Re: [Lcms-user] Using Eye One Display for colour measurements?
An ICC-based CMM for color management
Brought to you by:
mm2
From: Greg <sul...@in...> - 2004-01-23 09:20:02
|
And now I am quite sure what I have said below is wrong, or at the very least, exaggerated. I did a test where I purposely mis-calibrated my monitor (simulated by loading a vcgt which introduced a red cast, by adjusting the gamma of *only* the red channel), and then created a profile for my display in this mis-calibrated state with the Little CMS monitor profiler. This means that a colour managed application, using the monitor profile, *must* be used to to produce a neutral display. Photoshop 7.01 displayed a greyscale, assigned sRGB, perfectly. That is, the display matched a relative intent conversion to the monitor space. This means that Photoshop does honour the individual R, G, and B TRCs for on-the-fly display, and does not average them out. (the appearance of the greyscale was also very neutral - the Little CMS profiler did a wonderful job ;^) Greg. ----- Original Message ----- From: "Greg" <sul...@in...> To: "Marti Maria" <ma...@li...>; <lcm...@li...> Sent: Wednesday, January 21, 2004 10:02 PM Subject: Re: [Lcms-user] Using Eye One Display for colour measurements? > Another thing regarding the vcgt - my understanding is that Photoshop > doesn't actually use the monitor profile > fully, for real time display. It creates a "simplified" version of the > profile, which I believe involves using a single > gamma value, rather than the individual gammas of each channel. Thus, if the > vcgt is used to calibrate the > gamma's of each channel, this will result in an ICC profile which really > does have the same gamma value > for each channel, and Photoshop's simplified version of the profile will in > fact be just as accurate as the > full profile, resulting in a more accurate display than would be the case if > the vcgt were *not* used. > After saying this, however, I see that my Eye One Display profiles still > have *slightly* different > values of the gamma for each channel, despite the fact that it uses the > vcgt. > > Greg. > |