Re: Media Testing for maclife.de
Re: Media Testing for maclife.de
- Subject: Re: Media Testing for maclife.de
- From: Uli Zappe <email@hidden>
- Date: Thu, 11 Sep 2008 08:02:14 +0200
Am 11.09.2008 um 02:55 schrieb Chris Cox:
Let me point out that Capture One PRO (which I used for my ICC
profiles) performs the white balance adjustment *before* the ICC
profile assignment.
Ah, so there was another process involved.
And you are not using the profiles for calibration at all, but only
for artistic intent and post-process color correction.
Profiles are never used for calibration, but for profiling.
Calibration is the process you'll have to perform before profiling, be
it a monitor, printer, scanner or - a camera. For the camera, this
process is or at least includes setting the white balance.
It's clear from this processing sequence that the ICC profile has
no way to compensate for peculiarities of the spectral
response of the camera sensor, since after the white balance
adjustment, these have "moved".
Which means you are not profiling the camera at all, but rather
profiling the RAW conversion software and making corrections/
adjustments to the processed RAW output instead of the camera RAW
data.
So is *this* what in fact we're arguing about all the time?!?
The rose, by any other name, would smell as sweet; if you're happy if
we call the process "RAW conversion software profiling", then be it;
as long as it improves the quality of color reproduction significantly
- which it does -, I don't care. The manufacturers of all the software
packages I tested unequivocally call this very process "camera
profiling", though.
However, I would be interested to learn how a process that you would
be willing to call "camera profiling" could possibly work. All the
profiling packages that I'm aware of need the target image in TIFF
format. It is my understanding that to generate a TIFF from the RAW
data the rendering software *must* assume *some* value for the white
balance, the only question being which one. So are you saying that in
effect none of the camera profiling packages on the market can
actually do what it claims to do?
You improved the rendering of the RAW package to match whatever you
expected. Yes, ICC profiles can be used for color correction, and
yes, if the processing is decent they will do a good job of matching
your target. Unfortunately, that is not what you set out to test,
nor what you claimed to test.
??? I set out to test and claimed to test camera profiling packages. I
did follow the instructions in their manuals meticulously, which
included setting the white balance of the target image before profiling.
It is true that my ICC profile, built with a target shot at 6500 K,
provides better results for an image shot in sunlight than an image
shot in halogen lighting. *But* even for an image shot in halogen
lighting, the results for an image with the ICC profile are still
better than those for an image with the best Adobe profile I could
get.
See above. You aren't comparing like quantities or processes.
I compare different approaches to achieve correct color reproduction.
This is completely legitimate and mirrors exactly the viewpoint of the
user who wants accurate color, but does not care about the specific
implementation used for achieving this goal, as long as it does
achieve it. The beauty of a metrological approach is that the
parameters to check are clearly defined. For what it's worth, I could
metrologically compare the usage of a profile built with Adobe's DNG
Profile Editor and a ColorChecker 24 to singing "Hare Krishna" in
front of my camera three times at midnight. I haven't tested this, but
my guess is that Adobe would win.
But even this isn't the whole story. To explore this further, I
built two profiles with the DNG Profile Editor and a ColorChecker
24. I just used the target images with the DNG Profile Editor and
made no additional adjustments. I built one single-matrix profile,
using only one target image shot at 6500 K, and one dual-matrix
profile, using two target images shot at 6500 K and 2850 K. I then
tested the two profiles for images shot under various lighting
conditions.
Obviously, you would assume that the dual-matrix profile fared
better on average; however, surprisingly, the opposite was the case
(though not by much).
Again, you are not comparing like quantities.
Oops?!? Even an Adobe single-matrix and an Adobe dual-matrix profile
aren't comparable? You seem to live in an extremely segregated world.
So I can only conclude that while the idea of dual-matrix profiles
sounds nice in theory, it does not really work in practice, at
least not in its current implementation.
No, you cannot conclude that.
How so? Adobe explains that dual-matrix profiles are necessary because
they provide better results when applied to images shot under varying
lighting conditions.
My measurements show that single-matrix Adobe profiles deliver better
results (only slightly, but still).
I cannot see what hinders me logically to conclude that Adobe's
assumption is wrong.
In fact, you can't conclude much from your testing.
I can conclude which software to buy to achieve the best results (no
matter which technology exactly is used, but without the need for
manual tweaking). Which is is all I wanted to find out.
I used an ICC camera profile built by ProfileMaker.
But your process was flawed, your understanding of the processing
flawed,
So, since my understanding of the process stems from reading the
manuals of the software manufacturers, you are saying that e.g. X-
Rite's understanding of the process of using ProfileMaker is flawed?
and your results nearly useless.
They were useful for me in that I finally got the color quality out of
my camera that I wanted.
I don't know what this could possibly be, but wait for my upcoming
report where I will try to describe my procedure in exact detail
I would withhold that report until you correct your testing
procedure or account for all the processing stages.
How can I account for all the stages if I don't publish the report?
So while other members of this list asked me for an English
translation of my findings to be able to discuss them, you are asking
me to stay away from it? You know, it's not that this is so much fun
that I couldn't do without it. Sleep, for instance, is a tempting
alternative.
Why is it so hard for you to stay away from this kind of ad-hominem
arguments?
Because you are producing something with mistakes and
misunderstandings. And it is not an ad-hominem attack, just trying
to explain the seriousness of the situation to you.
What seriousness? Is disagreeing with Adobe dangerous for my health?
You aren't anywhere near the standards for a scientific publication.
Chris, sorry, but if this isn't an ad-hominem argument, then I don't
know what is.
We have a factual disagreement, yes. But what you're saying here
implies that you feel you are so much superior to me that not only is
it trivial that you are right and I am wrong, worse, I don't even have
the means of introspection that would be necessary to realize my
failure.
This might well be idiomatic English, but I still find it a telltale
detail that you didn't even write "*Your review* isn't anywhere near
the standards for a scientific publication" but rather "*You* aren't
anywhere near the standards for a scientific publication." The ad-
hominem is even in your words, and psychoanalysis tells us that this
rarely happens by chance.
Look, you know nothing about me, about my profession, about my
academic background. How can you say what you are saying here? Do you
really believe the Adobe perspective on the world is the intellectual
gold standard that cannot possibly be surpassed?
I really wonder what you're up to. Are you even aware that your words
from above effectively aim at destroying me as a public person? Is
this the seriousness you talked about? Do you want to prevent me from
providing the English translation?
The sad thing is, you almost managed to achieve the latter on a purely
factual basis. I hoped to be able to get this thing out today or
tomorrow, but now it seems more and more it will take weeks, simply
because I'm running out of time now, but cannot work on the
translation the next 2-3 weeks. :-(
Why is it unreasonable to assume that Adobe delivers default
settings as good as possible for them?
Because you don't know what the design goal was for those profiles.
True, but the logically only alternative to the intent to deliver
optimal profiles is to intentionally deliver suboptimal profiles. I
don't say such a behavior is impossible, but users would certainly
want to stay away from the products of such a company.
No, this does not mean that I think you are stupid -- just that you
need to do a bit more study before conducting such tests and trying
to publish conclusions.
If the amount of time I spent with studying this subject wasn't
sufficient, I *would be* stupid.
It would also help you greatly if, in the future, you asked
questions instead of making accusations.
??? What accusation?
P.S. my ipod contents include 3 Ships and Olias of Sunhillow.
So maybe we should restrict further discussions to alt.music.yes? :-|
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