Oculus Quest Hands-On: Freedom to Explore VR

hdmark

Distinguished
Feb 16, 2015
1,367
0
19,660
does this have the same out of range issues the windows MR sets have for the controllers? or do the extra cameras expand the FOV enough to cover a bigger range ?
 
You definitely need to clean up around your house. I think I might even see some cobwebs. >_>


I suspect that the extra cameras should improve on that. It looks like they cover a much wider combined field of view, likely covering everywhere except completely behind the wearer reasonably well. Though I suppose if you turned your head to one side, that could result in a hand on the other side not getting tracked properly in some cases.
 

bit_user

Polypheme
Ambassador

You probably know this, but that photo is clearly taken at a staged event. I guess it's themed around this demo:
... as I found out to creepy effect during Face Your Fears 2. I was quickly unnerved by the creaking doors, hapless screams and gentle skittering of legions of deadly spiders threatening to feast on my corpse.


They could still do coarse tracking of unseen controllers with an embedded IMU.
 

Joe Black

Honorable
Jul 3, 2013
88
0
10,640
Very impressive. Although I still feel that they should split the headset from the computational device. The headset should be as comfortable as possible. There are parts of the human body much better suited to carrying weight, and discomfort,than the human head. As small comfortable backpack could pack a lot more horsepower than this ever could. But on the other hand I suppose it's a worthy goal in the long run to aim for a lightweight, comfortable and powerful headset only solution.
 

The writer holds staged events at her house? All the more reason to tidy things up a bit! : D


I'm sure they do, but so do the Windows Mixed Reality headsets, and those are said to have tracking issues at times when operating outside of the field of view of the headset's cameras, where the tracking can become unreliable after some seconds. Something like pulling back an arrow and trying to aim it precisely with a hand over the shoulder might potentially cause issues, though it's also possible that better software interpretation of the data could maintain somewhat more reliable tracking in such situations. And in this case, the additional cameras would undoubtedly help a lot at increasing the size of the area that is fully-tracked.
 

bit_user

Polypheme
Ambassador

I definitely agree. This is the approach taken by Magic Leap, although it was more of a necessity for them.

Here's a thought - if the HMD supported some sort of wireless connectivity standard, then you could use the same HMD both with a PC and with a hip-pack or backpack compute unit. Maybe high-end cell phones would even pickup support for the wireless standard.

Anyway, the main downside would be more cost, if building two units + a connection cable. But the benefits would include more battery life, more comfort (obviously), and more horsepower (i.e. the compute unit could generate more heat and accommodate better cooling).
 

bit_user

Polypheme
Ambassador

True, there are cases where the hands could be out of view of the HMD for longer periods. Seconds is long enough to start getting drift.


Maybe they'll devise some clever deep learning-based calibration scheme. It could use online learning to automatically calibrate the drift of each IMU, as it's being used. It shouldn't take a very big network.


Also, this. Maybe even just one rear fisheye camera (which could also warn against backing into physical obstacles).