lightweight/portable QTVR setup for mountaineering/hiking

http://www.vrmag.org/vartist/guest_artist/AN_INTERVIEW_WITH_A_QTVR_ORIGINAL_-_KEN_TURKOWSKI.html

I swam several times a week, and rode my bike 50 miles in the Berkeley and Oakland hills. It was especially on the uphill bike climbs that I really appreciated the groves of trees and panoramic views that the San Francisco Bay has to offer. I bought a Minolta manual-everything camera, and tried to capture images of all of my outdoor experiences, so I could share it with my family and friends in Chicago. I got into backpacking big time, starting when one of my Chicago buddies came out so we could backpack together in King’s Canyon National Park before Christmas. Though I was naïve about winter mountaineering, we luckily had good weather because of the drought. That summer, I went backpacking in Mt. McKinley National Park with a couple of other Berkeley students. I tried to capture the awe of the mountainous environment while backpacking, but the pictures just couldn’t capture the grandeur of it all.

I own a D70 with Nikkor 10.5 mm. With this setup you have to shoot zenith and nadir for a full spherical.
I made a monopod construction which allows me to shoot zenith (1x) and nadir (2x)
I shoot 8 pictures around because it's better to estimate then 6 and gives enough overlap for ghost
retouching.

It works quite well to stitch. All my works on wwp are made this way:
http://geoimages.berkeley.edu/wwppeople/html/PeterNyfeler.html

And this additinal ones shot at the same spot:
http://www.nyfeler-mueller.ch/wwp305/wwp305.htm

An object movie how it works can be seen here.
http://www.nyfeler-mueller.ch/MonoPodHead/QTVR/MonopodHeadQTVR.html

I use a Canon 300D + Sigma 8mm/4 on a Bogen QTVR head (W/tripod), no need to shoot zenith or nadir. I have my setup in a backpack, & man..is it HEAVY!! I suppose my setup could be lightened with could be used in the above "flex bracket"+monopod.

I understand a 8Mpixel Canon 350D (with lens) can be obtained for like $450, which is an attractive price-point.

B) point&shoot camera using bracket

http://www.outsidethelines.com/qtvrmonopod/

C) point&shoot camera using plumbline bob & manual rotation (guessing overlap)

Can someone give me some recommendations for B) & C), like preferable state-of-the-art P&S cameras? Would you prefer those with a "SLR look", e.g.

http://www.dpreview.com/reviews/specs/Canon/canon_sx100is.asp [ $250 ]

than a "toyish" P&S camera, since it will have manual control & larger aperture lenses? Although, some "toyish" P&S cameras DO have manual control.

Can I get some URLs for some nice QTVR panos done with P&S cameras?

Once Upon A Time, email@hidden wrote:

> Aside from the expense of the stitching software and film, this is the

> cheapest VR rig I have ever found. Total expense: Less than 2 American

> dollars plus film processing.

a nice solution indeed, but i may have you beat... ;>

back in '95 when i was shooting my first pano, i merely took a clock from

the wall and centered my tripod over it using a makeshift plumb-line

(shoelace w/ a pen dangling from the end). of course it was far from

accurate, but i was able to get decent results-- shooting every clock

position (30 degrees).

in fact, the pano even ended up on a commercially released enhanced CD.

of course i know there are those on this list (and surely on helmet's

list)

who just shoot freehand w/ no tripod or rig at all and correct everything

post!

On Mon, 29 Jan 2001 03:56:18 -0800, Oleg March <email@hidden> wrote:

> Correct me if I'm wrong, but I'm under impression that the wider the lens is

> the more critical camera alignment should be. I think the reason is that wider

> lens show objects that are much closer to the camera for proper stitching of

> which observance of nodal point alignment is essential. If the camera moves

> from position A in first shot to position B for the second shot then the angle

> between "A", object and "B" is much greater if the distance from camera to

> object is short (such as when shooting with wide lens).

The lens FOV doesn't matter per se. The important thing is the ratio of the

closest object to the deviation/uncertainty in the nodal point. If this is

100 or greater, then you'll have a decent stitch; less than this, and you'll

probably have to do manual stitching and touch-up..

However, with a wider FOV lens, you would tend to have a wider angular

separation between adjacent shots (think 50% overlap), as well as less shots

, so less probability for parallax. Parallax is roughly proportional to the

the angle between shots, so if you double the angle, you double the angular

parallax. Then again, with the same number of pixels in the camera, you

have halve the angular pixel resolution, so in the end you'd be

misregistered by the same number of pixels.

In a nutshell then, with two handheld shots, the angular parallax due to a

moving nodal point is larger with a wider FOV, but the pixel parallax would

be about the same. With a wider FOV there's less shots, so there would be a

smaller probability of parallax.

Ken Turkowski email@hidden

Immersive Imaging Technologist http://www.worldserver.com/turk/

Apple Computer, Inc. +1 (408) 974-6699

1 Infinite Loop, MS 302-3VR

Cupertino, CA 95014

References:
	>Re: VR Photography Review of Nikon D3 now online (From: Scott Highton <email@hidden>)