[argyllcms] Re: FWA overcompensation bug in spec2cie
- From: Graeme Gill <graeme@xxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 15 Jul 2010 23:30:31 +1000
Alexey Gribunin wrote:
Yes, I know. I'm trying to do it to get Lab values from UV-cut simulated
spectrals, but with standard D50 illuminant used for tri-stimulus
calculations.
Well the FWA isn't designed to exactly simulate a UV cut instrument,
it's designed to simulate how something will appear in real life,
are far more useful function in my opinion.
I guess the approach to doing that would be to use FWA to
simulate an A type spectrum with the UV taken out of it,
and the multiply by D50. But really the reflectance values
a UV cut instrument returns for wavelengths it is filtering
from it's light source are rather dubious anyway.
[ie. you are really noting the difference between the way
a UV cut instrument really works and how it purports to work.
It purports to return a measurement "under D50" with UV cut,
when in fact it returns a reflectance measured under A with
UV cut multiplied by a D50 illuminant. In contrast, Argyll's
FWA compensated D50_0.0 measurement simulates the appearance
of being illuminated by D50 with a UV filter.]
But I thought that Lab values in Test_UVCut.ti3 should be calculated
with the standard D50 illuminant, that's why I decided that it's a bug.
If it's not a bug then it's necessary to mention in manual. Because as I
know I'm not alone who found this behavior little bit strange.
It's only strange if you were expecting it to emulate a traditional
instruments UV cut measurement. That's not the claimed purpose of FWA
compensation, which instead aims to overcome the limitations of an
instrument not actually measuring under the illuminant that the
XYZ values are computed from. So I don't see any issues with
the documentation - FWA does what it is specified to do, compute
the appearance under a given illuminant more accurately than
multiplying reflectance values by the illuminant spectrum.
Graeme Gill.
Other related posts:
