Re: The DeltaE 2000 color difference formula [was: Evaluation of Color Difference Formulas]
Re: The DeltaE 2000 color difference formula [was: Evaluation of Color Difference Formulas]
- Subject: Re: The DeltaE 2000 color difference formula [was: Evaluation of Color Difference Formulas]
- From: Marco Ugolini <email@hidden>
- Date: Fri, 20 Jul 2007 16:23:22 -0700
- Thread-topic: The DeltaE 2000 color difference formula [was: Evaluation of Color Difference Formulas]
In a message dated 7/20/07 7:04 AM, dpascale wrote:
>>> This is the idea!
>>
>> You mean, why I described is more or less the theory, but the practice is
>> an altogether different thing?
>
> No, this was simply a nice description of it!
Ah -- duh! -- thank you, then.
>> To those of us with "average" color vision, a color difference is
>> perceptible before we are able to describe it, and it's perceived as a
>> *composite* effect. Its component parts are not easily discernible or
>> described. Sometimes hue and saturation differences are extremely
>> difficult to separate in their effects. It takes training and study
>> for that.
>
> Agreed again, this is why we should transfer some (not all!) of this
> responsibilty in the "measured numbers" themselves, which are based on
> equations derived from very carefully measured experiments.
Yes, the numbers should function like a reliable "warning system" that we
then verify against our own direct visual assessment.
> I wanted to be polite since many "non-customers" reading these threads are
> more susceptible than you are ;-)
Meaning they're easily confused, or swayed, or uncertain even of what they
are actually seeing?
> The important thing is that we should all acknowledge that we do not know
> everything in color, because:
> i- it is a large science field, which mix many specialties. The scientific
> expert will rarely have the knowledge require to properly calibrate a
> real-life press, and the press expert will often be challenged to explain
> why there are different parameters for the CMC or CIE94 formulas.
The twain rarely meet. It's a small miracle that we on the "practical" side
of color management have finally been able to open enough of a tiny crack in
the thick wall of often mind-numbing and endlessly confusing scientific
jargon that often separates us from achieving our results.
> ii- it is a relatively immature science, which is being improved/challenged
> on a daily basis. What was true yesterday may not hold for tomorrow, but may
> still be enough for many purposes. CIECAM is an evolution of CIELAB; CIECAM
> is being improved; CIELAB is still used (and useful); these are moving
> sands!
Yes, the science is still in its infancy, as far as its practical
applications are concerned. For example, we still don't have a truly
reliable color difference equation based on a uniform color space whose
structure closely mimics the actual color relationships we perceive with our
eyes.
And color appearance models still remain very new as well, no matter how
much Microsoft goes around bragging about their use of a CIECAM-like model
in the color management capabilities of the Vista OS.
> iii- Numbers are one part of the end result.While they may be sufficient in
> many fields, think QC for a manufacturing plant, they cannot characterize
> everything we do in color. For example, you may well have a G7 approved
> supplier but the output can still look bad for your customer. You still need
> that "artisan" touch (and I am not talking of the display here!), which
> comes with trial and error, and training, i.e. experience.
That is of concern to me, because I do have a background that makes me
receptive to both the technical and esthetic sides of the color matching
watchamacallit. I often find that prepress technicians have blinders on, and
I can barely hold a gasp when I see the results of what some of them judge
to be a "good" image.
Marco Ugolini
_______________________________________________
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