20 cd/m2.
These luminance correspond to RGB values of (about) 64,64,64 to
128,128,128 (at 6500,2.2 and 80cd/m2 white) - a fairly important range of
grays for images. A small change in any black level adjustment results in a
measurable and visible change in gray balance. We should note that these
errors
will be "calibrated out" in most packages by adjusting the graphics card LUTs;
adjusting the digital black levels, however, gives a more linear lookup table
in the graphics card and a better calibration.
2. WRT colorimeter accuracy, it is true that colorimeters that use
light-to-frequency converters as the sensing IC, such as the Sequel Imaging
Chroma4 and the Colorvision Spyder, are prone to errors at low luminances.
Light-to-frequency sensors output a pulse train whose count is proportional
to the intensity of the light striking the IC. This means that a luminance
at full white may cause a sensor to output 1,000 pulses while a luminance
at gray may cause the sensor to output only 100 pulses. Thus, an error
of one count results in an error of .1% at white while a one count error in
gray
in a 1% error. The error at white is negligible, the error at gray is not. Now
consider
the total number of sensors in the calibrator (Chroma4 = 3, Spyder = 7) and
you can see how the
total measurement error becomes significant. To reduce this error, Sequel
has invented a unique method of measuring at low light levels.Sequel was
granted a patent on this algorithm making the Chroma4 the most accurate
device available today for measuring monitors.
I must apologize for taking so much space. I thought it might be beneficial
to discuss some of the principles and technologies that are embodied in a
monitor and their effect on its calibration.
Regards,
Nick Milley
Sequel Imaging