I feel it is only fair to warn you that this is a bit of speculation that is itself born from speculation. It's like you took something that "might be a seed" and planted it in something that "might be dirt" and this is resulting harvest. Bon Appetit.
Side Note: Oh Christ, I'm writing a side note already...
It is well worth mentioning that marketing departments can often be responsible for inflating expectations way out of line with reality. "Great Technological Improvements" might well refer only to an aesthetic redesign and better battery life and screw you Jack for thinking it was anything more than that.
A few days back HTC suffered the dreaded:"We really should not have given this to our web developer" problem. A few images were dug up of what looked to be the promised final version of the HTC Vive "VK2" developer kit. (and retail as well right? Nobody iterates THAT fast, there can't be a THIRD version, can there???).
These types of leaks happen so often in technology, if you were the paranoid type, you might think they were done on purpose.
Marketing Department: "Should we leak photos AHEAD of the holidays or AFTER?"
You'll recall that Oculus had the same kind of thing happen to them back in the summer of this year. And guess what? Those images ended up NOT being representative of their final product. Same thing could happen here... but that's no fun.
Let's brazenly (and conveniently) assume that these ARE official images of the final product and then try to coax that assumption to mate with our expectation of "Great Technological Breakthrough" and see what kind of offspring result.
It's neat that the leaked images of the Oculus Rift CV1 feature a HMD mounted camera in a very similar location shown in these recent HTC leaks. I think it indicates that both companies have been playing around with the concept and see value in it.
Or at least temporary value... since it was pulled from Oculus's final design.
It hasn't been made terrible clear what you would use it for. In fact, I don't recall Oculus or Valve ever officially commenting on what a camera might be good for. (please feel free to send me a note if you HAVE heard something)
"I could turn on the camera and take a peek at the real world in order to take a sip of my coffee/coke/beer/insert favorite beverage here."
Yes.... drinking and eating... using a keyboard... all fine, practical stuff but let's try to stretch a bit further, shall we?
"Wait!? The camera isn't stereoscopic!?"
You're right! The camera isn't stereoscopic. If you need to confirm this, just count the cameras. 1....
"That's terrible!!!! I though this was VR!?!?!? BLARRARRRG!"
I have a good hunch that monoscopic (not really a word apparently...) is fine for a lot of VR applications as long as it isn't the focal point and it might REALLY POTENTIALLY BE AWESOME for something...
Here's a slide the covers the specs for the Vive Developer kit (and is a good representation for commercial release as far as we know.) You might want to flip back to this image from time to time to confirm things as I keep rambling.
You'll note that the HTC Vive works by displaying stereo images on two separate panels giving each eye 1080X1200 of resolution. The image displayed on the HMD is then presented to the wearer as shown below, circular view portals bordered by black.
The camera displays a mono camera image unified across both panels at a resolution of 2160X1200 at 90hz and what if they also dispense with the black borders giving the user a feeling that their FOV is even larger than before?
What if they then overlaid that wide crisp camera view with stereo 3D objects projected at room scale, tracked to sub-milimeter accuarcy?
Note: It is worth pointing out that you might find yourself rather entranced watching a FPV camera view at 2160X1200 at 90 fps. This isn't you've likely ever enountered before and something quite novel in itself.
Note: I might also want to point out that the current iPhone 6S can swing 1080p at 120FPS while it is also handling encoding and write functions so...
A quick example of this would be Tiltbrush. You could either be drawing in a drab grey plane or you could be painting in your room.
Try to guess which one will appeal more to consumers?
Note: TiltBrush is awesome and the grey background is required for proper contrast with the artwork. If you asked me about this I would expect that you could dial in the brightness at which your video feed was displayed to allow you to enjoy your room in dreary grey, whenever it was called for.
Here's a quick summary:
Note: It is worth mentioning that this setup would also support multiple users in the same room. You could walk around 3D structures while viewing the room and the other users.
Note: In fact, if they AREN'T doing something along these lines then... I'm really puzzled. Lighthouse technology pretty much guarantees that this can't miss as an application. They could execute REALLY REALLY well on this.
Note: One day I am going to write a post entirely composed of notes. I am well on my way.
Note: Let me preface this section by saying this while this is 99% wishful thinking on my part, I ALSO think that this potentially is "the right way to do things" and we'll see this appproach sooner or later from some corner.
If you take a close look at the image that may represent the controllers and hum and haw a bit about the details, you might start to look at how it is put together and then wonder "what if it came apart?" or "What if it was MEANT to come apart?". You'd look at the seams and position of the sensors and realize that there are two distinct components separated by one seam. This one:
We have one portion on the controller dedicated to the buttons and the ring is its own module, housing the sensor array.
You could unscrew/detach the sensor ring and mount it on something else?
This comes to mind because I've noticed some weird things being highlighted by HTC in the last few days. During their Beijing developer conference last week, we were treated to a hack involving finger tracking from Noitom coupled with an HTC vive controller strapped to the users wrist.
HTC Vive has also promoted work from "Manus VR" that shows a simliar approach, this time using HTC Controllers that have been skinned alive. If you have a weak stomach, then you might want to skip this image:
So imagine you could unscrew the ring from the controllers and screw it into a different form factor, maybe a 3rd party controller like a gun or tennis racket or guitar...
A simple strap would allow it to be added to wrists, feet, elbows paving the way for better body tracking.
In short. After thinking about it a bit, I've concluded there should be a complete divorce of sensors from buttons. They really should not be seen in the same room together without their lawyers present. Sensors should be a stand alone unit that plugs into a secondary controller scheme with buttons appropriate to the application ranging from paint brushes to AK47s to surgical tools.
It is important to keep in mind that Lighthouse tracking is terribly efficient and in theory can track as many objects as you care to throw at it.
So perhaps one day...