Archived posting to the Leica Users Group, 2012/02/04

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] Pano stitch test: Sunset light
From: roark.paul at gmail.com (Paul Roark)
Date: Sat, 4 Feb 2012 15:07:49 -0800
References: <CAJ3Pgh7u6rxEftaF8498Z7RYrD8N5cESB6VaSoC=J-Svp_7baw@mail.gmail.com> <CB52E615.19E2F%mark@rabinergroup.com> <CAF8hL-H8M+Z=J2DkEox-6BahRN+CXy-9NHgNUSAaZz3CarPu2w@mail.gmail.com>

Mark, the very right lower corner of each frame is shown at
http://www.paulroark.com/BW-Info/Corners-18-v-21mmZM.jpg .

The blades of grass will not be exactly the same because the lenses
cover different areas. ?They were centered on the same central target.

Reasonable people can differ as how much the sharpness in the very
corners matters, particularly if you generally crop to, say 8x10, but
I don't always crop. ?The results correlate very closely with my real
world experiences over the years with symmettrical v. retrofocus
optics.  I favor rangefinders in no small part due to their ability to
use more symmetrical wide angles -- smaller and better.  With the
vignetting problems of electronic sensors, the number of top notch
symmetrical designs may be on the wane.  As one who does B&W, I don't
care that much about the red edge.  So, I grabbed what may be near the
end of a dying breed -- the 21mm C-Biogon.

I look at it this way: the more the optic has to bend the light, the
harder it is to balance out the various aberrations.   At any given
price point, the symmetrical design is probably going to be noticeably
better than the less symmetrical one. ?It seems to be a matter of
degree, and I don't know where Zeiss draws the line in its naming.
With the 21mm C-Biogon, I think Zeiss hit a nice compromise for film,
but with the dominance of color and digital, it's probably the end of
that line for that focal length.


Richard, you're ahead of me on the nodal point rotation. ?I've had to
do a lot of manual work on some of my old stitch jobs.

As a practical matter, what I find is that where I'm doing a major
foreground-background sweep, it's usually a vertical shot without
stitching that gets me into the parallax issue.

What I find I do rather frequently to get a bit more coverage, a few
more pixels, and extreme DOF in a horizontal image is to, first, take
a horizontal, stacked composition that is aimed somewhat down, and
then just take one shot at infinity with the camera moved up to
capture more of the sky. ?That way I'm using the full width of the
frame (and need those corners good), and I find skies easy to stitch,
manually if need be. ?That approach can't get an extreme pan, but with
a 21mm (with good corners) the coverage gives an 8x10 proportioned
final print with about the same coverage as a cropped 15mm, and with
great sharpness and relatively easy workflow.

I have been curious about the differences between Gigapan and CS5
stitching. ?Have you done a comparison?


By the way, I have tried to make a (roughly) matching pair of sky
images, one of which is the "Approaching Storm" M9 stitch, and other
is a single frame, Bronica RF645 with 45mm and Tech pan. ?It was an
interesting exercise to work up the two approaches and see how they
compare.  The coverages are quite different, of course.  It's sort of
a "before and after the storm" type of feeling I was looking for. ?See
http://www.paulroark.com/SYV-Skies.html


Paul
www.PaulRoark.


Replies: Reply from roark.paul at gmail.com (Paul Roark) ([Leica] Pano stitch test: Sunset light)
Reply from richard at richardmanphoto.com (Richard Man) ([Leica] Pano stitch test: Sunset light)
In reply to: Message from roark.paul at gmail.com (Paul Roark) ([Leica] Pano stitch test: Sunset light)
Message from mark at rabinergroup.com (Mark Rabiner) ([Leica] Pano stitch test: Sunset light)
Message from richard at richardmanphoto.com (Richard Man) ([Leica] Pano stitch test: Sunset light)