Meant to be Seen

I talked to someone who was at the SVVR meetup, they got the impression that you can have 3 STEM trackers working at once.

It wasn't clear if this was in addition to the controllers, or instead of, or if you could have more than 3.

One for each foot, one on the waist, and one in each hand, and the oculus rift as well .. be a pretty thorough motion cap system.

I was also at the SVVR meetup, and the gist I got was that it would be up to 3 STEM trackers, plus the controllers, for a total of 5. However, this wasn't set in stone yet -- the system specs hasn't been finalized yet, as they want to make sure they hit the right balance between cost and capability.

I can see the two extra being sold optionally.

I hope they produce them. It will enable people to produce ODTs without worrying about reinventing the electronics.

Itinerati wrote:I was also at the SVVR meetup, and the gist I got was that it would be up to 3 STEM trackers, plus the controllers, for a total of 5. However, this wasn't set in stone yet -- the system specs hasn't been finalized yet, as they want to make sure they hit the right balance between cost and capability.

I was really hoping that it would be modular, i.e., you could have the controllers and 1x STEM, but with more STEM controllers if desired (I very much desire them).

I was really hoping that it would be modular, i.e., you could have the controllers and 1x STEM, but with more STEM controllers if desired (I very much desire them).

It came up in the conversation -- also, if you look at the rendering closely, you'll see that there are two empty STEM slots on the base. Start with one, but make it capable of three, and let people buy the number they need.

Virtuality managed to generate excellent environments with just two tracked sensors. One on the head and one on a tracked joystick.

Adding two tracked joysticks of course adds more possibilities by they proved it was possible to have very compelling experiences without needing to track the torso, legs or feet.

As mentioned the head and arms/hand were connected by IK links so they always remained in the correct locations relative to each other, I'd imagine that with Unreal's or Unity's IK systems it will be relatively easy to implement these days especially as Virtuality had to do it all by hand.

With regards to patents on the magnetic tracking technology I'd think that Sixense are using techniques formerly used by Polhemus and Ascension on which the protection has now expired and so is freely available to anyone. I'd assume their IP lies in the software.

V8Griff wrote:With regards to patents on the magnetic tracking technology I'd think that Sixense are using techniques formerly used by Polhemus and Ascension on which the protection has now expired and so is freely available to anyone. I'd assume their IP lies in the software.

New patents were obtained for enhancements involving magnetic field modulation, allowing multiple base units to coexist within the same space by using different modulation frequencies. The older tech with expired patents can be used, but is also more limiting in some respects.

It has been awhile since I read and compared the patents, so for more details, I recommend you repeat that process...

geekmaster wrote:

V8Griff wrote:With regards to patents on the magnetic tracking technology I'd think that Sixense are using techniques formerly used by Polhemus and Ascension on which the protection has now expired and so is freely available to anyone. I'd assume their IP lies in the software.

New patents were obtained for enhancements involving magnetic field modulation, allowing multiple base units to coexist within the same space by using different modulation frequencies. The older tech with expired patents can be used, but is also more limiting in some respects.

It has been awhile since I read and compared the patents, so for more details, I recommend you repeat that process...

The 'InsideTrak' cards I bought and pulled from my SD2000 units (1996 vintage) and indeed the earlier 'FastTrak' boards used in the 1000SD (1994 if not based on earlier tech) had separate modules for defining different frequencies so I'd be surprised if separate modules are still under patent, as the limit is 17 years before 1995.

Ascension's SpacePad could control the frequencies in software so that may well be still under patent

It's obvious to me that the Razer Hydra uses cheaper/lower power coils which I assume is cost rather than technical, hopefully the new Sixense dev kits will be higher spec.

V8Griff wrote:... I'd be surprised if separate modules are still under patent, as the limit is 17 years before 1995. ...

Actually, you just triggered my memory about those patents. I think the distinguishing feature of modern patents for magnetic tracking is the calibration circuitry, where a magnetic field is generated by the controller receivers to detect and compensate for local magnetic anomolies. I will need to go back and read them again someday RSN...

Itinerati wrote: if you look at the rendering closely, you'll see that there are two empty STEM slots on the base. Start with one, but make it capable of three,

I did not see that till you pointed it out in your post. I think you are right, that is what that looks like, like two empty stem slots. I hope it is.

One thing not mentioned so far I think, is the possibility that for each additional tracker added, the refresh rate will slow.

Certainly this is how many systems from Polhemus and Ascension work. Not much point having 5 trackers if they are only being tracked at 30hz etc.

Hopefully they're passively measuring the magnetic field rather than oscillating it once for every sensor.

If it's passive then the refresh speeds shouldn't be hurt too much. The limiting factor is definitely measurement over transmission.

zalo wrote:Hopefully they're passively measuring the magnetic field rather than oscillating it once for every sensor.

If it's passive then the refresh speeds shouldn't be hurt too much. The limiting factor is definitely measurement over transmission.

It should use the same technology as the Razer Hydra, which has a four-phase cycle. Each of the three field coils in the base is activated, followed by a fourth "silence" phase, and then it repeats. The Hydra controllers have coils that sense these fields, and the processing circuitry knows which base coil was activated at a given time, synchronized to the silent "sync" phase.

The base signal has nothing to do with how many controllers are sensing its magnetic field signals. There could be a potentially unlimited number of sensors, limited only by other design considerations such as USB packet update rate and processing speed of the wireless base unit.

I have not studied the details yet though, but this is how I understand it. There is more discussion about the details of the Razer Hydra internals in another thread:
http://www.mtbs3d.com/phpBB/viewtopic.php?f=120&t=14036

It would still be nice to find a way to get multiple Hydras working together, perhaps my using one base coil set and somehow synchronizing (genlocking) their timing.

sixense website says kickstarter coming this month!!!!!!

VR enthusiasts.. time to get them wallets ready!

this os what tactical haptics said about the stem:

Tried out the Sixense STEM tracking system today.
It's got great range and appears to be much more accurate than the Hydra.

Kickstarter starting September 12th:
http://sixense.com/wireless

Hopefully with some detailed specs like accuracy, latency and range. Unlike some trackgun KS from not long ago...

just thought probably everybody who has or plans to get a rift will get a STEM. Plus ppl wanting to do trackguns/swords, etc...

that's a lot of ppl! And I doubt it will be cheap. This KS will go very high!

Happy for sixense! They deserve it

Well I hope it won't be too high, we have a Reactive Grip to back just after this one

drifter wrote:Well I hope it won't be too high, we have a Reactive Grip to back just after this one

Yeah really curious about price too... Preparing myself for 150-250$ depending on number of trackers.

I foresee quite a few STEM based Kickstarters coming in the next 12+ months..

- Assault rifle
- Gun
- Glove with finger/wrist articulation sensors
- Haptic feedback glove
- sword/staff/shield/...

please somebody do them gloves!

first specs: 240Hz system with 4.2ms latency

juicy

http://www.eetimes.com/document.asp?doc ... itedefault

Great, the big joystick controllers house smaller STEM cells developers can put whereever they want to. So you could put them onto the back of your hands. Now only gloves with a very primitive grab function/sensor are missing and gameplay will be so intuitive and intriguing.
And all on an ODT. Oh, my, god! By the way, what are those handles for?
Only for storing the STEMs?

yeah, they probably hope to mass produce the stem units. Makes sense. The more you have, the more motion you can capture.

Have you all forgotten?

THE STEM KICKSTARTER STARTS AT 7:00AM PDT TOMORROW

zalo wrote:Have you all forgotten?

THE STEM KICKSTARTER STARTS AT 7:00AM PDT TOMORROW

Nope... Breath is held and hoping that I'm dizzy enough not to notice the cost when I pledge to the tier including the 5 trackers.

I'm really excited about this. Hope they haven't set the goal too high.

Plastic ready to melt...

(btw my home-made avatar just became real ! )

The Kickstarter is live!

http://www.kickstarter.com/projects/895 ... th-virtual

drifter wrote:Plastic ready to melt...

(btw my home-made avatar just became real ! )

Nighteyes wrote:The Kickstarter is live!

http://www.kickstarter.com/projects/895 ... th-virtual

I went with the $299 version with 5 trackers.

It is trending toward a megabuck already!
http://www.kicktraq.com/projects/895778 ... h-virtual/

Opted for the 3 sensor bundle for $199. Pretty good value considering the enhancements over the hydra, plus the ability to track the head/chest with the 3rd sensor. Pretty amazing what the kickstarter has got too already on the first day!

Ok i don't like being a party spoiler, but isn't anyone concerned about the following?? (self-quote from another topic):

As explained very well by Doc OK in the following video (and also mentioned by Carmack at QuakeCon 2012), magnetic tracking is pretty good for relative positional tracking but not so good for absolute positional tracking, which i believe is what we really need:
http://www.youtube.com/watch?v=IERHs7yYsWI#t=3m30s

From my experience with the Razer Hydra this is very noticeable and annoying, i've never managed to feel that 1 to 1 correspondence between my real hands location and what i was seeing inside VR. Maybe Sixense have found a way around this with their new STEM system, but I remain skeptical for now.. (especially since i didn't see mentioned anywhere an improvement in this regard over the Hydra).

Also there's the big doubt about how is this going to interact with the Oculus positional tracking solution when it comes out. And finally there's the possibility of Oculus VR releasing their own hand-tracking system (which in my opinion would be the best-case scenario) and that would render the basic STEM system quite useless.. Personally i'm going to wait a bit to see how things pan out before parting with $250..

Good concerns about the STEM accuracy, knowing the Hydra biggest flaw...
Our only hope : the guys from Tactical Haptics (collaborating with sixsense) find that it "appears to be much more accurate than the Hydra." (dixit their FB page)
True, a bit of competition will be nice (as soon as there is a market)

geekmaster wrote:I went with the $299 version with 5 trackers.

Idem. Needed 2 extra STEMs for the Reactive Grip...

So we'll get it more or less when the Rift will be out... still wandering if we'll have redundant positional tracking system or what ?

The positioning accuracy problem described in those videos is only because a suitable calibration procedure was not used.

As I mentioned in other threads, Hydra/STEM positioning precision is repeatable to within one millimeter (or smaller for the STEM). Positioning accuracy with respect to the real world is relative to how well you calibrate it with a displacement map. A simple way would be to display "gloves" at various places in in VR, then you put your real hands into those virtual gloves and press a button to anchor that (repeatable) measured point in the magnetic field to that virtual (and physical) point in space. You do that for a number of locations and interpolate between them. Then your VR hands will be exactly where you expect them to be, and you can accurately use physical props in the real world that you can see represented in VR.

We just need a calibration displacement map, recalibrated each time the Hydra/STEM base is moved relative to the surrounding magnetic distortion fields.

There is no reason that accurate repeatable positioning error cannot be compensate with displacement mapping as I suggested in previous posts (and repeated above).

PatimPatam wrote:absolute positional tracking

In the first couple of minutes of this other vid, Doc OK himself says basically what geekmaster just reiterated:

http://www.youtube.com/watch?v=H5bSzVByLjM

So Sixense tech can do absolute 1:1 tracking, but this "virtual glove" calibration routine isn't worth the effort for consumer gaming use? Or is it just that no one has bothered to code something up that will make this relatively foolproof? With all of that sweet Kickstarter money sloshing around, hopefully Sixense is able to perfect a 1:1 calibration routine suitable for even luddite grandmothers.

Now for the real question: Where the hell am I going to get $349???

I thought I noticed that too in their demo vid. The positional head tracking doesn't look 1-1.

A displacement map lookup sounds like it would fix this if it doesnt introduce too much latency.

http://m.youtube.com/watch?v=UxSZ0fk_ubA

Edit: Actually latency is not a problem but seems that you would need to calibrate each sensor every time
To get to 1-1 tracking. But the most important one is probably the head.

@nateight: Great video. Funny at the end with the lightsaber. Wasn't expecting that.

Yes in theory it could be solved, but as of now it IS still a problem, and as nateight said the question remains on how practical this calibration procedure would be for the standard gamer..

I think it's a bit misleading to "forget" to mention this issue on the Kickstarter (as well as possible solutions to it) and not putting a row saying something like "absolute position tracking capability" or "position accuracy" on their comparison chart. I know it's standard advertising practice to ignore the weak points about your system and highlight the problems of others, but i still don't like it when it happens.

Don't get me wrong i believe it's going to be a fantastic product, i am very happy they succeeded with the funding and hope they reach at least the 1st stretch goal. It's always good to have more options, and i think it will be great for all you lot getting Omnis and such.

But for the standard sitting down (or standing without turning) scenario, where hands occlusion is not a big problem, i believe an outside-in optical system combined with some clever IMU sensor fusion could be a cheaper and more robust solution. I still have hope that either OculusVR or Sony will use a similar approach. We will see!

@ G33kmaster:

As I mentioned in other threads, Hydra/STEM positioning precision is repeatable to within one millimeter (or smaller for the STEM)

I can't find where I saw it (thought it was on the main KS page but cannot find it), but they are only claiming 3mm accuracy I believe. The Kickstarter page itself says that the precision on the STEM system is 'under 3mm'.
I wonder if its actually less accurate, due to using a stronger emitter - Ascension gear is less accurate with the long range emitter I believe.

WiredEarp wrote:Ascension gear is less accurate with the long range emitter I believe.

The accuracy decreases as distance increases due to a decrease in the signal to noise ratio. See this paper page 19 for a graph of inaccuracy vs range with the Ascension Flock of Birds system.

I would assume their "under 3mm" accuracy is for positions that are close to the emitter coil. It's listed on the comparison chart on their Kickstarter page.