Finding best no-parallax-point

180 views
Skip to first unread message

Panorama-Fotografie - Carola de Poel

unread,
Feb 6, 2022, 4:39:31 PM2/6/22
to PTGui Support
Hi all,

I'm trying to find the best no-parallax-point for my new camera. 
I use the info on https://ptgui.com/cullman_noparallax.pdf
But I am afraid that I don't understand it right, because I cannot figure out why 62mm is the best position in the first example.
I always thought that the best position is the one with the smallest amounts in all three values (average, min distance and max distance).

I made a batch of photo's this afternoon, varying from 9 to 11 mm (on a nodal ninja 4 panoramic head).

I made a excel sheet with the data of mm, average distance and max distance (I understood that min distance was not important for this). I created a graph from it (see attachment). 

Can someone tell me how many mm is the best to use. I think 10.8 mm, but maybe I'm wrong as I can't interpret the example well either.

I found that when you re-stitch the same photos, the optimizer gives different values each time, which makes it even more difficult to understand.

My gear:
Canon EOS R5
Canon 8-15 mm fisheye lens
Control ring mount adapter EF-EOS R
Nodal Ninja MK4

Thank you very much.
Kind regards Carola 

Schermafbeelding 2022-02-06 om 22.04.51.png

Panorama-Fotografie - Carola de Poel

unread,
Feb 6, 2022, 5:56:51 PM2/6/22
to PTGui Support
I think I was not clear in my explanation.
I used the 8-15 mm lens on 15 mm (this time, some times I use 12 mm because I have to take less photo's to make a full 360).
I varied on the nodal ninja MK4 between 9 and 11 cm, not mm. See attachment.




Op zondag 6 februari 2022 om 22:39:31 UTC+1 schreef Panorama-Fotografie - Carola de Poel:
20220206_151735.jpg

PTGui Support

unread,
Feb 7, 2022, 4:05:19 AM2/7/22
to pt...@googlegroups.com
Hi Carola,

As you noticed the optimizer gives different values each time, this is
because randomness is involved in the control point generator. This
random placement causes more variation in control point errors than the
actual parallax.

So actually I don't think this method works well and I've removed the
link to the article from the support page.

I think the classic method of using live view and observing the parallax
works best:
https://www.youtube.com/watch?v=zEgLgReARxs

But I don't think you've wasted your time, you can probably use your
images to find the best setup: create a project for each image set, run
Align Images. In the Panorama Editor, disable Find Optimum Seams. In the
Image Parameters tab, give the first image a high blend priority (1000).
Check the control points and perhaps use the 'Generate Control Points
Here' function to ensure you have many control points spread our over
the overlap area. Run the optimizer.

Open the detail viewer and zoom into the overlap area. In the Create
Panorama tab, toggle the display of image 1 on and off. This will
alternate the visibility of the two images, so you can observe the
parallax shift and see which setup gives you the lowest parallax.

Kind regards,

Joost Nieuwenhuijse
www.ptgui.com
> --
> You received this message because you are subscribed to the Google
> Groups "PTGui Support" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ptgui+un...@googlegroups.com
> <mailto:ptgui+un...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ptgui/fa8c7fe4-bcb5-4461-ba09-cee45a7c3102n%40googlegroups.com
> <https://groups.google.com/d/msgid/ptgui/fa8c7fe4-bcb5-4461-ba09-cee45a7c3102n%40googlegroups.com?utm_medium=email&utm_source=footer>.

Panorama-Fotografie - Carola de Poel

unread,
Feb 17, 2022, 7:05:35 PM2/17/22
to PTGui Support
Thank you very much Joost! 
I'm going to work with your suggestions!
Kind regards, Carola

Op maandag 7 februari 2022 om 10:05:19 UTC+1 schreef PTGui Support:

Steve Taylor

unread,
Feb 18, 2022, 12:24:38 PM2/18/22
to PTGui Support
Hi Joost, 
That technique works fine for wide angle and short focal length lenses, but is a lot more complicated with a 100mm mm lens with at f32 0r f22 with a minimum depth of field focus of 5m+ to get infinity sharp.
 steve

Erik Krause

unread,
Feb 18, 2022, 1:45:37 PM2/18/22
to pt...@googlegroups.com
Am 18.02.2022 um 18:24 schrieb Steve Taylor:
> That technique works fine for wide angle and short focal length lenses, but
> is a lot more complicated with a 100mm mm lens with at f32 0r f22 with a
> minimum depth of field focus of 5m+ to get infinity sharp.

Why should it be more complicated? You obviously need two features which
are (more or less) inside the depth of field, but apart from this it's
just the same as for wide angle lenses.

Or the other way round: If you don't see parallax in live view at 100%
zoom you won't see it in the resulting panorama either.

--
Erik Krause
http://www.erik-krause.de
Reply all
Reply to author
Forward
0 new messages