|
 |
Monday, May 17, 2004 |
Many people write me to ask why their panoramas aren't sharp, clear,
and relatively free of stitching artifacts like mine (obviously they
aren't referring to my 1996 panos that are still encoded with the
Cinepak codec). Hopefully the
following example will help to answer that question.
Awhile back I was scanning panoramas shot on JMT3, a backpack trip from
a couple summers ago in which a couple friends and I went from Onion
Valley to Whitney Portal on the John Muir Trail. I was using a good
quality 4000ppi scanner,
and
VueScan. Unlike some scanning software, VueScan has an option that is
tailor-made for panoramic photographers -- the ability to lock both
color
and exposure settings for future scans based on an initial scan.
Previously I found that when scanning panoramas with VueScan, choosing the
initial image
you choose to lock all subsequent exposure and color values from is
critical. Even then, sometimes it is just not possible to find one
image which is representative of the color/exposure values of the other
11 shots. So I ended up with this panorama,
which was not good enough to post to VirtualParks. In hindsight, it
wasn't only VueScan but the scanner was just not as good as my current
scanner.
This scene on a rocky hilltop overlooking Guitar Lake under the
towering edifice of Mt. Whitney is extra challenging to produce. It was
shot on a monopod where gusts of wind pushed the nodal point off. The
cloudy lighting kept changing, and in fact at least one exposure was
completely shaded while others were brightly lit by sun. The scenery is
mostly rocks, so there is little frame of reference for getting the
color right.
So I set that scene aside. Last year I got a new scanner, a Nikon Super
CoolScan 4000. I tried VueScan with it but it consistently didn't get
the film offsets right when using the film strip adapter. So I'm using
Nikon's software, which also allows you to lock exposure. That same scene is now on VirtualParks here.
It's not perfect, not my favorite from the trip, but it's one of very
few that was scanned twice, and I'm satisfied that it is close to my
memory of what the place was like.
The point is sometimes you have to go back to square one, as hard as it
is to lose all the time invested to date. Also, every time you upgrade
your scanner, every single one of your film-based scenes has the
potential to improve in quality. Can't say that about digital-camera
based panoramas!
7:19:43 AM
|
|
© Copyright 2006 erik goetze.
|
|
|
|
May 2004 |
Sun |
Mon |
Tue |
Wed |
Thu |
Fri |
Sat |
|
|
|
|
|
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 |
17 |
18 |
19 |
20 |
21 |
22 |
23 |
24 |
25 |
26 |
27 |
28 |
29 |
30 |
31 |
|
|
|
|
|
Apr Jun |
|
Purpose |
VRlog provides news, developments and analysis of the virtual reality (VR) world from a nature photographer's perspective. Since I am not connected to or funded by any VR vendor, I intend to objectively appraise what's going on, and the direction VR is headed in. -- erik goetze
|
|
|
|
Syndicate VRlog |
The items on
this site are available in an RSS newsfeed, an XML file format.

|
|
|