Hello all!
I've been struggling twice so far to obtain the so called "UniWB"
on a Pentax K-30, but without any "real" success so far.
Has anyone been able to succeed at this? (And maybe share a raw
photo that has the appropriate WB settings?)
I've tried the method described on
http://www.guillermoluijk.com/tutorial/uniwb/index_en.htm , and I've
obtained the coefficients: (RGB) 0.98 1.00 1.00 or 1.01 1.00 1.02.
(Coefficients are obtained by running `dcraw -i -w -v ./photo.dng`,
and also checked against the output of `rawshack`.)
However when using any of these WB settings, the in-camera
histogram just ignores the blue channel (although in RAW it exists and
is well exposed)...
I've then tried a more "sane" approach: I've played with the
manual WB by setting the temperature (in Kelvin) and the bias, and see
what coefficients I obtain. The best one (i.e. closer to 1) I've
obtained (RGB) 1.51 1.00 1.51, for 3700K and a bias towards green and
a little towards blue (I think).
Moreover I've also set the following knobs in the hopes that the
histogram is more "liniar":
* custom image to "neutral";
* contrast to the extreme left (i.e. negative value);
* "high/low adj" to the extreme left (i.e. negative value); (I
don't exactly know what this implies, but the result is satisfying;)
I've compared the resulting JPEG histogram with the actual RAW
histogram and these settings seemed the "closest" (by a visual
interpretation, however without taking into account WB).
I don't know yet how to handle the following:
* saturation --- some say that it should be set to the left;
* sharpness --- what is the difference in Pentax between
"sharpness" and "fine sharpness";
* high/low adj. --- what does it actually do?
Is there a more "in depth" documentation about the implication of
all these? (The K-30 manual is a little bit sketchy and shallow on
these subjects...)
As seen I'm a little bit puzzled, however I'm confident that given
time I'll figure things out.
Thanks,
Ciprian.
--
PDML Pentax-Discuss Mail List
[email protected]
http://pdml.net/mailman/listinfo/pdml_pdml.net
to UNSUBSCRIBE from the PDML, please visit the link directly above and follow
the directions.
