> Interesting. I guess the need for explanation belongs to Phil
> Askey. Wonder why he couldn't do normal sharpening?
I reskimmed his review and I didn't see where he tried
post-processing a JPEG file. He made lots of comments about the
in-camera sharpening, though. His conclusion is somewhat
similar to mine, though.
What it appears to me is the in-camera sharpening algorithm
isn't USM-based, but rather uses an increase in contrast to
achieve the appearance of sharpening. From appearances it is
doing so by darkening a dark "edge", but unlike USM, doesn't
lighten the bright side of the "edge". The advantage to this
would be a significant decrease in the appearance of halos.
In Phil's comparision of RAW images, it looks like the K10D
takes a back seat to no one. Those are pretty sharp files. And
in my testing of his sample images, I was able to achieve pretty
much equivelent sharpnesses with similar sharpening as what he
applied (80%, 1-pixel, 0-threshold) to the RAW file in ACR.
However, I did get better results with about a .7-pixel diameter
to keep halos under control.
I fully expect Pentax to issue a software upgrade to the K10D
once they figure out what went "wrong" in their algorithm, but
this is pointing out a serious flaw in Phil's comparison tests.
He's always tested JPEG files, but it's really in the RAW files
where we find out the foibles of the camera. For instance, the
400XTi's hidious noise levels in RAW, but "ultra-clean" CMOS
characteristics in the JPEG files. Hmmmm. If all you compared
were the JPEG files, you'd believe that the 400XTi had the
cleanest sensor around.
AG
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|