Vuzix Forums / VR920 / Tracking / yet more head tracking problems / Latest Posts

RE: yet more head tracking problems

nuclearhythmics — Sun, 02 Aug 2009 22:04:24 GMT

My yaw troubles were solved after following vic256's advice. I took my PSU off my desk and actually put it into the next room (no more loud cpu fan). Also I've noticed that my metal eyeglasses effects yaw and it's calibration (maybe because they have little magnets on them for clip-on sunglasses) so I've been using my plastic framed ones, or contact lenses instead. The documentation even mentions to calibrate the VR920 with the temples extended because the headphones affect the magnetic field (I've detached the headphones anyways). There is still a little interference or noise so I used the smooth opcode on yaw in the Glovepie script which settles it but creates a slight delay.

If you have serious problems you might consider getting a Wiimote and wireless sensobar (they're pretty cheap on Amazon):
http://forums.vr920.com/Topic3261-8-1.aspx?Highlight=freetrack
I ordered them for around $30 to get the additional 3 axis x/y/z for head tracking. I'll probably just continue using the VR920 yaw/pitch/roll and use the Wiimote for x/y/z when I receive it.

RE: yet more head tracking problems

Comicus — Mon, 23 Mar 2009 09:02:33 GMT

That doesn't work, at least not for me, because the deflection of viewing angle is never consistent. I've been looking through the forums for help with the issues I've been having, very similar to this thread, and will try to increase the yaw filtering as suggested. Now if only I could figure out where that setting is located...

I'm also finding that my axis are coming through a bit wonky (technical term). Meaning, my roll axis and yaw axis seem to cross-reference in FSX. This doesn't happen in AxySnake and the iWear Monitor shows the values change correctly, so it must be something in the translation to FSX. I'm in the process of trying to get IL2 up and running and see how that goes.

RE: yet more head tracking problems

Mnemonic — Mon, 23 Mar 2009 03:43:02 GMT

It's very strange guys, I don't have any problem flying FSX with VR920.

And if you told that you have stable wrong zero-position in FSX, compensate this when calibrating, try to set zero position not in front of you but a bit left/right (depend where you have stable zero-point in FSX), that's all.

RE: yet more head tracking problems

daytonabeach — Sun, 22 Mar 2009 11:43:55 GMT

The conclusion on this problem is, VR920 does NOT work with FSX, PERIOD!

RE: yet more head tracking problems

kevaquel — Fri, 14 Nov 2008 13:22:22 GMT

Lots of words but the question is:

How to SOLVE THIS PROBLEM?

11/14/2008 - and problem still extists.. :/

RE: yet more head tracking problems

Alopex — Sun, 03 Aug 2008 09:21:05 GMT

Your magnetic compass is a small metalic arrow which was exposed to a magnet so that if it is free to move, it aligns with the local magnetic field.

Since your magnetic compass is magnetized, it is attracted by anything metalic (containing iron).

So you are correct: your magnetic compass will be attracted to the iron inside the cage, or anything magnetic inside, for that matter. :)

But inside a faraday cage, you have no way of knowing where the magnetic north is though ;-) (nothing from outside goes in, nothing from inside goes out, magneticaly speaking of course :))

---

The important thing in our case, I believe, is to offer your VR920 (which if I understand well, contains the electronic version of a magnetic compas) a strong stable magnetic field.:w00t:

There is no need for a faraday cage (too clumsy, too expensive, too impractical).

If your environnment is not magnetically stable, you can suppress the noise it creates by shouting (magnetically) louder.

Shouting louder can be done by creating a stronger magnetic field than whatever is available around you.

To avoid magnetic perturbation, you can put anything you detect as being magnetically active inside a faraday cage (everything inside you pc box shouldn't be a threat if the box is metalic and closed). The screen can be a problem, the speakers too. But your mobile phone, your earphones, can be a threat too.

To detect what is perturbing the field, my suggestion stands: use a compas, move it around where you head is supposed to stand, and see if the arrow moves eraticaly while the game is functionning. See what is the main perturbator, turn it off or shield it (for your phone, you can make a pouch out of aluminium paper).

---

small story about magnetic perturbations: the doors of shops usually have coils which create a magnetic field. Anything inside that field is going to interact with the emiter: if it catches the field, it perturbs it and the emiter can "feel" the change.

The thing with anti theft tags, is that they interact with the field in such a way that the emiter detects a change in the way the field is perturbed. So what some knowledgeable people can do is put the stolen goods inside a pouch made out of aluminium to avoid detection. ;-)

Some people do this with their bank notes, or their passport too ;-):cool:

RE: yet more head tracking problems

Graiser — Sat, 02 Aug 2008 23:11:19 GMT

Are you sure about that? I'm pretty sure a compass continues to work in a faraday cage, at least as well as it does in an incompletely finished faraday cage.
The load of metal might mess with the field, just as the compass in the hilt of my old cheapy survival knife always pointed odd, (metal hilt.) but the fact it's a faraday cage shouldn't make any difference.

RE: yet more head tracking problems

Alopex — Wed, 30 Jul 2008 12:34:56 GMT

haha... I like your idea...

Makes me think about the detector a the door of shops: they carry coils that act as antennas.

The thing is that if you look at the magnetic field inside a coil, it is usualy quite stable. same in between coils...

So you could try 2 coils at opposite directions (or even, magnets, why not :w00t: ), to create a near stable field where your head moves or stands... (by stable, I mean straight, but can't find teh exact technical word in proper english, you'll have to excuse me ;-))

If you go for this setup, try using a magnetic compas to check that the direction of the field is approximately stable... espcially, you can turn them magnets a bit to "clean" up your field.

Maybe also, you'll find out that your screen is definitely blurring things as it functions in specific modes... or even you computer, or if you are like me, your power suppply, your printer, your speakers standing on your desk...

My wild guess is that if you manage to magnetically clean the place where your head stands you get good results (ie: making it in such a way that the arrow of your compass points in the same direction in the approximate place where you put your head).

Note: Cleaning doesnt mean erasing: putting yourself in a faraday cage will erase all magnetic interference from teh outside, but the compas in the vr920 will be lost:hehe:

RE: yet more head tracking problems

vic256 — Fri, 25 Jul 2008 15:59:49 GMT

I actually tried creating a 'artificial' reference point for the yaw using an electromagnet ripped from an old door stopper, it certainly did something to my yaw tracking and if close enough it did work somewhat but it had to be too close to my face to be practical... As if I didn't look geeky enough with the glasses on, staring at a large plywood board with a doorstopper in the center, wearing the glasses certainly added to the geek factor according to my wife. ;)

I belive that magnetism is subject to the inverse square law which means that the magnetic field quickly gets weaker the further away the magnet is (It's spreading the same force over a larger area) meaning that it will at relatively small distances become an interference rather than a reference point ...

Now I'm not an electronics engineer so maybe someone else could comment and even suggest alternatives or maybe have tried with strong permanent rare earth magnets like neodyms or weaker ferrite magnets ...

Rgds,
Troels

RE: yet more head tracking problems

Alopex — Fri, 25 Jul 2008 01:23:01 GMT

[quote][b]benoit.schillings (2/7/2008)[/b][hr]I noticed that certain rooms are so well shielded that the earth magnetic field is not picked much by the sensor.

Kind of a pain to take your machine outside, but this is a possible test, or at least near a window pointing north.

-- benoit[/quote]

This post made me think about this:

Maybe when my game is on, maybe the speakers are creating a magnetic field that corrupts the VR920 readings.

Could it be a good idea to set up a strong magnet in some place in the room as a reference point?

Maybe that could help the vr920 work?:w00t:

What do you think?

RE: yet more head tracking problems

electromaggot — Thu, 24 Jul 2008 15:43:35 GMT

Thankfully the people paying for it had deep pockets.

Heck, a Xybernaut wearable computer would run you $10,000 (US) and that was 8 years ago!

Today, these "ultraportable" PCs (like the Asus Eee) are 20x cheaper and much much more powerful! I have a Eee and love using it with the VR920, which even adapts to & displays the 1024x600 resolution perfectly (which you can easily change or turn off the main LCD). The USB & VGA ports are right next to each other, so I coil up & twist-tie the extra USB cable length and it's like the two were made for each other.

I don't game with it (I should try sometime but I don't think there's any 3D acceleration), although I can do everything else, including surfing and videos, whether from YouTube or a plugged-in slim DVD drive (which is also powered over USB).

Sorry, we're way off topic... should've started another thread... but maybe we've exhausted it? ;)

Tadd

RE: yet more head tracking problems

Graiser — Mon, 21 Jul 2008 22:07:02 GMT

Wow. 2.5-3 times as much for Just the tracking.

RE: yet more head tracking problems

electromaggot — Mon, 21 Jul 2008 00:19:57 GMT

Graiser:
Thanks for schooling me and for the interesting info. I'm actually an EE but I don't know a lot about piezoelectric speakers. So I bet you're right that they don't affect the tracker much, if at all.

Troels:
Good points! I'd say your PSUs are very suspect (which it sounds like you verified). A power supply contains a sizable transformer, and transformers are all about magnetism. (For anyone who doesn't know how transformers work: they contain two coils of wire - that's two separate conductors, but they're wound together. The electron flow through one coil [b]couples magnetically[/b] to the electrons in the other coil. If one coil has twice the number of winds as the other, you're stepping down the voltage by half. With AC (alternating current), that magnetic field will be fluctuating wildly, swapping poles every 1/120th of a second (60 Hz in the US, that is).) So maybe those PSUs "emit" enough EMF to mess up your tracker calibration.

Graiser:
I'm intrigued by your plans to "wear" magnets to influence the tracker. I will be [i]very interested[/i] to hear how that works! Especially, how the earth's magnetic field or other ambient EMF will affect it as you move around.

Believe it or not, I did something similar with a wearable computer once. However, I used two separate trackers (two USB InterTrax2's), one attached to the belt-worn wearable and the other attached to the HMD. I took the difference between them to arrive at my HMD reading. That way, I could display something relative to the position the user was facing (or at least the way his belt buckle was facing!). It actually worked well. So that's one option for ya, if the permanent magnets don't work out.

RE: yet more head tracking problems

Graiser — Sun, 20 Jul 2008 12:48:58 GMT

First of all, I don't own a pair, yet. So take all I say as the speculation it is.

The headphones are almost certainly piezoelectric units. This means there's a thin layer of something like quartz that deforms when hit with electricity on the earphone membrane. Interestingly enough, if you strike it, it will generate a small electric charge, (this is the basis of both some microphones and those little electric ignition cigerette lighters. They shouldn't have driver coils like you find in large speakers, so shouldn't produce any significant magnetic field. A bunch of flourescent lights would probably give you greater issues.

Second of all...I'm contemplating sewing small magnets into my daily-wear jacket when I get a pair, so I can, hopefully, use them anywhere, at any orientation, without having to recalibrate, (remember, I'm thinking wearable computer.)

RE: yet more head tracking problems

vic256 — Sat, 19 Jul 2008 14:22:53 GMT

Well, I've tried to calibrate the set like shown in the manual, slowly and carefully 1 axis at a time and that doesnt give me a good calibratoin, EVER! ... I do as I described in my last post and it almost always is a success.

Second thing, I've got like 3 laptop PSU's, a router PSU and a WiFi access point PSU under my sofa and if I sit on the sofa in a radius of about 1,5 meters I NEVER get a good calibration, PSU's unplugged or outside that radius I almost always get a good calibration.

Third thing, I've got a heater (radiator using hot water for heating) made of metal, 3 meters long under my windows, if I sit too close to that, one side of the yaw will give me less travel than the other..

I have no doubt that at least electromagnetic fields and large metal object interfere with my calibration :w00t: I'm not sure about radio interference though , my mobile phone never did interfere from what Im aware of, I just remove it or shut it down to be on the safe side :) I hate it when it doesnt track right!

Rgds,
Troels

RE: yet more head tracking problems

electromaggot — Sat, 19 Jul 2008 12:44:18 GMT

Today for the first time, I used my VR920 in front of an old 21" Trinitron CRT that I still have (the only thing I can still use shutter glasses with for stereo 3D). I noticed the yaw tracking wasn't acting the same as when I'm in front of LCD monitors. Yes I did recalibrate.

So, if the tracker does indeed use "earth magnetic field" or otherwise, the room's "magnetic signature" for its yaw reference... I wonder how susceptible it is to magnetic fluctuations from everything around us that has magnets -- just as Troels mentioned. A 21" CRT has some massive electromagnetic coils in it to steer the scanning electron beam.

I'm curious... Show of hands: How many people having trouble with their yaw tracking are using CRTs? Big CRTs?
(Can we do a poll on here?)

For that matter, anything with electrons flowing through it has a magnetic signature. Anyone with yaw problems live nearby any high-tension power lines? Or... do you live in a building with lots of metal in the walls or ceiling? Maybe that could make a difference? I wonder how RF affects the sensor as well -- any type of EMF -- like living too close to a radio broadcast tower.

Final thought: as Troels also mentioned, maybe unplug the heaphones from the HMD entirely and try that? I wonder how well they are shielded, especially if you turn them up real loud. These are just ideas I'm kickin' 'round.

Tadd

RE: yet more head tracking problems

vic256 — Sun, 13 Jul 2008 13:57:50 GMT

What's all this about the yaw not working? Yes it's difficult getting a good calibration, sometimes it takes me a couple of tries where each calibration session takes me around 3-6 mins but when it properly calibrated it works fine here ....

When I calibrate I do not do as instructed in the manual though, I turn the glases very casually and not necessarely slowly around all axis', even vertically but except upside down (Always gives me a bad calibration) for 3-6 minutes (Untill I can't get the bars to move anymore), ensuring that all positions possible has been used during calibration (360 degrees while wobling the glasses, tilting them, holding them vertically etc.).. I've never done a calibration more than twice without having a 'good enough for fsx' or perfect calibration. I allways do the calibration where the glasses are to be used and remove PSU's, speakers, subwoofers, mobile phones and everything else with either strong magnetic fields or electronic interference before calibration and playing. I even extend the the buildin headphones for them not to interfere too much with the calibration.

Second thing, I increase the filtering amount in FSX so that the 'deadzone' for yaw was widened a bit, not much, just to remove the twitching that sometimes occur as if some magnetic interference was introduced in the environment once in a while.

Anyways, this can't be a general thing as mine works fine, if yours are defective, why not return them and have them fixed under warrenty?

Rgds,

Troels

RE: yet more head tracking problems

SuperDre — Fri, 11 Jul 2008 14:59:12 GMT

I guess we're just screwed with that unless they bring out a software solution.. I also can't get the yaw to really act like it should.. I wonder why this wasn't fixed (in other words, replaced with a better hardwaresolution) before the vizor was even released (or maybe they started using even cheaper solutions).. I would at least as a Q&A manager never have released a device this buggy, almost every (professional) review has some comment on it.. and how is it that some older helmet like the VFX-1 (which was also developed by some of the vuzix people if I'm not mistaken) really do work..

But I guess they will never fix it or provide a solution... it's almost 4 months since the last software update, and not a peep of any vuzix employee on these forums...

RE: yet more head tracking problems

Loony — Thu, 10 Jul 2008 03:33:06 GMT

So basically, what's the cure ?
I have exactly the same problem as of all of you, the horizontal axis doesn't work at all (I have to tilt my head to see a bit of the left/right of my airplane on FSX...)

It's frustating, since the rest is really great !

RE: yet more head tracking problems

tzetsin — Sat, 08 Mar 2008 15:34:39 GMT

I finally got my unit updated, and after all that fuss and waiting.... All my complaints were addressed! yay!! The tracking is still buggy but its actually working now, and the graphics look ALOT better, i can actually watch movies with them now with nice colour, though with games i still find it looks dark when i have desktop settings on and it looks ok after putting brightness / contrast sliders all the way right. If i may be so bold Mr designers could you add a profile option to the monitor... or the glasses themselves that will allow me to have one setting for desktop/games/movies ect. or just find out what causes it an make it work better ;)

RE: yet more head tracking problems

DesH — Sat, 08 Mar 2008 09:24:07 GMT

Sounds similar to the problem I've been having. Uninstall your VR920 software and then re-install it and that seems to cure the problem for me. Thing is, everytime I start my PC and plug in the VR920 I have to go through the calibration all over again which is a real pain. I'm using latest 110 driver so not sure if its just a bug with this ?

Regards

Des

RE: yet more head tracking problems

pondo22 — Wed, 20 Feb 2008 12:46:21 GMT

So are we saying the tracking is screwed?

RE: yet more head tracking problems

CarlKenner — Tue, 19 Feb 2008 09:40:13 GMT

Using accelerometers for measuring yaw is very tricky, and would probably be less reliable than using a compass. Pitch and roll are measured with accelerometers because those angles change which direction gravity is, and accelerometers mainly measure gravity.

Accelerometers can also measure acceleration, but they don't measure it reliably enough to calculate velocity and position from it. Also it is impossible to tell the difference between moving and turning using only one accelerometer. If you had two accelerometers at different ends of the device, it might be possible, but it wouldn't work well.

Their other real choice besides a compass for measuring yaw would be to use rate gyros. Rate gyros have lots of drift, and no fixed reference point, so they are not much good for playing sitting down in front of a keyboard and mouse. But they work OK if you play standing up. Basically they measure yaw speed, not yaw itself. The Trimersion uses rate gyros for measuring yaw, so does the PS3's SIXAXIS. The Z800 uses a mixture of the rate gyros and a compass.

They also could have tried some kind of optical tracking, like the TrackIR, the P5 Glove, and ARToolkit use.

RE: yet more head tracking problems

tzetsin — Tue, 19 Feb 2008 09:00:47 GMT

I can tell you what i did to solve the problem... i stuck em on the shelf till the firmware patch comes along and pray to whatever gods'll listen that it fixes it. Using them in the state they're in just upsets me. I had really high hopes for them too. The 3d part is still pretty cool, but i'm a scenery flyer, i like the eye candy and with the colour being so iffy on them anyway i cant justify using them to fly if i cant use the tracking. If the tracking would would i could easily overlook the colour because of the added depth of being able to look around.

I've been trying and trying to get them to work... i've got Vista & XP dual booted on my home pc and my laptop and tracking just wont work with any amount of usablity. Its always the yaw... and it sounds like its always the yaw for EVERYONE so i gotta ask... Why didnt vusix use acceleromiters for the yaw like they did with the other two axis?? I'm sure theres a good and technical kinda reason, so i dont expect anyone to smack their forhead and ask themselves why THEY didnt think of that... but theres gotta be a better way?

Carl, i was only using your program to try to fix the problem i was having, so it wasnt glovePIE that was causing the problem for me.

RE: yet more head tracking problems

pondo22 — Fri, 15 Feb 2008 15:31:24 GMT

Since my last post its got a lot worse I can't even use the things now in FSX. I thought it was FSX that was the problem but after testing it would seem the Calibrator is incorrect.....looks like it's just the Yaw thats effected pitch and roll are ok. Does anyone else know of the problem being the hardware? Really hope it's not I just got them working after the firmware problems.

Ps I have tried calibrating for 5 mins and it's still the same apart from my arms falling off!

RE: yet more head tracking problems

pondo22 — Wed, 13 Feb 2008 13:22:39 GMT

I'm having the same problem with FSX all seems ok in the calibration but as soon as I start FSX and fly about for a while then set the head set down for a while and go back to the game the tracking is all out and can't be sorted. I have shut the game down and start again? Also when it is working in game it's not quite right....if you look directly behind it gets all confused and jitters from side to side. Maybe is theres quite a few having the same problem it needs a patch to sort it out?

RE: yet more head tracking problems

CarlKenner — Wed, 13 Feb 2008 03:27:44 GMT

You didn't change anything in the script.

It is the placebo effect.

The bottom IF part is the part that probably applies to your game.

RE: yet more head tracking problems

frogger — Tue, 12 Feb 2008 15:59:43 GMT

i got the pitch problem fixed by changing sum numbers in the script

PIE.FrameRate = 120Hz
if FakeTrackIR.ProfileID=332DH then
FakeTrackIR.yaw = smooth(vr920.Yaw - var.offset)*10
FakeTrackIR.pitch = vr920.Pitch+50 degrees
FakeTrackIR.roll = -vr920.Roll
else if FakeTrackIR.ProfileID=3A35H or FakeTrackIR.ProfileID=3A38H then
// Arvoch Conflict and Evochron Alliance use a different range
FakeTrackIR.yaw = (vr920.Yaw - var.offset)*0
FakeTrackIR.pitch = vr920.Pitch*5
FakeTrackIR.roll = -vr920.Roll*10
else
FakeTrackIR.yaw = (vr920.Yaw - var.offset)
FakeTrackIR.pitch = vr920.Pitch
FakeTrackIR.roll = -vr920.Roll
end if

is the smooth part right for the filter ?

RE: yet more head tracking problems

CarlKenner — Tue, 12 Feb 2008 13:58:22 GMT

That's probably a glovepie (script) bug.

If you play facing whatever angle you set as zero, then it will work fine, otherwise you will have that bug.

Sorry, it's my fault. I'll have a go at fixing it.

Edit: I don't think the yaw changing when you pitch is my fault. I think it is just lazy programming in the driver. Or I might be misinterpreting the euler angles and applying them in the wrong order.

RE: yet more head tracking problems

frogger — Tue, 12 Feb 2008 04:20:31 GMT

im having this type of problem also

in the calibrator it seems perfect i am getting 0 jitter even , i think this is cause when calibrating it i was rotating it about 2mm a sec , it took longer but it seems good "within the calibrator , in games like lomac and fs9 when i turn my head left i get about 15 to 20 degrees of turn if i go right i get the same and then it flights around to the 180 degrees instantly , when i get it to not do this i have problems where if i look up past say 15 degrees it starts using the yaw . this is using glovepie as the iwear monitor program doesnt run anymore , i get an error saying "runtime error ... floating point support not loaded" then the program closes , although im getting no jitter in the calibrator i am still getting it in games . when i try to add the filter with the smooth(vr920) thing the faketrackir stops working ( carl can u help with this one please :D )

ill try calibrating it outside since i have a laptop and ill see if its any better.

Edit:ok just went out side and it didnt help its going in the 40's to the left and 120's to the right , was only out there for 1 try since its almost 40 degrees outside and they were heating up like a bitch :hehe: now im inside and tryed in in the living room with the aircon blowing in my face seems to be getting yaw both ways in lomac nicely but still when i look up it yaws to the left :( i realy dont like the way this thing tracks its yaw but i dont want to stick ir lights on it just yet :P

RE: yet more head tracking problems

tzetsin — Thu, 07 Feb 2008 17:36:18 GMT

when i turn my head to the left and right in IL2 and in FSX the camera just jiggles around a couple degrees, when i turn my head fast in one direction it kinda goes that direction just a bit but then yo yos back the opposite direction soon as i stop, again, updown and roll work fine. Like i said though, when i exit the game (during the off times i actually get the left and right yaw on the calibrator to actually work) it seems to be working. i've exited the games and restarted both games again... using glovepie i'm getting the same thing when using the glasses to run the mouse curser, i can get it to go up and down, but no left and right other than just uncontrolable jiggling around left and right.

RE: yet more head tracking problems

CarlKenner — Thu, 07 Feb 2008 02:21:55 GMT

If it is working in the calibrator, then it should work in the game. It may just be a problem with the game not using enough filtering. When you say it doesn't work in the game, what problem is it having?

RE: yet more head tracking problems

benoit.schillings — Thu, 07 Feb 2008 01:04:41 GMT

I noticed that certain rooms are so well shielded that the earth magnetic field is not picked much by the sensor.

Kind of a pain to take your machine outside, but this is a possible test, or at least near a window pointing north.

-- benoit

yet more head tracking problems

tzetsin — Wed, 06 Feb 2008 23:45:16 GMT

ok i dont want to be a [meanie] but at what point do i finnaly say i think my glasses are pooched? i've been trying and trying and trying and i absolutely CANNOT get turning left and right to work with any kind of reliablility. on the off chance i get it to work in the calibrator it'll be completly useless in fsx or il2, then i'll switch back to the calibrator to see if its functioning and it "seems" to be moving left and right properly, just not within the application.

dont say calibrate it for an hour, if i gotta calibrate these sobs for anouther 10 mins i'll run em over with my truck.