NN3 MkII & 7d - My lower rail settings are odd

[mitchfx]

After multiple calibration sessions with my 7d and the NN3 MKII I consistently come up with 57-58mm for the lower rail setting. Everyone else here and on other boards say that 53mm should be right, but if I try that the center of rotation on the rotator target is WAY off. Could I possibly have something assembled incorrectly or maybe doing something else wrong?

Thanks,
Mitch

[hindenhaag]

Hi Mitch,

What camera are you using? Does the long side of the vertical rail's foot show to rotator? Are you using canon 5D II?

Heinz

[mitchfx]

As I said in my op it's a canon 7d. Yes, the "deeper" foot of vertical rail is facing inward towards rotator head.

Thanks

[hindenhaag]

Ups, sorry, missed that early in the morning.

http://wiki.panotools.org/Entrance_Pupil_Database H for 7D is 42 mm in case the colleagues measured right. Just check it on your own again. My NN3 adds 13mm offset to CP-2. This means LRS should be 55mm.

Have a look to this thread: http://www.nodalninja.com/forum/show...r+rail+setting

They way I do it with smooth's tutorial: http://www.easypano.com/forum/displa...PagePosition=1

Heinz

[mitchfx]

Thanks for the reply. I have double checked all the tutorials as well as my settings and keep coming up with 58mm for the lower rail. I've uploaded some images that show my rig as well as the live view (with grid) in various degrees of rotation about the head. Keep in mind the upper rail has not been set here, only the lower (upper is at an arbitrary position in these photos)...

https://plus.google.com/photos/10996...KDJv8_jjOTnyAE

I've also uploaded a video showing the camera rotating around...

http://youtu.be/ZQF2AHLwEcA

I guess as long as it works correctly there really isn't a problem, but its just a bit confusing that everyone else gets a lower rail position that is SO much different with the same setup. Is there any chance that the camera plate that came with my NN3 MKII kit is slightly different?

[John Houghton]

Mitch, Far more useful would be two actual camera shots of the nadir taken with the camera pointing vertically down, incrementing the yaw setting of the head by 180 degrees between them. But it doesn't much matter what settings anybody else is using. It's good that you have taken the trouble to set up your head carefully yourself, and the proof of the pudding is in the eating. If your settings work well for you then be happy and shoot some panoramas. I have only ever used my own settings and I really don't care what settings anybody else comes up with.
.
John

[mitchfx]

I just posted two photos at 180deg. (last two in same gallery as others)

https://plus.google.com/u/0/photos/1...KDJv8_jjOTnyAE

https://plus.google.com/u/0/photos/1...KDJv8_jjOTnyAE

Thanks,
Mitch

Mitch, Far more useful would be two actual camera shots of the nadir taken with the camera pointing vertically down, incrementing the yaw setting of the head by 180 degrees between them. But it doesn't much matter what settings anybody else is using. It's good that you have taken the trouble to set up your head carefully yourself, and the proof of the pudding is in the eating. If your settings work well for you then be happy and shoot some panoramas. I have only ever used my own settings and I really don't care what settings anybody else comes up with.
.
John

[John Houghton]

Mitch, Thanks for posting the images. When aligned with PTGui, they reveal slight parallax shift, suggesting that the lower rail setting is off by maybe 1-2mm or so:



John

[mitchfx]

John,

Thanks for checking this out. What is your process for test aligning in PTGui. I have the software as well and would like to replicate what you are doing exactly so I can properly calibrate.

-M

[John Houghton]

Mitch, The process for aligning the nadir images was no more than generating a stitched image in layered PSD format from the complete project, but with only the two nadir images selected for inclusion in the output. I applied a -90 pitch adjustment via the numerical transform option to bring the nadir to the centre of the output area and adjusted the fov to 50x50 degrees and rectilinear projection. I created an animated gif file from the psd file in Photoshop for the purpose of displaying the result here, but you can simply open the psd file in Photoshop and toggle the top layer on/off to get the same effect, or you can adjust the opacity of the top layer to 50% to compare the positions of the head.
.
John

[mitchfx]

John,

I've uploaded two more sets of calib photos in the same google gallery, one at 57 and one at 56 (images named accordingly). From my own tests in ptgui it look as if 57 is closer than 56 but I'd love a second opinion if you have the time. There is still some offset in the opposite axis with both pairs but I think this is due to a tiny amount of rotation of the camera on the mount.

Thanks,
Mitch

[John Houghton]

Mitch, I don't know why the two images are in different orientations in each pair. Anyway, the 56mm setting is clearly much better than the 57mm one. The offset is about 0.13mm compared to 0.75mm.
.
John