3 ms·

Has Occulus ever tried electromagnetic motion sensors. Sixense probably did that once, thought they never went into production but they said it was super precis

Has Occulus ever tried electromagnetic motion sensors. Sixense probably did that once, thought they never went into production but they said it was super precise and has very low latency. Occulus is tracking motion with camera which is way more CPU intensive than calculating position based on electromagnetic signals. One trouble they had though that it wouldnt work well with high energy electric items nearby. So dooes anyone know if Occulus or HTC vive or other well in market consumer VR device ever tried that or has plans to do so? is this even possible? EDIT: They also made controllers for Razer. https://support.razer.com/console/razer-hydra/ https://support.razer.com/console/razer-hydra/

jackbrookes 6y ago
Magic Leap's controller uses this technology, and as far as I am aware most people said it has inferior tracking to Oculus/HTC/Valve's approaches. https://developer.magicleap.com/en-us/learn/guides/design-magic-leap-one-control https://developer.magicleap.com/en-us/learn/guides/design-ma...
GuB-42 6y ago
I think that the goal for Oculus is to do everything using headset-mounted cameras, without the need for a base station, so it makes sense for them to focus on image processing instead of specialized sensors. Cost also matters, especially for a $300 standalone headset, and extra sensors aren't free. One thing to realize is also that cameras (or lighthouses in the case of the Vive) are not the primary sensors. What really do the job are accelerometers and gyroscopes. Cameras only provide reference points to correct for drift.
- Robotbeat 6y ago
  Cameras are also nice for enabling pass-through views. Works much better than you might think on the Quest. My phone hasn't been working the last few days, so I've been watching videos (etc) in the Oculus browser with the passthrough environment, and I can walk around a bit and do stuff (brush teeth, etc) while watching a video on a large screen, and it's relatively practical (hand tracking is still annoying and largely 2D so it's still better to use a controller), but this could be improved with software updates and the improved Quest 2. The view is stereoscopic and so eye-hand coordination works basically flawlessly (much better than I thought it would). It'd be sweet if someone eventually made a headset with pass-through view with high resolution and ability to choose wavelength (UV to IR, maybe even a phased audio array to visualize sound sources for the deaf). If you could make the tracking cameras work in the Sun and figured out a way to harden the display (perhaps using a protective electronic shutter system) to sunlight, it'd be closer to a general platform.
the_hoser 6y ago
The problem with magnetic tracking is distance. Unless the motion you're tracking is in a relatively small space, you need a terribly powerful magnet to overcome background EMI, and that introduces problems of its own. Adding compute power to compensate is a much easier problem to solve.
numpad0 6y ago
Someone said you can’t have a PC in the same room at all while magnetometer based tracking is operating, so...
- kevin_thibedeau 6y ago
  Wouldn't be a problem if homebrew gamer PCs were FCC compliant and didn't have gaping RF holes in them.
TwoBit 6y ago
Yes they did, and it didn't work well enough.

jackbrookes 6y ago

Magic Leap's controller uses this technology, and as far as I am aware most people said it has inferior tracking to Oculus/HTC/Valve's approaches. https://developer.magicleap.com/en-us/learn/guides/design-magic-leap-one-control https://developer.magicleap.com/en-us/learn/guides/design-ma...

GuB-42 6y ago

I think that the goal for Oculus is to do everything using headset-mounted cameras, without the need for a base station, so it makes sense for them to focus on image processing instead of specialized sensors. Cost also matters, especially for a $300 standalone headset, and extra sensors aren't free. One thing to realize is also that cameras (or lighthouses in the case of the Vive) are not the primary sensors. What really do the job are accelerometers and gyroscopes. Cameras only provide reference points to correct for drift.

Robotbeat 6y ago
Cameras are also nice for enabling pass-through views. Works much better than you might think on the Quest. My phone hasn't been working the last few days, so I've been watching videos (etc) in the Oculus browser with the passthrough environment, and I can walk around a bit and do stuff (brush teeth, etc) while watching a video on a large screen, and it's relatively practical (hand tracking is still annoying and largely 2D so it's still better to use a controller), but this could be improved with software updates and the improved Quest 2. The view is stereoscopic and so eye-hand coordination works basically flawlessly (much better than I thought it would). It'd be sweet if someone eventually made a headset with pass-through view with high resolution and ability to choose wavelength (UV to IR, maybe even a phased audio array to visualize sound sources for the deaf). If you could make the tracking cameras work in the Sun and figured out a way to harden the display (perhaps using a protective electronic shutter system) to sunlight, it'd be closer to a general platform.

the_hoser 6y ago

The problem with magnetic tracking is distance. Unless the motion you're tracking is in a relatively small space, you need a terribly powerful magnet to overcome background EMI, and that introduces problems of its own. Adding compute power to compensate is a much easier problem to solve.

numpad0 6y ago

Someone said you can’t have a PC in the same room at all while magnetometer based tracking is operating, so...

kevin_thibedeau 6y ago
Wouldn't be a problem if homebrew gamer PCs were FCC compliant and didn't have gaping RF holes in them.

TwoBit 6y ago

Yes they did, and it didn't work well enough.