[argyllcms] Amount of patches
- From: Pascal de Bruijn <pmjdebruijn@xxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Mon, 13 Apr 2009 19:14:55 +0200
Hi,
I just did an experiment while characterizing the TFT screen of my new
inexpensive laptop. I put my Huey to work, and first measured 29 patches,
then 288 and finally Argyll's default of 836 patches. After generating the
profile, and applying them to some images, the visual difference was
negligible.
I took 29 patches (for the low end) because this is what the Huey Pro
software measures.
To be honest I was baffled to see a 29 patches read, to produce pretty
much to same results as a 836 patch read. Though it didn't surprise
me that the 836 patch read wasn't any better than the 288 patch read.
Some notes though, I was using -a S -q h, and had my own shaper
smoothing patch applied, where I multiplied the XSHAPE variables
by a factor of 100.
All three profiles give excellent results, as they all compensate for the
total lack of contrast the inexpensive screen has. Which is it's biggest
problem.
When looking some things up, I noticed the standard Huey reads 26
patches, and the Huey Pro reads 29 patches. Does anybody know
what these numbers are based on?
Does anybody know how many patches the vendor software of
DataColor's Spyders read?
What about the vendor software of X-Rite's i1's? And the ColorMunki?
So my basic question would be, is there a "logical" "sensible" amount
of patches to read for creating a matrix profile?
Regards,
Pascal de Bruijn
Other related posts:
