...There is much more to sharpening than trying to establish
"optimal" Thresholds, radius, strength, etc., which Nik's does with
a good deal of effort and attention to detail.
What is really happening here is that Nik's approach to the matter
is rather a forward-looking one. In other words, whenever you jump
into Nik's to perform the sharpening (and final) steps of your
workflow, it does look towards the final output "goal" but... it
does forget much of the past or prior work done during your
workflow, including the attributes of the actual capture (Subject,
ISO speed), etc.
On the other hand, FLEXSharp offers a different value proposition:
let's have a true, "back-to-back" look of the entire workflow,
giving us the opportunity to take into account everything we have
done so far to the images (from capture to the side-effects of our
basic workflow steps), and let's also consider the final output
goal-media, so we factor in all these variables and produce the
best possible results, while still being able to preserve our
images' integrity as high as we can.
As a result, FLEXSharp is flexible enough to offer:
-Control of Noise levels derived from a wide range of ISO speeds
-Control Noise derived from intermediate workflow steps (Lelvels,
curves, etc. which all increase noise).
-Sensitivity to distribution of spatial frequency on the original
capture, translating into the ability to separate skies and clouds,
from skin, hair, leaves in the shadows of trees, grass, coarse
surfaces of roads or mid frequency textures pulled out from noise
in purely flat areas, etc.
-Control of low and high frequency artifacts (dark or bright halos,
abnormally bright spots or edges, compression "crystals", etc.).
-Control of image-enhacement attributes of the sharpening process,
such as MicroContrast levels, and overall dimensionality (a benefit
of shooting digital). Micro-contrast is KEY for achieving an
optimal balance of "edge" sharpness, low-to-mid frequency detail
enhancement, actual strenght of sharpening required, and final
output goals (DPI, size, etc.)
-Ability to choose balance of spatial frequency enhancement bias
(low and high, or mid-to-high, for instance). This is extremely
important in order to extract true detail from ISO 200-800 shots,
while still preserving dimensionality and maximizing S/N ratios,
where practically ALL sharpening tools avaialble out there fall
apart.
-Overall image quality and integrity preservation, regardless of
how soft or hard you go on the sharpening throttle.
-Keep things
simple , with minimal or no intervention of end user.
Below you can find the example on hand (from which I have the
original ORF file, shot with Sigma's 50mm EX Macro). This tree
log/trunk lives in the back of
my yard, where I can evaluate how
true-to-live the results posted here are (FLEXSharp-v1.4 and NIK's
left):
The sample posted here is intended of
very high DPI output. Not
necessarily for a "soft-and-pleasant" look on your screen (even
though, in real life, it is does not look soft either).
Compare (in both samples) the brigh-colored spots and dark halos
around the edges of leaves, etc., as well as the levels of noise in
shadows
and detail-intensive areas (
do not confuse noise with
detail, as I see this log every day and I can tell you how it
really looks).
In this case, I ran FLEXSharp directly over the entire crop, with
no help from the FLEXMask masking routines, and, as you can see,
there is a strict control of the signal-to-noise ratio built-on the
sharpening filters and parameters. That's why Nik's sample look
rather "grainy" and less-dimenstional, whereas FLEXSharp looks
crisper and more dimensional (sharpened areas stand-out more, and
they will do so in paper, as well).
Well, that's enough for today. Just my two cents on what I believe
sharpening needs to be when it comes to enhancing and extracting
the best out of our work.
Happy shooting!
