On 2010-01-07 15:21, Ken Waller wrote:
Doug - The shot of the defect is from a Tiff that was made directly from
the RAW capture (DNG).
On other images, the black line (column) is constant in width & color.
I've printed off the mailing label & boxed the body - I'll mail it
tomorrow.
Oh, shucks, man. You don't need to do that. I was just thinking of an
email or a Flickr posting or something. In fact, if you could just
email me [either this address, or franklin (at) shootingshark (dot) com,
or DougF (at) NutDriver (dot) org] the original "TIFF from DNG" version
of that small crop file with no compression, that'd be great. I just
want to see what I can see at extreme magnifications, on the histogram,
and stuff like that. Might even play around with stuff like edge
sharpening.
As clean as the circular edge of the bright spot is, it appears to be
covering a /lot/ of pixels, based on the smoothness of that edge, but
from the size on the original shot, it doesn't look like it's actually
covering that many pixels. And I want to see if the black line fades in
and out smoothly, or at all, in the original RAW data.
This is the sort of case where I'd really like to be able to get deeper
access to the sensor data and the downstream in-camera processing. For
example, the "de-Bayerization" of the sensor data could be why that
white circle ends up so circular, even though apparently covering too
few pixels to appear that smoothly circular. That sort of effect can
happen at several different levels of the processing, from the
electronics of the sensor's photo sites all the way to the compression
algorithm used on the image data.
--
Thanks,
DougF (KG4LMZ)
--
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.
