Capturing Reality Camera Rig

Development Asset Creation
1

Hi Everyone,

After a quick community post search, it looks like there were plans for fixed camera rigs to be implemented into RC but I am unable to find if it ever came to be?

We are working with a Ladybug 5 (See below) which is an extremely rigid camera rig comprised of 5 sensors.

https://www.flir.com/products/ladybug5plus/ 

In ContextCapture we are able to create a relative rig whereby the orientation and position of each photo of a “Camera Event” are known and held fixed, while the absolute position and orientation of the block can be adjusted.

Looking forward to hearing if this is a current feature of one which is planning on be implemented in the near future.

Cheers,

John

1 Like
2

Hi EvenPheven,

it is still in our feature request database. I will add your vote there.

In some cases it should be possible using XMPs data (https://support.capturingreality.com/hc/en-us/community/posts/360007680199-Rigging-and-XMP-Command-line). XMP file defines the rig and also which image is in concrete rigs slot. These files have to be created manually.

If your cameras will have GPS, you can also use the flight log for this, where each camera will have its position it real time.

 

3

375471859899 did you get this to work in RC?

I curently am staring at a dataset from a ladybug camera.

It’s unprocessed 6 images per position and per position there is a csv avaiable with x y x , orientation and accuracy data.

374823146340 what is the status on supporting a Rig like this in RC? 

4

Hi OldRaven, there weren’t changes in this since. I will add your vote into a feature request.

5

Hi,
Does this feature still on the list of request ?

I have a stereocamera system where I know exactly the position between both cameras and I would like to use it to scale my models.

Thanks,

6

Hi MissSwing,
yes, this is still a feature request.
But it should be doable using XMPs. Unfortunately, we don’t have an example of this exact process, but you can find more about XMPs in the application’s help.

1 Like
7

Is this still a feature request? I have bundles of images split from 360 camera frames and would like to tell Reality Capture (now Reality Scan) that the sub-frames of each bundle should all have the same origin.

1 Like
8

Hello @kotavy
Yes, this is still a feature request.

9

Hi team, I am using a modified Qoocam3Ultra (not in the lens database) 3D180VR camera that shoots 2x 6944x6944 stereoscopic images. I can create individual Left and Right images from what is originally LRSBS 2:1 equirectangular. Seeing that the original stereoscopic frame is not recognized, I’d like to import the individual images (from an interval sequence) as a ‘camera rig’ and define the known horizontal offset of the camera (IPD) so that I can use those camera positions for 3D Gaussian Splatting reconstruction. Camera is 1.7mm but not sure what distortion model to use either for a 3D180. Any ideas as how to make this happen? Have read the user manual but cannot work it out, can’t see YT tuts either.

Your help is appreciated

10

Hello @AVsupport
The only way here for now is to use XMP workflow (Assigning Set (Static) Camera Locations - For Multi-Camera Photogrammetry Rig). But this wasn’t tested internally.
This is the page Re-use Alignment.

As the camera is fish-eye, I would use Division distortion model.

11

Hi @otrhan thanks for the tips, here’s a sample image (I couldn’t upload earlier as noob)
PS, not sure if I mentioned this, the camera is moving, not static, as it rotates around the object. The ‘rig’ is the 2 lenses of the camera

Screenshot 2025-07-15 132621
Screenshot 2025-07-15 132621796×750 86.5 KB

Screenshot 2025-07-15 103126
Screenshot 2025-07-15 1031261827×1860 504 KB

12

@AVsupport
This could be a problematic dataset for RealityScan, as the distortion is quite big there. Also, the capturing from one spot is not ideal: https://www.youtube.com/watch?v=NYdHk9PhS6o&pp=0gcJCc0JAYcqIYzv