<<Moose, just a query about deconvolution sharpening:
<<Does the algorithm for deconvolution take into account the lens used
Very late and trying to get home. In brief, it seems clear DXO uses "blind"
R-L deconvolution but weights it where the lens in softest with their
profiles--so yes, it does take the lens into effect but the algorithm is not
really changed, just how much is applied.
It likely assumes the PSF (point spread function) is Gaussian blur as it
works well and the Gaussian deconvolution kernel is fast to compute. FM and
Topoz in Focus likely uses blind R-L deconvolution with a Gaussian assumption
for the PSF.
I have not used Topaz but is has an option to suppress ringing artifacts as
well as picking the blur radius. How well that works I don't know, but suspect
FM is just as good if carefully used in a layer and applied judiciously.
Canyon's DLO from their explanation does use the specific PSF from each lens
for aberrations as well as for diffraction--very cool really. Deconvolution
sharpening used to undue diffraction effects would not be predicted to work
well, as there is a hard high freq cut off in the data.
In empiric use it does work fine to a degree--long post on this in the past.
http://lists.tako.de/Olympus-OM/2013-05/msg02068.html
Motion blur is a whole other PSF and FM and Topaz allow one to specify this.
Mike
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|