to some recent posts by Wolf Faust>>>
.......................................................................
I don't know where Wolf Faust gets his data from but there is overwhelming
evidence from real-world, high-end scanner and digital camera users that the
HCT does a better job than the IT8. As for his claims that I'm just creating
extra patches for marketing value, he's sorely mistaken. The current HCT
design is the product of about eight year's research and experimentation,
with all available profiling packages and dozens of different scanners. In
fact I started out with a 250-patch design where I simply re-occupied the
IT8 patches with different colors. (Didn't produce enough improvement).
Interestingly, I never planned to commercialize the HCT. I created it just
for my own use and for my own consulting clients. I was forced to go public
with it through word-of-mouth testimonials and strong user-demand.
As for Faust's claims that only 170 patches are necessary to create a good
scanner profile, that's a theoretical dream. It's like saying a 300-patch
printer target can do the same job as a 1400-patch target. (I wish!)
Sure, if you already know the basic characteristics of a particular scanner
model and if that scanner applies absolutely no color transform except the
normal 1-D curve functions, you would (theoretically) only need a few
'hinting' patches (less than 170, by the way) to create a good profile. But
in reality, all scanners work differently, often with annoying little
non-linearities in unexpected places. And individual scanners within a
family are usually different enough that most modeling theories can only
make an approximate prediction, not an exact analysis, of a certain
scanner's characteristics. Multiply this by the enormous number of variables
in the typical user interface and you can see why many more than 170, 250
(or even 500, for that matter) patches may be needed to successfully decode
the performance of any particular scanner for which you do NOT already know
the basic characteristics.
An even worse case is if a scanner applies any color transformation like a
3xn matrix, a 3-D LUT or an ICC profile (which many do). In that case even
500 patches will probably be too few to successfully analyze that scanner's
RGB color space accurately.
In spite of all this, I am happy to say that I have achieved quite good
results with some IT8s on some scanners (using my 'extended profile' trick)
and, as I have always said, casual users should be quite satisfied with the
IT8. But high-end users continually tell me the HCT blows the IT8 away for
critical matching to the original. The differences they notice are usually
in deep colors and shadow tones and the reasons are pretty obvious when you
compare the IT8 to the HCT.
The bottom line is that to do the best possible job you have to start with
the best possible tools. Any way you look at it, a hand-measured 500 patch
target is quite obviously a better tool than a hand-measured 250 patch
target. Period.
I'm copying this to Dr. Chris Edge (KPG-Imation) - one of the world's most
experienced and respected color scientists. Chris tells me his tests show
the HCT clearly improves the quality of scanner profiles compared to
profiles from the IT8. That, along with many unsolicited letters of thanks
and congratulation from demanding color separators and photographers over
the last two years is good enough for me.
Regards,
Don
******************************
Don Hutcheson Hutcheson Consulting (Color Management Solutions) 11 Turnburry Rd Washington, NJ 07882 Phone: (908) 689 7403 Mobile: (908) 500 0341 Fax: (908) 689 5305 E-mail: [EMAIL PROTECTED]
******************************
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Lcms-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/lcms-user
