Re: maclife.de
Re: maclife.de
- Subject: Re: maclife.de
- From: Uli Zappe <email@hidden>
- Date: Fri, 15 Aug 2008 23:53:29 +0200
Hi Eldon,
thanks a lot for your interesting comments!
Am 15.08.2008 um 22:53 schrieb Eldon:
The inCamera plug-in from Pictocolor gave me a profile with the
lowest average and max deltaE of any of the packages I tried,
Interesting.
but I was disappointed in the smoothness of the profile, which was
evident after assigning it to a Granger rainbow and also in a number
of real-world images.
Yes, this seem to be a general trade-off I also experienced with
printer profiling.
The logic behind this seem to be that if you do rigorous curve-
fitting, you'll get very good delta E values, but on the other hand,
since you have only so many discrete steps in your LUT, end up with
some unevenness in the color areas between the color patches you
fitted the curve to.
Of course, this is especially pronounced if you test against the
target you build the profile with, because basically, if only you get
these 288 colors correct, you'll already get very good delta E values,
and still it says (theoretically) nothing about the remaining almost
16.7 million colors ...
Overall, I was most pleased with the scanner profiles from Monaco
EZcolor and Argyll, with an edge to EZcolor. Both had a respectable
deltaE and were commendably smooth.
Interesting you say this. MonacoProfiler/EZColor (which deliver
identical results for all practical purposes, as do ProfileMaker/
i1Match) do indeed sport a visual characteristic similar to that of
Argyll (as far as I could tell in what little time I had for the
visual evaluation of Argyll profiles).
I take it from your list that you did not evaluate one of the Logo
apps (ProfileMaker/i1Match)? I ask because metrologically, the Monaco
apps were even closer to the Logo apps than the 1 point score
difference (4.5 - 3.5) suggests. In fact, the original calculation
would have resulted in 4.0 for the Monaco apps. However, the
perceptual rendering intent of the Monaco profiles (which was not used
for the metrological evaluation) expanded the luminosity range so much
that the images always looked a bit "too saturated" in comparison with
the Logo profiles and more importantly, tended to clip relatively
often, again in contrast to the Logo profiles - Logo seems to take a
more conservative approach when expanding luminosity in the perceptual
intent.
Since this effect was clearly visible in all test scans, I tried to
take that into account by detracting 0.5 points from the Monaco score,
thus resulting in 3.5.
Now, all I just wrote about Monaco seemed to be true for Argyll, as
well (but again, I had very little time for visual evaluation). This
means I also got 4.0 points for Argyll originally, and deducted 0.5
points to make the results comparable.
Just FYI, SilverFast, OTOH, does not expand luminosity at all in the
perceptual intent (in fact, there simply is no different perceptual
intent in these profiles), making the images always look dim, but they
never clip. So SilverFast was also deducted 0.5 points for the
opposite reason, or else it would have been on par with the Logo
profiles. The Logo profiles seem to hit the sweet spot in between.
The delta E values of the Monaco profiles and Argyll are certainly not
bad, it's just that the Logo profiles are better.
Incidentally, after searching the web and fooling around a bit, I
figured out how to convert a standard IT8 reference text file to
the .mrf format that EZcolor requires for flatbed scanning and
printer profiling. It turns out that .mrf files are just gzip
archives. Here are the steps I took:
Most interesting, thanks a lot! I googled for that as well, but was
less successful than you were.
Bye
Uli
________________________________________________________
Uli Zappe, Solmsstraße 5, D-65189 Wiesbaden, Germany
http://www.ritual.org
Fon: +49-700-ULIZAPPE
Fax: +49-700-ZAPPEFAX
________________________________________________________
_______________________________________________
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