camera offset

[Mihai Bodea]

hello!

I'm new. from the wild yonder. and have medium level English knowledge.

I have in mind improvising a camera rig that alows panning without changing perspective; for that I need to know how far to offset the camera backwards with respect to the lens. and even more, how about when I use a zoomlens.

I'd be greatfull for any reply.

misu

[Michael Collier]

find out where the plane of focus is. On a film camera, there is a marking on the body and usually a pin to attach a tape measure. On a video camera its not generally marked, so you sort of have to guess. It will be a few cms behind where the lens mounts, where the chip(s) are. Offset the camera so the plane lines up with the axis of rotation in the pan head. You can also offset the tilt so that too is in the axis of rotation, but to do that you need a specialized head, a standard fluid head would be impossible to offset the tilt into the axis of rotation. The pan is easy though. Can't imagine why it would matter except for shots really really close to the front element of the lens. Even cameras mounted relatively far from the focal plane don't present much perspective change past three feet or so. Let me know if that helps.

[Mihai Bodea]

thank you, Michael!

already did what you suggest. it isn't working. the axis of the movement has to cross some nodal point in the lens. it works the same with still panoramic cameras, where the lens pans by itself around this virtual nodal point, covering the wide angle. the problem is that I don't know the math for that nodal point. it has to be on the axis of the lens, in front of the focal plane, but how far? I imagine it varies with lens. that is why I'm wondering about zoom lens correction. with the tilt, even if I don't need it for now, I know how it's working.

my problem is that I have objects close to the camera, as well as a distant background, and I don't want the these different planes to float, one in front of the other, it has to be static.

still waiting,

misu

[Michael Nash]

You need to find the nodal point of the lens, not the film plane.

 

http://www.edb.utexas.edu/teachnet/qtvr/nodalpoint.htm

 

http://www.path.unimelb.edu.au/~bernardk/t...hoto/nodal.html

[Brad Grimmett]

Misiulica,

Please change your user name to your real first and last name as per the forum rules.

Thanks.

[Mihai Bodea]

I found what I needed. except it is an empirical approach that is suggested in those links. finding the iris plane (aka nodal point) o given lens might be impossible so I'll be tweaking the rigg to get it right.

thanks!

misu

[Will Earl]

The way I was taught to find the nodal point is to place two objects directly in line with the camera both at a set distance apart, so the first object is 1 metre away from the camera and the second object is 2 metres away from the camera. By sliding the camera backwards and forwards along the baseplate (or a rotating base of some sort) pan the camera left and right. If the camera is rear of the nodal point then the first object will stay closer to centre of frame, where if the camera is forward of nodal the rear object will stay closer to the centre of frame. If the camera is nodal then both objects will line up during the pan.

[Chris Keth]

The way I was taught to find the nodal point is to place two objects directly in line with the camera both at a set distance apart, so the first object is 1 metre away from the camera and the second object is 2 metres away from the camera. By sliding the camera backwards and forwards along the baseplate (or a rotating base of some sort) pan the camera left and right. If the camera is rear of the nodal point then the first object will stay closer to centre of frame, where if the camera is forward of nodal the rear object will stay closer to the centre of frame. If the camera is nodal then both objects will line up during the pan.

 

You beat me to it. The trick works beautifully, I have used it to set up DSLR panos before. C-stands work brilliantly.