Uploaded May 2021 | Updated September 2026, 12 minutes ago
Very Important when setting up SteamVR to use both an Oculus Quest and Basestations you MUST set "activateMultipleDrivers": true in: SteamVR\resources\settings\default.vrsettings
When I first received my Quest 2 I tested with SteamVR trackers hoping it would be simple to setup full body tracking. I realized quickly that was not the case and that the Quest 2 and SteamVR playspaces would not agree and were totally out of alignment from one another. After research I could not find any convenient or built-in solutions for re-aligning the spaces back together.
The technique I applied for calibrating the space offsets was something I picked up from a machine learning course at Stanford. Using Eigendecomposition for minimizing the sum of squares error, the alignment of the two playspaces can be calibrated in seconds using purely best-fit approximation and eliminating the need for the user to have to set any offsets manually. Integration occurs as the player moves the controller-and-tracker together as a single unit. Thus for the user to calibrate the space they need only move the controller-and-tracker together allowing the regressor to find a "best-fit" which aligns the two spaces back together. Simple to use, but not to explain!
A decent calibration can be achieved very quickly using this method, needing only integrate over a few feet to make a close approximation. But the approximation continues to improve with larger sweeps.
And was the most elegant (eg. least time consuming) approach I was able to think of. And eliminates the need for adjusting manual dials or entering offset values manually after each time the Quest 2 origin changes after the HMD enters sleep mode!!
---
This feature is still in development and is currently available in the latest beta v3.5.8. Really hoping this helps many Quest users align their basestations more quickly and create better full body motion capture! Please let me know how you think of the new features! =D
Please join the Discord for more info on the project:
discord.com/invite/ErZcKaQ
Thanks for watching!
Very Important when setting up SteamVR to use both an Oculus Quest and Basestations you MUST set "activateMultipleDrivers": true in: SteamVR\resources\settings\default.vrsettings
When I first received my Quest 2 I tested with SteamVR trackers hoping it would be simple to setup full body tracking. I realized quickly that was not the case and that the Quest 2 and SteamVR playspaces would not agree and were totally out of alignment from one another. After research I could not find any convenient or built-in solutions for re-aligning the spaces back together.
The technique I applied for calibrating the space offsets was something I picked up from a machine learning course at Stanford. Using Eigendecomposition for minimizing the sum of squares error, the alignment of the two playspaces can be calibrated in seconds using purely best-fit approximation and eliminating the need for the user to have to set any offsets manually. Integration occurs as the player moves the controller-and-tracker together as a single unit. Thus for the user to calibrate the space they need only move the controller-and-tracker together allowing the regressor to find a "best-fit" which aligns the two spaces back together. Simple to use, but not to explain!
A decent calibration can be achieved very quickly using this method, needing only integrate over a few feet to make a close approximation. But the approximation continues to improve with larger sweeps.
And was the most elegant (eg. least time consuming) approach I was able to think of. And eliminates the need for adjusting manual dials or entering offset values manually after each time the Quest 2 origin changes after the HMD enters sleep mode!!
---
This feature is still in development and is currently available in the latest beta v3.5.8. Really hoping this helps many Quest users align their basestations more quickly and create better full body motion capture! Please let me know how you think of the new features! =D
Please join the Discord for more info on the project:
discord.com/invite/ErZcKaQ
Thanks for watching!
