[argyllcms] Re: Accuracy of ccss versus ccmx

  • From: Ver Greeneyes <vergreeneyes1@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sun, 13 Dec 2015 16:15:29 +0100

Thanks for the replies, both of you! I'm not really in a situation where
ccss files would have a speed advantage, so I'll use ccmx files instead.
Off-topic from my original question, but it does feel odd to have such an
expensive device as an i1Pro 2 and only use it to tweak a less expensive
device. I've used my i1Pro 2 on a monitor that has a pretty awful black
level (so I don't expect the light sensitivity to be an issue), but I
noticed that it's also a lot slower. Would you expect the i1Pro 2 to be
more accurate than the i1Display Pro, even when the latter has been tweaked
using a ccmx file produced with the former?


Emanuel Hoogeveen

On Thu, Dec 10, 2015 at 6:14 PM, Florian Höch <lists+argyllcms@xxxxxxxxx>

Am 10.12.2015 um 17:48 schrieb Ben Goren:
Rather than give you an answer for that exact question, let me cut to
the chase: if you’re after the best quality results, use a ccmx file
specific to your actual hardware.

The advantage of the ccss file is flexibility. And, in environments
where you have multiple workstations with multiple instruments
instruments that you move around and the like, it’s the most
manageable solution and therefore the most accurate. But if you have
the luxury of having both a spectrometer and a colorimeter and the
time to use them both for each display (as I myself do — and I even
just happen to be in the process of profiling displays this morning),
then go the ccmx route.



@OP, as a side-note, afaik technically the way a CCSS is used by Argyll
internally also results in a matrix being created (on-the-fly), but with
a different method than in the CCMX case (the CCMX matrix is calculated
from measured XYZ of both spectro and colorimeter, the CCSS on-the-fly
matrix is created using a least squares method involving the measured
spectral data, also depending on the number of samples measured, of the
spectrometer and the sensor data of the colorimeter stored in supported
meters like the i1D3. Graeme will surely correct me if I'm wrong :-)


Other related posts: