Multi Sensor Alpha

Development on Multi Sensor has been abandoned in favor of a complete rewrite for the v3 branch.

Which is being redesigned with support for multiple sensors from different brands.

Brekel Body v3 is currently in beta for all customers, PointCloud v3 will follow after.




The info below regarding Multi Sensor – Alpha is here for historic purposes:

connects to instances of any v2 application
offers calibration tools to align multiple sensors to share the same coordinate system
record pointcloud data from multiple sensors
fuse skeleton data from multiple sensors (not yet)
fill occlusions by placing multiple sensors around a capture volume from different angles
extend your capture volume by placing multiple sensors next to each other (with a bit of overlap)



Frequently Asked Questions:

  • So I need a Pro PointCloud/Body/Face v2 license right?

Yes, you will run the base applications as normal, in GUI or headless mode, to access the sensor’s data and run Multi Sensor as a master control.


  • How many licenses will I need?

You will need 1 license for Multi Sensor Calibrator.
You will need 1 license for Pro PointCloud v2 in case you want to capture (and convert) pointclouds.
You will NOT need additional licenses for additional sensors, you can simply install your Pro PointCloud license on multiple machines.


  • But the alpha/beta of Multi Sensor Calibrator is free right?

Yes, you get free early access to the software in order to test and help improve it.


  • How many v2 sensors can I connect to a single machine?

Due to PCI-Express bandwidth limitations the Microsoft’s drivers only support a single sensor per machine.
Note that on (highend & expensive) hardware that would theoretically support multiple sensors there is still the Kinect driver/SDK limitation.


  • Which sensors are supported?

Only Kinect v2 at the moment (Orbbec and Intel RealSense are not supported for Multi-Sensor)


  • So I will need multiple machines right?

Yes, you will need one machine per sensor.


  • What is the cheapest/smallest machine you recommend?

I have been getting good results with the Intel NUC i5 models when running the applications in headless mode.


  • How is PointCloud data supported?

The Multi Sensor app will send signals to all Pro PointCloud v2 apps to record at the same time to ensure best performance.
You can then process the individual BPC files as you normally would and things should align based on the calibration.


  • How is Body data supported?

At the moment this is not supported yet.


  • Is Pro Face v2 supported?

At the moment this is not supported yet.


  • How many sensors can be used simultaneously?

There is no real limit to the amount of sensors you can use. (well internally 32 communication ports are reserved)
In practice I’ve tested with 2, 3 and 4 sensors.


  • What does Alpha mean?

It means not all features have been implemented, there will be bugs, crashes and things like workflow, UI and implemented features are likely to change before official release.
It also means you can have early (free) access and can help find bugs and improve by making suggestions and giving feedback.


  • Why is this currently for existing license owners only?

This is complex software which takes a lot of time to create.
By buying or owning a license you will help me dedicate time for development.
Limiting to existing license owners (for now) helps focus support to people that are already familiar with the single-sensor applications.


  • Does this, or will it, support v1 sensors/software?

For the moment only the v2 sensors/software are supported, there are no plans yet for v1 support in the initial version.


  • Will overlapping sensors interfere with each other?

There is a small chance of interference since these consumer sensors don’t have an option to sync their hardware shutters.
In practice interference is usually minimal, can phase in and out and manifests itself as jittering backwards and forwards on overlapping pixels.
Depending on your setup you may be able to play with sensor placement, distance and overlap to minimize interference.
Please share your experiences during alpha/beta testing regarding this!
I’ve also noticed that interference usually happens when the sensors are not warmed up yet, possibly due to the hardware internals performing slightly different than the factory calibration.