Re: gamut warning failure
Re: gamut warning failure
- Subject: Re: gamut warning failure
- From: Graeme Gill <email@hidden>
- Date: Tue, 17 Oct 2006 09:42:37 +1000
Pat Herold wrote:
ColorThink does use the Device to Lab table by default to create the 3D
gamut. Most other graphing applications don't - they go Lab to
Device.
I'd be highly surprised if that were the case, since there is no straightforward
way of using the PCS to device lookup to generate a gamut surface, whereas
the device to PCS conversion can be used directly.
It is still unknown what Adobe actually uses to come up with the gamut
warning. I know Steve himself talked with Adobe concerning this very
subject and was never able to get a definitive answer from them.
There are four possible approaches to determining whether a color
is in or out of gamut that I can think of:
Using the gamut table of the ICC profile. This is direct, but
generally highly inaccurate, since the quality of the tables created
by the profiling applications often isn't great, and the fact that
it's indexed by PCS value limits the accuracy.
Using a round trip threshold. The value is round tripped using the
relative colorimetric conversions (ie. PCS->device->PCS), and some
delta E is used as a threshold. I would guess this is the most commonly
used, reasonably accurate method.
Gamut hull testing. Create a gamut hull (using similar methods to that
used for 3D plotting), and test the color to see if it's inside or
outside.
Inversion test. This is the most accurate method, but consists of
trying to invert the PCS value, and checking if it's possible to
invert it (ie. if any device value can exactly reproduce the color
according to the device to PCS conversion).
Gamut plotting and testing of CMYK (or other more than 3 colorant)
devices is complicated by a couple of things. One is that the
ink limit (TAC) affects the gamut. A good 3D plotting application
will take this into account. The other is that the gamut
surface need not be a simple hull, but generally corresponds to
a 4 dimensional hyper-cube, folded into 3D space, so any
smooth representation of the gamut surface is probably an approximation
of a much more complicated underlying object.
Graeme Gill.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden