On 8/25/2013 3:11 AM, Nathan Wajsman wrote:
> I know that some degree of unsharp masking is usually necessary, but if my
> understanding is correct, this software artificially converts out.-of-focus
> shots into something that is more or less in focus. That seems a degree of
> manipulation well beyond standard USM.
I believe USM is more a form of artificial manipulation than deconvolution. If
you read my treatise at the link, you
should understand that it changes the image at edges based on entirely
arbitrary parameters that the user sets. There is
no reference at all to what the original image the lens projected on
film/sensor was like.
One can, however, see how proper use makes details otherwise invisible to the
eye visible, and deduce from that that the
detail was in the original projection, as the process cannot invent detail
(unlike the software on cop shows).
OTOH, deconvolution, in its pure form, uses measurements of aberrations in a
particular lens to reconstruct what the
lens would have projected, had it been less imperfect. This seems to me to be
no more, and probably less, artificial
manipulation than USM.
I suspect that you are simply used to USM, so it seems more OK.
In practice, with generic deconvolution apps, such as FM, Topaz In-Focus, etc.,
the algorithm is not tuned to the
specific lens. DXO does provide lens specific parameters in its processor. In
my one comparison, however, it did no
better than FM.
As to "this software artificially converts out-of-focus shots into something
that is more or less in focus", I was not
referring to such use, only to mild use to make web size images that more
accurately reflect the visual qualities of the
large originals.
Indeed, both USM and DC may be used to do that. I have used them, together with
artificial blur, even to move the
apparent plane of focus of shallow DOF images forward or back. That is clearly
a couple of bridges too far for you, and
I respect that. I've only done that, so far ;-) , to show roughly what an image
with the plane of focus properly set, to
my eye/taste, might have looked like.
I only propose that, used for its intended purpose, FM is no more manipulative
than the sliders in LR.
Side notes:
Mike G and others have pointed out that in theory, DC should not work for
correcting diffraction blur. In practice, with
my lenses and images, it increases detail visibility and 'sharpness' in images
as apertures small enough that folks like
Chuck say they are diffraction limited.
I don't know why, nor much care, as a pragmatic processor, but FM generally
works better for me for post down sampling
resharpening than USM, but that would not be true in LR, where (I believe) it
couldn't be done on a separate layer and
the layer opacity adjusted to taste, or even masked. One of the other DC apps
(Topaz?) allows sub pixel settings to make
the effect more subtly adjustable.
In my mind, when theory does not accurately predict real world results, it's
the theory that needs correction. :-)
> Anyway, it is all a moot point.
Until, as is likely, it does work in LR. Despite the artifact damaged images
that people who overuse the LR controls not
uncommonly produce*, it is a dominant app. So the producers of FM are probably
hard at work right now.
> If it doesn't run from within LR, it doesn't run on my computer. I am not
> willing to use more than one piece of software on an image. The only
> exception is panorama stitching.
Makes perfect sense to me. Barring something really special, I only use PS,
including for panorama stitching.** ;-)
P. S. Moose
* Exactly analogous to misuse of SC and USM, BTW.
** And, perhaps, focus stacking, from what was posted while I was traveling.
--
What if the Hokey Pokey *IS* what it's all about?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|