Olympus-OM
[Top] [All Lists]

[OM] Calculating parallax error, was: Panorama Heads

Subject: [OM] Calculating parallax error, was: Panorama Heads
From: Chuck Norcutt <chucknorcutt@xxxxxxxxxxxxxxxx>
Date: Sat, 25 Aug 2007 16:57:39 -0400
I figured this had to exist somewhere.  I looked around and found a nice 
page of photo related calculators one of which is the amount of parallax 
error from being off the no-parallax-point.  The output is specified in 
terms of degrees and also number of pixels in the image.  Focal length 
is not mentioned as an input parameter but is certainly there in another 
form which is the image field of view.  From another page I found some 
discussion of having reasonable stitching results with errors of about 6 
pixels.  But there are probably a lot of caveats around that number.

Anyhow, after playing around with the calculator a bit (and assuming the 
  "nodal point offset" is about equal to the focal length of the lens 
away from the tripod socket) it appears that one can make a rough 
generalization for 35mm frames that no panorama head is needed if the 
closest object is at least as far away in feet as the twice the focal 
length of the lens in millimeters.  eg; if you're shooting with a 50mm 
lens and the closest object is 100 feet away you'll probably do OK with 
no panorama head.  You may even get away with less but, YMMV.

Chuck Norcutt

Chuck Norcutt wrote:
> I should also have added; before running out to buy a pano head try 
> shooting some on a tripod but without a head.  You may find that you 
> really don't need a head.  I've probably shot less than 15 or 20 panos 
> but I don't have a head and haven't needed one.  Probably half of them 
> are shot hand held and are also successful.  But shooting hand held I 
> still try and make my body mimic the rotation on a tripod head.  All of 
> my panos are landscape type with the subject probably at least 2500 
> times the focal length of the camera away from me.  There's a value 
> somewhere in the void and I don't know what it is that says if the 
> subject is X times the focal length away you don't need to worry about 
> parallax error.
> 
> Chuck Norcutt

==============================================
List usage info:     http://www.zuikoholic.com
List nannies:        olympusadmin@xxxxxxxxxx
==============================================

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz