Where's our Oculus Hand Tracker?

Completely agree, but this has been discussed to death before. Here for instance: ;-)
https://developer.oculusvr.com/forums/viewtopic.php?f=25&t=5912

Anyway, i'm still pretty optimistic that we will see something by next CES or GDC. For now developing using my Hydra!

Worst case scenario we'll have to rely on Sixense, which is not too bad; i'm quite confident that their consumer product (or other companies products based on their technology) will be much more affordable than their current Dev Kit pre-order prices.

Are you sure those are Dev Kit prices? I got the impression $300 was their consumer release. It doesn't say Developer version anywhere on their pre-order page.

Still, couldn't hurt to keep the topic in the eyes of Oculus. Wouldn't want them to think demand for a cheap solution has disappeared and we're completely cool with being robbed by Sixence. $300 is absurd. A PS Move kit is less then $100.

"PatimPatam" wrote:
Completely agree, but this has been discussed to death before. Here for instance: ;-)
https://developer.oculusvr.com/forums/viewtopic.php?f=25&t=5912

Anyway, i'm still pretty optimistic that we will see something by next CES or GDC. For now developing using my Hydra!

Worst case scenario we'll have to rely on Sixense, which is not too bad; i'm quite confident that their consumer product (or other companies products based on their technology) will be much more affordable than their current Dev Kit pre-order prices.

Good thread, Patim, I see we're of similar mind on the controller setup. It seems so straightforward! How can Oculus be slacking off on this important item?!

After a few dozens hours on the DK2, the biggest problem with it isn't the resolution or the framerate, anymore. Both of those are at good/acceptable levels. But there is NO HAND TRACKING!!! That's not a minor issue, that's a HUGE ISSUE! We need hand tracking desperately!

Also, as Tbone pointed out in your older thread, even if Sixense gets off their butt, releases the STEM, and even gets it's price down to something that isn't STUPID-EXPENSIVE, the computer will *STILL* will have to merge hand tracking and head tracking data from both systems. That's two positional locations, from two completely different sensors.

If Oculus were to make a hand controler, they could EASILY keep track of the hands, relative to the head. Adding 2 more position/rotation points should be easy for the camera to handle, and feeding all of that data down one pipe would make all of the programmer's lives *MUCH* easier! Why torture the people who are making the games with two separate tracking pipelines??

Patim, you're using both the Hydra tracking and DK2 tracking. Wouldn't it be vastly easier if the Oculus drivers could handle all of that in one place?

"MaxSMoke777" wrote:
Outside of *maybe* the LEAP

What's wrong with the Leap? I was considering ordering one along with the Rift mount kit to try out some ideas. I know only a few current demos support it, but for development it seems like an affordable and intuitive way to use your hands in VR.

(Don't get me wrong, I'm not disagreeing with you about Oculus coming up with their own solution, just curious about your dismissal of the Leap.)

The Leap only works in a small box worth of space, directly in over the top of it. I have actually advocated the use of the Leap in the past, and was the first person, I'm aware of, that suggested putting a Leap on the front of the Rift itself. I was quite happy to see Leap make an official mounting bracket for the Rift.

But now the Rift uses optical tracking, which might get very obscured by the Leap, and with DK2's camera system, it could be possible to observe a MUCH larger area then the Leap is capable of. Even if the Leap does not obscure the DK2's LED's, it's still a very small box to work with. Also physical buttons add some much needed tactical feedback.

I'd personally love to see some demos using the Leap strapped to the DK2. As of yet, I haven't seen anything released, just some videos on YouTube. It might still turn out to be a great thing. I just think that optically tracked joysticks just have a greater potential.

I really am curious what you come up with in your own Leap experimentation though. Btw, the programmer who did take an early interest in my suggestion found that he got the best results by angling his Leap slightly downward. I don't think the Leap's official mounting bracket does that, so I suggest you make some modifications to do so.

I also made some suggestions which I don't think anyone has tried yet for movement using the Leap. I suggested using a pinch-zoom to drag the player new locations. For instances, bring together the two pointing fingers of your hand and where you pointed with both would mark a location. Then pull the fingers apart to zoom over to that new location.

The two fingers together would create an invisible ray based on their orientation to your vision. The impact point of that ray would create a visible marker so the player knows where they are going. Pulling the fingers slowly apart would decrease the distance to that marker. Closing the hands or moving the fingers apart quickly would break the marker. It would feel like a grappling hook, pulling you around. No joystick input required.

"MaxSMoke777" wrote:
Are you sure those are Dev Kit prices? I got the impression $300 was their consumer release. It doesn't say Developer version anywhere on their pre-order page.

Well i'm pretty sure i read a few times that their current prices are due to being mainly aimed at developers for now, and that it will come down, but sorry can't find a link at the moment.. maybe they were just Reddit rumors.. :-P

"MaxSMoke777" wrote:
Patim, you're using both the Hydra tracking and DK2 tracking. Wouldn't it be vastly easier if the Oculus drivers could handle all of that in one place?

Yep i know, you don't have to convince me! :-)

The way Sixense is going around this is by having a 3rd STEM tracker on the head that serves as a "link" between both tracking systems (or by entirely ignoring Oculus' positional tracking i'm not 100% sure), you can see they use it in all their latest "lightsaber" demos. Either way i think this is less than ideal for a consumer product, and we don't even know if the Rift CV1 will have a strap where you could easily attach a STEM pack.

"SnottyBison" wrote:
What's wrong with the Leap? I was considering ordering one along with the Rift mount kit to try out some ideas. I know only a few current demos support it, but for development it seems like an affordable and intuitive way to use your hands in VR.

Yeah i had a back and forth with one of the Leap developers at the end of the thread pointed before.. i think it could be a cool option in some cases, but as a general input device i believe a motion controller would be much better.

Here are some of the problems i see with the Leap Motion:

- Unreliable: it doesn't work as you expect a 100% of the time, there are problems and glitches tracking the fingers when you can only see the side of your hand for example (which is understandable). They say it will improve, but not sure how much it can realistically improve.

- Latency: they also say they are going to improve it, but in my experience it's still way too high.

- Small FOV: even if it's slightly bigger than the Rift's FOV, would not work for simulating some type of interactions like sword fighting or tennis for instance.

- Lack of physical feedback: simple things like holding a virtual gun would feel very gimmicky.

- Interference with Rift tracking: having both hands in front of your face will block the LEDs of the HMD and stop the positional tracking, certain interactions like drawing a bow or aiming a sniper rifle require you to do so.

@MaxSMoke777 There are a number of Leap VR demos hosted on our developer gallery, and lots more out in the wild, especially with the 3D Jam currently underway. The ones you can find in the gallery include our VR Intro, Collider VR, and Plasma Ball VR.

It's important to note that today's Leap Motion Controller is not what we envision for consumer-level VR/AR. Far from it! This is an experimental implementation of existing hardware that opens up a range of possibilities within the peripheral's current limitations -- with UX principles, assets, and APIs that carry forward to Dragonfly, our embeddable prototype. Dragonfly is designed for OEMs to embed into their hardware and features significantly expanded FOV and greater-than-HD camera resolution, as well as full-color passthrough.

As for mounting the Leap on a downwards angle, we definitely considered that during the design process. But we believe that optical hand tracking and augmented reality (if you'll pardon the pun) go hand-in-hand, and based on what we've heard from the community, there's massive interest in that direction. Even if you don't go in for AR, there's the "wanna-Coke" scenario -- meaning the ability to toggle between real and virtual worlds without taking off the headset, so you can grab a drink (or quickly escape from a survival horror game).

Even if a lower angle presented some slight tracking benefits for the current generation of Leap hardware, it would make any significant developments in AR untenable. You'd always feel disoriented because you'd be looking at a downward angle. For these reasons, we went with 90 degrees as a standard.

I have been thinking about it quite a lot and wouldn't it make sense for oculus to stop using an screen mounter camera and track everything including legs, hands, head, positional tracking and all you want with something like a couple of leap motions with two 180º fisheye lens mounted at the front and back of the hmd?

A couple of leds over the screen kind of like a wii bar would allow for easy positional tracking or maybe even no leds, just draw something in the screen corners.
It could also warn you when you are going to walk into a wall.
Being 360º it could track hands and legs everywhere.
It would not only work for AR but it would also serve as an awesome 360º video camera.

Project PERCEPTION NEURON

https://www.kickstarter.com/projects/1663270989/project-perception-neuron

"Pickman" wrote:
two 180º fisheye lens

Lenses don't work like that.... :p