on 10/2/01 4:30 PM, Enoch's Vision, Inc. (Cary Enoch R...) at
> Excellent objective review:
It's like most reviews, the critical tone is overly neutral so as not to
offend purveyors, and so doesn't properly inform us either. Prose like
"Despite its complexity, the Super CoolScan 4000ED is a remarkable,
eminently useful device once you figure out how best to exploit its
capabilities" is a waste of any reader's time. Terms like "pro-quality" and
"tack-sharp" just sound like they mean something. Then there's the "Not so
Cool Stuff" (at the end) about Nikon Scan 3 repeatedly crashing the host
computer. Most users would say this is a hell of a lot more serious than "a
Bruce introduces that the scanner is complex, but doesn't ever explain why.
Only through inference can we understand that he means complexity in terms
of when you engage the product's more desirable and necessary features (esp.
GEM) performance becomes mind-numbingly slow and the operator has to use
heuristics to develop an effective workflow. Bruce says "If ICE is cool, GEM
is nothing short of amazing." Blah, blah. What he doesn't say is that these
features are partly designed to overcome a liability for this product
because of how the Nikon optical subsystem (unlike many others) emphasizes
film contamination and grain.
Ditto for color management complexity. Bruce's doesn't like Nikon's "closed"
color management design. Yet he doesn't describe what features relate to
closed versus open CM design, so it's impossible to understand why the Nikon
is lacking. He mentions that the supplied profiles are apparently useless
when applied to raw scans, but otherwise doesn't consider this point worthy
of deeper consideration. Kind of like the crashing quirk. It would have been
illuminating to compare Nikon Scan with even one other scanner driver in
regards to features for ICC color management. As an aside, I don't find
using Nikon Scan 3 with an LS-2000 any more closed or cumbersome than the
designs I've seen from Umax, Hewlett Packard, Epson, Polaroid or even
Silverfast. I've found Nikon Scan a tad easier to understand than the others
I've seen. In spite of support for ICC color management, Bruce seems to
finds the color uninspiring. I found this too on the LS-2000. This matters
to me a lot, but doesn't seem to matter to Bruce much. I wish someone with
Bruce's reputation might invoke some helpful rhetoric, in the service of the
consumers, like: 'This color management professional (Kent Brockman) can't
understand how a company of Nikon's experience and reputation can produce a
scanner product (4000 ED) that's this complex and expensive, with features
invoking established color management trends and technology, yet somehow
provides output with mediocre color!' (Not to mention that crashing problem)
Bruce mentions the importance of "raw" scans for profiling but doesn't
explain what raw means for this product or how it relates to profiling
There's hardly an industry standard.
Bruce's observations about optical density and dynamic range & noise
performance are noise. If it was Joe Blow making this review, I would
overlook it, but Bruce knows better. We all should know the industry has
shown little interest and barely any restraint with resolution and density
claims. The scanner purveyors are repeating the mistakes of the Hi-Fi
industry in the 1970's when audio amplifier power specifications which were
simply a measure of the instantaneous peak power while clipping without
regard to whether the output signal bore any resemblance to the input. The
FTC finally put an end to such antics in the audio business. Computer users
and the industry press are strangely tolerant of complete product failures.
Bruce parrots the purveyors hypnotic and senseless specification of density
performance in terms of the word size of the sampling subsystem (e.g., 4.2).
Bruce is completely aware of how absurd or misleading the purveyor's claims
can be, yet rather than providing any real insights into these performance
claims, he says only "In other words, we're not accusing Nikon of
artificially inflating the 4000's dynamic range spec." Wink, wink. The
review then moves on to some uninspiring density analysis including a
side-by-side comparison with the output of an Imacon Flextight II. While the
Flextight II comparison may have been intended to show the Nikon's prowess,
it can make can make one, and only one, point: there must be some incredibly
disappointed Flextight II owners.
Bruce makes no mention of the conventional wisdom that the LS-2000, Coolscan
IV, 4000 ED and 8000 ED all show depth of field limits that manifest as poor
corner to edge sharpness. This complaint is oft repeated on newsgroups and
Flare (bloom) in high contrast images can be a problem with the LS-2000, but
Bruce makes no mention of this performance concern either.
It was a review, I'll grant it that...