[argyllcms] Re: dispcal - blackpoint still corrected despite -k 0.0

  • From: Gerhard Fuernkranz <nospam456@xxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 27 Oct 2010 17:52:33 +0200

Am 27.10.2010 15:21, schrieb Anders Torger:
> So if I display an image with RGB color 0,0,0 (black) I would expect it to be 
> sent to the screen as 0,0,0 and not some non-zero calibrated value,...

Assume for instance a display where RGB numbers of 0, 1, 2, 3, ..., 6, 7
all result in a black output on the screen (within measurement
tolerances) and only numbers >= 8 result in an output brighter than
black. Then it does not matter whether you drive the display with 0 or 3
or 7 - the result you see will always be the same, namely black. In this
case, using the RGB value 7 for black, and not 0, has however the
advantage of a relatively flat gradient (only a difference of 1) from
the RGB value used for black to the first non-black value, while using 0
for black would imply a steep gradient (i.e. a jump from 0 to 8) from
black to the first non-black value.

Now let's look what for instance may happen if the max. RGB number which
gives a black output drift a little bit downwards after the calibrations
has been established (say from 7 to 4). Then the big step from 0 to 8
may suddendly become visible as discontinuity/banding, because RGB=8
would then become brighter, due to this drift, while the output for 0
won't change. However, if value 7 is used for black then the same kind
of drift would certainly have the effect that "black" is now displayed a
little bit brighter, but the step from 7 to 8 (i.e. from "black" to the
first non-black) will likely still result just in a small visible
difference, avoiding the discontinuity/banding near black.


Other related posts: