DIY Head Tracker For A Tenner

Hi.

So I've come back to Elite after about 6mths away and can't get my old EDTracker 9150 working ingame. I'm not using the same system, having upgraded cpu/mboard and moved to Win10, so have installed everything from scratch again.

The EDTracker works fine in the GUI and I have set all the ingame control options correctly as far as I know. However when enabling Headlook it only moves if I turn my head to the extremes and acts like I'm pressing headlook buttons. i.e. I move my head left and nothing happens until my head is far left, then the camera smoothly pans left. If I look straight ahead again, or even slightly less left the panning stops where it is and doesn't return to centre.

Headlook Axis Mode: DIRECT
LOOK UP [+JOY Y-AXIS]
LOOK DOWN [-JOY Y-AXIS]
Deadzone: 0
LOOK LEFT [-JOY X-AXIS]
LOOK RIGHT [+JOY X-AXIS]
Deadzone:0

I did notice when setting the direction axis I had to move all the way before it would register, so it's like Elite is treating the EDTracker axis like buttons where 100% is ON and anything else is OFF.

Maybe I'm missing something obvious, any ideas?
 
Try assigning EDTracker axes to "Look Up/Down" + "Look Left/Right" instead of the individual directions. You describe assigning an analogue input to a digital control.
 
Last edited:
Ah yes! It looked to me like Look Up/Down was just a deadzone control but yes you assign the axis there too. I feel like a fool lol.

Thanks! *blushes*
 
Just bought "IL-2 Sturmovik: 1946" - Any tips for getting ED Tracker working with it? If indeed it can?

ps: I use Opentrack with EDTracker when in ED for example...
 
Last edited:
Okay, I got the right supples and the gyro is working. But how do I output to ED now? Windows doesn't seem to recognize it as a game controller despite installing the drivers from your site.

EDIT: I have emulation in opentrack, although it doesn't emulate pitch or roll. is that normal for the 6050?
 
Last edited:
Might be worth just clarifying exactly what you mean by "right supples"... what components are you using then? A Pro Micro now? MPU6050 breakout for sure?
What firmware have you flashed in, what version? Was it definitely the one for the 6050?
 
Brumster.

Hoping you can help me, this isn't Elite related but when I'm playing The Division my head is basically controlling my character, when I tilt my head down my chap runs forward, tip back and he runs backwards etc etc lol. I need to disconnect it I think but will that mean I need to calibrate the magnetometer every time or will it remember its settings when next plugged in?
 
Last edited:
I just got the v3 PCB from Edtracker and the Arduino and MPU-9250 from Hobby Components. I put everything together and was able to flash it successfully via the 4.0.4.0 GUI (The status windows showed that the flash was complete). And that's where the fun ended. The GUI doesn't show any data for the temperature or any thing else it should be receiving from the MPU. The Arduino has a solid red light, and intermittent has an amber light flickering. I haven't seen any lights on the 9250 (are there even LEDs on this thing?). I got to the point where I even tried desoldering the MPU (wow, what a pain) so that I could verify all the solder points on the Arduino and then resolder the MPU back on, but still no luck.

I spent the last few hours searching and troubleshooting, but I'm at a loss. Any suggestions? Is there a way to validate that either the Arduino or the MPU is functioning properly? I read about someone experiencing a similar issue and there solution was to remove the button. Why would this work? In order to remove the button, I'd have to desolder the Arduino board which is not something I want to do unless I'm fairly confident it will solve my problem.

Any guidance is appreciated!
 
found this works on headtracker setup.JPGmine hope its some help
 
Another slightly odd bit of behaviour from the Client.

When it's minimised to the task bar, and you double click it to maximised it (one click won't do), it very strangely results in you clicking the icon in the task bar in effect next to it.

ie: You (double) click it, so it disappears (as its being maximised), but one of the clicks then goes to the icon that replaces it, so (if that requires one click) that gets opened too.

So I double click it in the task bar and I get it and another application opening up...



This is an oddity ontop of an oddity for me, as I also have strange behaviour that when I first minimised the Client, the hotkey doesn't work, so I have to maximised it and minimise it again. And of course, in doing this, I then get the odd behaviour of a (random) second program opening up from the task bar too.
 
Hey brumster, any news on your wireless version?

Last I checked you were going through the certification process. Waiting with bated breath here, I think I'm turning blue!

Same here! :)

It's coming along really well, I have been playtesting it the last few days (as has Rob) and it works "awesomely" :). I'll do a teaser video some time. Obviously it's a balance between talking about it and getting everyone excited, and then having to let them down when we find some aspect that is going to delay it. A bit like Frontier in that respect - "Soon"(tm)!

But it's going really well, all positive, the device itself is working and totally practical in terms of it's performance and behaviour, so it really is just "hoop jumping" at this stage. We're refining the software and the power management to get battery life as good as we can get it, but to be honest even with a 300mAH LiPo and no power management whatsoever, it was running solid for 13 hours for me.... so I think we can probably down-spec the battery on the final revision. The power management is very cool with what Rob has done, actually :) if the device detects it's been left still for a period of time, it goes into a "deep sleep" to conserve charge. Give it a shake/move, and it wakes up ready for play... this way users don't have to worry too much about turning it off with the hard switch (although for extended periods of non-use this will be advised).

Like I say, I don't want to get peoples hopes up and state firm dates because we're at the whim of processes and procedures outside of our control. But rest assured it is coming, it works, it IS a thing, and it is pretty darn cool :)
 
Last edited:
When this is a thing, you could get in touch with the people at www.RiftCat.com, who are doing an Oculus emulation streamer to smartphones. Good head tracking is the missing component in making cheap Cardboard like headsets into actually tolerable VR.
 
Like I say, I don't want to get peoples hopes up and state firm dates because we're at the whim of processes and procedures outside of our control. But rest assured it is coming, it works, it IS a thing, and it is pretty darn cool :)
Are you sure you can't send someone over to hurry things up a little?

yoLgGdY.jpg
 
I just had a fantastic idea!

I know we've been throwing around positional (translational) tracking for the EDTracker, and a good attempt was made using the code from FacetrackNoir. Current implementations of positional tracking include a stationary camera and some sort of moving target mounted on the observed's head. Brumster, is it possible to put a tiny camera in the EDTracker and a stationary target on the monitor? It could be a paper target, an IR led, and reflective IR tape with an IR light source on the EDTracker. And all the processing would occur within the EDTracker module itself.

Is that possible?
 
Technically yes, but it wouldn't make much difference which way round things were (camera on head or camera on monitor) and I'd have thought people would prefer to minimise the amount of bobbins on their head... :) unless I'm missing some advantage to having the camera "user-side".... :) ?
 
Back
Top Bottom