Re: maclife.de
Re: maclife.de
- Subject: Re: maclife.de
- From: Uli Zappe <email@hidden>
- Date: Wed, 3 Sep 2008 17:16:52 +0200
Am 03.09.2008 um 14:06 schrieb Bob Frost:
As a former scientist myself, I would point out that your research
into camer profiles seems very limited at present.
Yes and no.
The big problem is that several devices/parameters have to be combined
for a specific test scenario, and with each new device you add, the
number of scenarios you'll have to handle grows exponentially. The
primary focus of the review as a whole (not only camera profiling) was
to test profiling software and hardware, so I tried to include all
profiling packages and all camera targets on the market (I don't say I
succeeded - I missed at least one -, but I tried). Knowing the debate
about the correct way to light the targets, I also decided to test
different target lighting variants. Finally, there were the 8
different lighting conditions I measured.
As I said, these combinations alone resulted in 1000 test images that
needed to be evaluated.
When I became aware of the importance of the interaction between RAW
converters, I tried to include as many RAW convertes in my testing as
possible, but I didn't aspire to completeness (and, as Edmund pointed
out, thereby missed at least one converter I should have included,
which is RAW Converter). Rather, I looked for RAW converters that are
either explicitly ICC aware or popular. I performed preliminary tests
with these converters and the various software packages, looked which
converter worked best with which profiling software, and then sticked
to the optimal combination for each profiling software - otherwise,
the number of test images would have exploded to many thousands.
Because of the controversy about whether camera profiles are an
improvement at all, I also evaluated the default profiles of the RAW
converters, but keep in mind that this was not the primary focus of
the review. Therefore, when it came to the question of how many
cameras to test, I sticked to one since it's not very probable that
some ICC profiling software works decidedly better with one specific
camera than another (just as I also used only one film scanner, one
CMYK printer etc etc. for my respective profiling reviews, or else the
material would have become endless). It's certainly more probable that
the default camera profiles of a specific RAW converter work better
with one camera model than with another, but as I said, testing the
RAW converters wasn't the primary focus.
I did cross-check my results on an "anecdotal" basis with other
cameras and found no hint that including them would add meaningful
data to the review, so I left it at that. I still think that the 1000
images that were evaluated are a statistically solid base for the
focus of the review.
You've only tested one camera, I believe, whereas Adobe has specific
camera profiles for many, many cameras, and now has the new camera-
matching profiles for many Canon and Nikon cameras. Have you tested
the ones for your Canon?
No. I used the software Adobe provided me with, which did not include
Canon specific profiles. I just did a quick search on their site and
cannot find those profiles there, either. Can you provide me with some
pointer?
Have you tested Canon's own raw converter?
Yes, of course. That's how I was able to determine the color
deviations of Canon's own "picture styles" which were my "factory
settings" reference point for the question if (third-party) RAW
converters or individual ICC profiles could improve the color
reproduction.
Has your work with the one camera been confirmed by others to be
superior?
Somehow I don't understand this question. By other cameras? By other
testers? What is superior? My work, the camera, ... ?
Without it being translated into English (or American) few people
will be able to try to confirm it or point out its inadequacies (if
it has any).
Well, there are a bit more than only a few people on this earth who do
speak German. ;-) As I wrote at the beginning of this thread, I'd
like to put the results of this test on my website later on (in
English), but currently, I honestly cannot say if I will have the
resources to do so. This test already ate up a lot of resources as it
is.
I can see that there is a very small (?) market for 'perfect color
reproduction', but the main market is surely for color that looks
good,
I can only repeat that technically induced color deviations are in no
way identical with a "good look".
Adobe's take on the use of ICC camera profiles is spelt out on their
website FAQs
That sounds convincing, but as I already wrote, my measurement results
simply did not confirm several of the arguments that sounded
convincing in theory.
But if a few of you want to do something different, nothing is
stopping you from doing just that. You obviously have the means of
making and using your ICC camera profiles, so what's the problem?
The problem is the lack of a comfortable workflow. I'd like to use ICC
camera profiles and still be able to have Aperture loading my new
images automagically when I connect my camera to my Mac.
That's easier and therefore less error-prone in everyday life, but of
course, you can work around it. But no-one ever said this is a life-or-
death issue. This whole argument started when Edmund remarked that
it's a shame that Lightroom doesn't support this, and I agreed (and
added Aperture to the mix). That was all, and I'm still completely
baffled which kind of reaction this has produced. I would have thought
that other results I reported would have been far more interesting. So
probably there's more going on behind the scenes of this argument (cf.
below).
Your statement about your dislike of Adobe (and Microsoft) as a
Company starts to raise question marks in my mind as to your
scientific 'neutrality', and to my mind has no place in a scientific
comparison of different workflows.
In a way I'm glad you brought this to the table, as I supposed it was
prone to misunderstanding.
I completely agree that my feelings about Adobe as a company have no
place in the review (it wasn't primarily about different workflows,
but anyway), and it did, in fact, not influence any of it. I measured
Lightroom's default conversion quality in exactly the same way as
those of the other packages, and reported the numerical results. (I
hope you don't assume that I somehow manipulated the numbers.) In
fact, I expected Lightroom's results to be better and was surprised
they weren't. Also keep in mind that my criticism of the missing ICC
camera profile option applies in exactly the same way to Apple's
Aperture, whereas it does not apply to Capture One PRO. Still, the
test results of default conversion quality were in this sequence:
Apple – Capture One - Lightroom, so this sequence has nothing to do
with my liking of those companies.
However, every scientist is also a member of society, and as such,
cannot be neutral, as there is no such thing as neutrality in social
relations (staying away from any kind of judgement or action is a
social statement in itself which cannot be avoided). When Andrew
confronted me with the authority argument that something cannot be
correct because some specific group of "smart people" disagrees, this
was a social argument, not a scientific one. As such, I can only
answer with an argument on that level, i.e. that I don't accept thee
people as the authority as which they were presented to me, for
reasons I at least hinted at.
Neutrality on this level isn't possible - either you accept somebody
as an authority or you don't, and neither position is neutral. But if
no neutrality is possible, I think it's better to clearly say where
you stand than to keep that for yourself - for the benefit of
neutrality in the area of scientific questions, which, of course, must
prevail.
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