Panoramas

[neonlamp]

The article on panorama images describing a new camera feature was interesting. I haven't the luxury of a new camera with that option, but use old-style stitching. To ensure that images blend correctly it is important to eliminate parallax errors. In order to do that, the centre of the panning circle has to be positioned at the front of the lens, not on the camera film plane which would be the case for most situations if a camera is mounted on a tripod with its standard mounting screw.
I would be interested to know how Canon's panorama mode deals with that.
To eliminate parallax errors in the old style stitching method I use a slide rail which can set the camera back, and it might have been worthwhile mentioning whether a slide rail would be useful with the new cameras.

[Farider]

I haven't used the panorama option on my R6 yet but I do have it on a DLSR from my pre Canon days and I have used it there, albeit a long time ago. It worked well as I recall and gave a good result. Can't comment on how it was achieved though!

My more recent ones with the Canon cameras have been individual shots stitched together in Photoshop and parallax doesn't appear to have been an issue.

Ian

[John Liddle]

In order to do that, the centre of the panning circle has to be positioned at the front of the lens, not on the camera film plane which would be the case for most situations if a camera is mounted on a tripod with its standard mounting screw.

Not necessarily at the front of the lens!

The camera needs to be rotated about the point which minimises (not necessarily eliminates) parallax errors. This point seems to go under a number of different names - "Nodal point" being one, but I think it is more correctly the "Entrance pupil".

Lots of videos etc. on the web to help you find it - it will be different for every lens.

A very useful database here https://wiki.panotools.org/Entrance_Pupil_Database which covers a lot of the Canon lenses (amongst others).

[neonlamp]

That is a useful reference, thanks.
As a rule of thumb, though, I suspect pivoting around the front of the lens will be more accurate than from the sensor plane.If you don't have the data to hand.