Squaring up camera so camera is perpendicular to central axis of subject

[rogerxnz]

The full subject of this message is: Squaring up camera so camera is perpendicular to central axis of subject and rear left and right edges of subject are sharp.

I want to photo a bird's skull so the rear right and left edges are sharp and then take shots at closer distances coming forward for focus stacking.

I am using a Linhof Techno with a sliding back and a P1 4-150 sensor unit. I have the skull extending over the left and right frames of the sliding back.

I thought I could centre the skull over the frames from the sliding back (so the edges of the skull are the same distance from the edges of their respective frames), note the difference between the distances from the sensor when focusing on each of the right and left edges of the skull, move the lens to the midpoint of these distances and rotate the camera (not sure which way) and, hopefully, the two edges will eventually be in focus.

I have spent hours trying to find which way I should rotate the camera without success and, no matter what I try, getting the two rear edges in focus has eluded me. I know that the right edge of the skull is captured on the left edge of the sensor and vice versa but this has not helped me!

I think swinging the sensor should not be necessary and that all I should have to do is centre the sensor over the midpoint between the rear right and left edges (that is, make each edge the same distance from the external edge of their respective frames) and make the plane of the sensor parallel to a line drawn between the two rear edges.

I also think the fact that the side of the skull is curved should be irrelevant to getting the two edges sharp.

I attach some pictures of my set up.

Any assistance would be gratefully received!
Roger

[Gigi]

Roger -
A bit more info please:
What lens are you using? Aperture?
Distance to subject?
Front and back distance?

This may be a DOF issue (razor thin up close), possibly from field curvature of the lens.  Or could be too much side to side, if you are stitching on the horizontal axis. Or possibly front plane parallelism issue w the Techno.

Geoff 

[rogerxnz]

Hi, Geoff
Thank you for your interest.

I am using a 120mm macro lens. I am focusing wide open. The distance to subject varies but is about 200mm from front of lens. I haven’t measured the other distances.

What I have since found is that by using a ruler to measure the distances from the rear edges of a subject to the sensor unit and rotating the camera to even the distances, I get great results! It’s a pain but it works.

A laser measuring device might solve the difficulty of holding a ruler and not disturbing the subject.

I, therefore, think the concerns you have are not relevant.

What I do not understand is why it does not seem possible to find focus with Liveview when rotating the camera to the correct subject to camera distance, no matter how slowly I turn the knob on the tripod. I know the gearing on the tripod would be coarser than that on the camera focus knob but I would have thought that turning the tripod control slowly would overcome that.
Roger

[Gigi]

If I get this right, you are trying to pivot the camera to get the left and right points on the back of the skull equidistant? That can only work if the camera is located along the right axis (imagine a triangle, whose base is determined by the two L/R  points, the axis is at the midpoint of that line and at 90º). If the camera is not along that axis, then you can only rotate it to get the base line in focus if you first get the camera parallel to the L/R point line, and at the same time, use just the right amount of shift, to get back to the center line...

Again, this is assuming everything in the camera is parallel too, including the front lens plane. Sometimes that can be off. I had a heck of a time with a 35mm lens wide open, never could get sides both in focus. Didn't matter stopped down, tho. 

Hope this helps.

[John Nollendorfs]

Roger:
A trick that I use in copying art to get the camera parallel to the art, is a mirror in the same plane as the art. In the view finder, position the camera so the camera reflection of lens is centered in view finder.

[rogerxnz]

Thanks, Geoff
My camera is mounted on a sideways macro rails, which are mounted on a front to back macro rails, which are mounted on a tripod. So, there is no doubt that, when I rotate the tripod head, the camera is NOT rotating around the tripod centre!

However, I wonder whether this is really much of a problem because, going back to your triangle analogy, in my case, none of the points of the triangle are fixed. I can refocus if the apex changes and I can accomodate changes to the L and R by moving the camera along the sideways macro rails.

Anyway, I have found that measuring the distances between the sides of the subject and the sensor unit tells me which way to rotate the camera and, by rotating in increments, I eventually get them even and the sides in focus at the same focus point.
Roger

If I get this right, you are trying to pivot the camera to get the left and right points on the back of the skull equidistant? That can only work if the camera is located along the right axis (imagine a triangle, whose base is determined by the two L/R  points, the axis is at the midpoint of that line and at 90º). If the camera is not along that axis, then you can only rotate it to get the base line in focus if you first get the camera parallel to the L/R point line, and at the same time, use just the right amount of shift, to get back to the center line...

Again, this is assuming everything in the camera is parallel too, including the front lens plane. Sometimes that can be off. I had a heck of a time with a 35mm lens wide open, never could get sides both in focus. Didn't matter stopped down, tho. 

Hope this helps.

[rogerxnz]

That’s a great idea, John!

I guess you have to place the mirror in the centre of the art? Is that correct?

There are commercially available mirror solutions—see my next post.
Roger

Roger:
A trick that I use in copying art to get the camera parallel to the art, is a mirror in the same plane as the art. In the view finder, position the camera so the camera reflection of lens is centered in view finder.

[rogerxnz]

Info from Bernard Languillier posted in the Medium format forum:

http://www.zig-align.com/

or

https://heritage-digitaltransitions.com/product/dt-laseralign-system/

There alternative laserwise, I don't remember the name of the one I am using.

Cheers,
Bernard

[BAB]

Arrange the Skull in a vertical position looks like the wire holding the Skull are not parallel mounted anyway.
At 200mm deep filling the frame with your sensor you would need to make appx 150-180 images to stack to keep focus rolling around the object until it reaches the back. (As the curve of the object rounds sharply so does the focus fall off sharply in that case you will need smaller steps for those really rounded surfaces)
Then stack in Zerene or Helicon and your done!