## Re: Precision required for nodal point?

Expand Messages
• ... As always: It depends. If you have structures in the near forground it should be more precise than if you shoot in wide open places. For general use I d
Message 1 of 7 , Oct 1, 2007
On Monday, October 01, 2007 at 10:10, paul womack wrote:

> But HOW precise do I need to be in
> evaluating and building the nodal
> point?
>
> 1/2 a mm?
> 1 mm?
> 5 mm?

As always: It depends. If you have structures in the near forground
it should be more precise than if you shoot in wide open places. For
general use I'd say 3mm precision is enough. However, if you use a
fisheye lens the no-parallax-point moves anyway:
http://tinyurl.com/94g7c (scroll down to Results and look at the
Graph)

Apologies that I didn't write the wiki article I proposed more than
one year ago about how to calculate the possible error. Perhaps I
should do it right now...

best regards
--
http://www.erik-krause.de
• ... My widest lens (wide end of zoom) is 41 mm BugBear
Message 2 of 7 , Oct 1, 2007
Erik Krause wrote:
> On Monday, October 01, 2007 at 10:10, paul womack wrote:
>
>> But HOW precise do I need to be in
>> evaluating and building the nodal
>> point?
>>
>> 1/2 a mm?
>> 1 mm?
>> 5 mm?
>
> As always: It depends. If you have structures in the near forground
> it should be more precise than if you shoot in wide open places. For
> general use I'd say 3mm precision is enough. However, if you use a
> fisheye lens the no-parallax-point moves anyway:
> http://tinyurl.com/94g7c (scroll down to Results and look at the
> Graph)

My widest lens (wide end of zoom) is 41 mm

BugBear
• ... And which angle do you intend to rotate between shots? Meanwhile I ve added the formula to http://wiki.panotools.org/Parallax Some more info and a neat
Message 3 of 7 , Oct 1, 2007
On Monday, October 01, 2007 at 13:21, paul womack wrote:

> > As always: It depends. If you have structures in the near forground
> > it should be more precise than if you shoot in wide open places. For
> > general use I'd say 3mm precision is enough. However, if you use a
> > fisheye lens the no-parallax-point moves anyway:
> > http://tinyurl.com/94g7c (scroll down to Results and look at the
> > Graph)
>
> My widest lens (wide end of zoom) is 41 mm

And which angle do you intend to rotate between shots? Meanwhile I've
missing, but may be it helps...

best regards

--
http://www.erik-krause.de
• ... [snip] ... The actual precision needed will depend on the angular field of view of your lens, how far you rotate the camera between adjacent shots, how
Message 4 of 7 , Oct 1, 2007
--- In PanoToolsNG@yahoogroups.com, paul womack <pwomack@...> wrote:
>
> I am just setting out to make a wooden
> pano head for my Canon a630.

[snip]

> But HOW precise do I need to be in
> evaluating and building the nodal
> point?
>
> 1/2 a mm?
> 1 mm?
> 5 mm?

The actual precision needed will depend on the angular field of view
of your lens, how far you rotate the camera between adjacent shots,
how close to the no-parallax point the objects in the field of view
approach, especially at the seams, and in general the amount of
parallax mismatch you can tolerate in your panorama. For landscapes
and even for most indoor shots, 5 mm would be accurate enough. For
tabletop setups and the like where the subject approaches very close,
you might need greater accuracy than that.

Alan Ball
scoundrel1728
• This is entirely possible. I built a wooden panorama head a few years ago as a prototype, but never got around to the next step of refinement. Other than
Message 5 of 7 , Oct 1, 2007
This is entirely possible.

I built a wooden panorama head a few years ago as a "prototype," but
never got around to the next step of refinement. Other than being
rather larger and heavier than I'd like, it works fine! :)

Since I thought I might be using different lenses or cameras (and
because I expected that my "first cut" would not be perfect), I
included fore and aft + left/right adjustment capability by including
a couple of "Manfrotto 3273 Quick Release Adapter with 2372 Plate"s
installed at right angles between the camera attachment point and the
rotating base. The 3273/3272 pair seems to be priced about \$45.00-
50.00 (US).

[And GOOD LUCK finding the product on the Impossibly Awful To Use
Manfrotto site. Good Grief!!]

If you don't mind re-creating your pano head several times ("OK, now
drill the hole 1/64 inch *lower* this time"), then you don't need
that stuff.

Also, my camera is attached by way of a Quick Release Clamp, which

If you're using only a single camera and lens, you in theory could
get away without using such things. You might have to make several
versions before you decided that your positioning was good enough.

A very useful tool in all this: a self-leveling Laser plumb/level
thingy!

It is easy to pay far too much for such a thing. Hunt around during
the "holiday/Father's Day sales" times and see what you can find.
Mine cost about \$35.00 (US). The principle is astonishingly
simple... A horizontal and/or vertical laser line is projected from
an emitter that is DANGLING from an attachment inside the unit.
Dangle means pendulum means... GRAVITY is running the show. What a
deal... thank you Mr. Newton.

ASSUMING that the little tiny laser emitter was correctly calibrated/
oriented during manufacture, it will project a *very* accurate
horizontal or vertical line. Get out your string-and-weight plumb
bob or 6 foot (2 metre) construction level and challenge the results.

What this is useful for: Adjusting or Confirming that the center of
the lens is aligned with the vertical rotation axis (yaw) of the pano
head. Similarly, that the center of the lens is aligned with the
*horizontal* rotation axis (pitch) of the pano head.

That is: you can project a laser line against your pano head, and
pretty easily see how close the pivot points (holes) you've created
match up with the center of the lens axis.

Have Fun!!

eo

On Oct 1, 2007, at 2:10 AM, paul womack wrote:

> I am just setting out to make a wooden
> pano head for my Canon a630.
>
> Along these lines:
>
> http://www.worth1000.com/tutorial.asp?sid=161123
>
> I understand about the nodal point, and
> needing to rotate about it.
>
> But HOW precise do I need to be in
> evaluating and building the nodal
> point?
>
> 1/2 a mm?
> 1 mm?
> 5 mm?
>
> BugBear
• ... The wiki artikle (s.a.) is finished for the time being (diagram still missing). You should be able now to estimate what precision you need. Interesting
Message 6 of 7 , Oct 2, 2007
On Monday, October 01, 2007 at 17:04, Erik Krause wrote:

> > > As always: It depends. If you have structures in the near forground
> > > it should be more precise than if you shoot in wide open places. For
> > > general use I'd say 3mm precision is enough. However, if you use a fisheye
> > > lens the no-parallax-point moves anyway: http://tinyurl.com/94g7c (scroll
> > > down to Results and look at the Graph)
> >
> > My widest lens (wide end of zoom) is 41 mm
>
> And which angle do you intend to rotate between shots? Meanwhile I've
> added the formula to http://wiki.panotools.org/Parallax
> Some more info and a neat diagram (someone volunteer?) is still
> missing, but may be it helps...

The wiki artikle (s.a.) is finished for the time being (diagram still
missing). You should be able now to estimate what precision you need.
Interesting thing that the parallax error in the finished panorama
does not depend (directly) on the used focal length (only on the
rotation angle between shots and the displacement of the no-parallax
point).

Please tell me if there are some errors, since I did the calculation
all on my own ;-)

best regards
--
http://www.erik-krause.de
Your message has been successfully submitted and would be delivered to recipients shortly.