A7R4 lens / imaging sensor shifted optical axis?

[aroakley]

Ok, so here's a weird question / observation!

I shoot 360's panoramas and am transitioning from a Canon 5D Mk3 to my new Sony A7R4.

Whilst trying to set up the parallax adjustment, I've noticed something odd which is making adjustment problematic. The lens optical axis seems to be shifted from the camera imaging axis by several mm. Referring to the attached photo, if I align using Live view so that the circle of the bubble is exactly centre full frame, the physical lens axis isn't aligned vertically above the physical bubble. Looking at the attached photo, it's apparent that the lens is slightly to the right of being directly above the actual bubble disk centre. I've checked all the mounting plates are square and true and that the camera magnification rectangle is centred. The lens is a Laowa 12mm F2.8 Zero-D EF mount  Adapter is a Metabones V. Lens physical centre axis does not align with camera imaging axis

The problem is that when the camera is rotated 90deg to point horizontally, the optical centre of the lens will orbit the physical centre instead of being perfectly aligned above it. This degrades stitching performance.

Am I going crazy or does anyone know whether this mirrorless design does have an off-axis sensor....or could be a characteristic of this particular lens? Having done some optical engineering in a previous life, I know that it's not an impossibility....but I'd be surprised.

I guess the only way to answer this is to take the lens off and try and determine whether the sensor is exactly in the centre....unless anyone can offer some insight?

Hmmm, I might be fooling myself. It's less than I thought.  It's about 3mm. That would only take 1.3 degrees of camera angle misalignment to produce that offset over the 13cm between lens and bubble plate. Although the measurement of the mounting plate and camera look bang on, 1.3 deg isn't much.

[aroakley]

UPDATE!

So, there is a definite shift with the Laowa. I'm not going crazy. I put on a 50mm lens and aligned using Live View. The physical and optical axis were perfectly aligned as the capture below shows (blurred as I'm too close to focus the bubble plate)50mm

With the 50mm in perfect alignment, I put back on the 12mm Laowa and....a 4mm offset appeared as before!

So in conclusion, for some reason, the Laowa 12mm optical axis is 4mm shifted from the physical centre axis.

[IamNic]

Hello @aroakley,

for me it looks like the white "center" line is off. While it might be center to the lens hood, it is not to the lens itself from my point of view.

I am currently not in front of my computer, otherwise I would measure in Affinity Photo, but just from looking at it, the line seems off.

That is the best I can do right now:

- Nic

[aroakley]

Hi Nic,

Thanks for your post. The white centre line actually represented the centre of the tripod bubble and appeared to be shifted by 4mm from the  lens centre when Live View indicated optical alignment with the bubble centre.. The red line is the physical centre of the lens.

Having done further tests, I'm convinced it's a characteristic of the Laowa 12mm. When I replaced it with a 50mm lens, the physical lens centre alignment with the bubble centre exactly matched the Live View optical alignment. Without moving anything and reattaching the 12mm, the 4mm divergence re-appeared.  

Andrew

[IamNic]

Hello @aroakley,

thank you for clearing that up, seems like I misunderstood your post.

- Nic