FreeTrack Forum

Welcome, you're not connected. ( Log in - Register )

RSS >  Hardware only Solution
jm2 #1 11/05/2008 - 17h27

Class : Apprenti
Posts : 3
Registered on : 11/05/2008

Off line

Hi.
I got a link to the freetrack page and i like this tool, but i noticed that it isnt
supported by all games.
I want to play IL-2 1946 with freetrack, but when i found some things in the
web about viruses, needed game modefications that can casue that i can
get handled like a chaeter on some servers ... that gets me scared.
So i thought about that problem a got another solution.

To get the information of head moving by using a webcam or a trackIR like
system is not the only way to make such a system to get working.
A second way, thats my idea, is to use mechanic sensors that gets mounted
on a basecap and converting the position information to electronic usable
signals by using potentiometers.

The simple point on this way is, that the electronik that could be used to
get this sensor informations into the PC and the game is a cheap and simple
analog joystick or joypad.
Some joypads like the Playstation ones with two analog sticks could be
used for this, every stick has two standard potentiometers in it.
This two small ones (not very accurate in cause of the little size) simply
has to be removed and replaced by higer quallity ones in industrial quallity
to become more accurcy.

Perhaps this solution ist not as perfectly accurate as the webcam based
ones and it only allows two axles of control, like changing the view with
a mouse (the "roll" axle is not available), but this solution doenst need
any kind of software, oder special support by any game.
Every game where an analog input device (joystick / analog joypad)
can be used to controll the view would support this solution.
There is no software modefication needed that can cause the suspicion.

Also this way is not expensive as webcam solutions, cause theeres only
a simple analog device and seom smaller parts needed.
So even people with enough money for other solutions could use this one.

And then there are still two potentiometers and (at my joypad) 11 keys
left that could be bound on game functions (putting them into a wooden
or aloy panel would give a nice labelled controllbord).

This 11 keys also could be used in combinition with a little speech
recognition unit (small board, 5V power supply, all functions integrated)
to realize a software indipendent speech controlling of some functions,
like it is in the real Eurofighter 2000.



What do think about this idea, should i work on it or waste it.
I have to say, all parts are allready in my partsbox (yes, also a speech
recognition module with continiuos listening technologie lays since a
year in this box)


thx, jm
didja #2 11/05/2008 - 19h30

Webmaster (admin)
Class : Webmaster (admin)
Posts : 1043
Registered on : 09/07/2007

Off line Www

Hi jm :D !

I don't know much about electronics, but I think your solution may work. However, it would be quite unprecise and hard to get it well calibrated. But I'm not sure about what I've just said ^^ .

But the point is that you can get 2DOF working into IL2 without any modification. The modifications are only needed if you want to play the 6 DOF. It's a limitation of IL2 that you can get rid of by using the famous mod, but you're not going to be able to play the game on some servers since you could be considered as a cheater.

But your idea is still interesting ;) ...

Rémy :)
Posted Image
Deimos #3 11/05/2008 - 19h45

Class : Beta Tester
Posts : 120
Registered on : 07/11/2007

Off line

Well, you weren't the first one who thought of using a joypad for headtracking:
http://www.youtube.com/watch?v=ajGzyTCwsTY
But i wouldn't expect too much from such setup... If you really want a hardware-only solution, you might want to look up some head mounted displays (HMD) -  the more expensive models have head tracking built in. And even more expensive ones enable you to see the virtual world stereoscopically - using two separate video inputs and independent displays for each eye.
But coming back to ground, i wouldn't agree that webcam-based solutions are more expensive than ones using potentiometers - remembers that potentiometers tend to wear out with time, while webcam can last for years. And so can the hat/clip rig. Also, you don't really need a microsoft or logitech camera - there are lots of cheap webcams with parameters comparable, or sometimes even better that the expensive ones. Just check the webcam compatibility tread - in some cases there are camera prices posted.
Also, i think that the LED rig for freetrack is much easier to build (and probably more comfortable to wear) than anything mechanically atached to your head.
If you really want a purely hardware solution, then gyros (for rotation measurement) and accelerometers (for translation measurement) are the way to go - that's how headtracking HMDs work. But gyros and accelerometers are rather expensive, and rather hard to use - you need to know your way around electronics and microcontroller programming if you wanted to interface them with a PC.
Edited by Deimos on 11/05/2008 at 19h56.
jm2 #4 11/05/2008 - 20h50

Class : Apprenti
Posts : 3
Registered on : 11/05/2008

Off line

This solution u can see in this youtube video ist a little bit to simple.
I also think that this would not work accurate and over long times.
And i dont want to get such a construction on my shoulders.
But when u see this simple one works so good,
a more complex one with better sensors should work too.

Accelleration Sensors and a gyros stabilisator could not be used,
they dont give u a chance to get a relative position of your head to
the electronic.
The potentiometers are not the problem, there are ways to use sensors
withouft scratching contacts that works digital/optical.

And when i see what some guys are wearing kind of strange modified
basecaps with LEDs, i think a other construction that is as heavy or
less heavy as this webcam based sensor caps, i dont think a lightwight
system built from a headlamp rubber ribbon and some small sensors
would be acceptable for me :D

Perhaps i can find a way to built a hardware only solution thats based
on a webcam like sensor thats connected to a joypad electronic.



Look at this too, nice video:
http://www.youtube.com/watch?v=AcFAC-hsps8&NR=1
Deimos #5 12/05/2008 - 00h34

Class : Beta Tester
Posts : 120
Registered on : 07/11/2007

Off line

Well, good luck then. I must say i really like the solution shown in the video and are looking forward to whatever you come up with :)

However, i'd like to point out a few things:

but this solution doenst need
any kind of software, oder special support by any game.
Every game where an analog input device (joystick / analog joypad)
can be used to control the view would support this solution.
There is no software modification needed that can cause the suspicion.


Of course you know that FT can emulate mouse, keyboard, and joystick output so that it can be used in games that aren't officially supported, don't you? Besides, IL2 supports TrackIR/FT even without any modifications, as didja said.

Accelleration Sensors and a gyros stabilisator could not be used,
they don't give u a chance to get a relative position of your head to
the electronic.


Yes, they can - it's true that their raw output is measured linear and angular acceleration, but if you differentiate their output, you'll get their angular and linear positions. And numerical differentiation is not that hard, especially if done on PC side, or with a DSP (digital signal processing) microcontroller.

i don't think a lightweight
system built from a headlamp rubber ribbon and some small sensors
would be acceptable for me :D


Well, there are a number of FT rigs - practically every single one is different, and not all of them are heavy or sophisticated. For example this:
Posted Image
is my clip that is attached to a baseball cap when using FT. It's just a piece of 0.5mm plastic and some reflective tape, so it wieghts almost nothing. It is illuminated by infrared LEDs mounted near the webcam, and the camera sees this light reflected off the reflective tape. And if you still want it lighter and less weird-looking, you could probably just use pieces of reflective material attached to some cap/hat as well.

I strongly do encourage you to get on with your idea, but in the meantime, i'd also suggest trying out the freetrack in some game supporting all 6 degrees of freedom (there are even some freeware ones around), especially if you have a few LEDs and resistors laying around - believe me, the experience is well worth even wearing a stupid-looking cap :D
Edited by Deimos on 12/05/2008 at 00h36.
Kestrel #6 12/05/2008 - 03h16

Webmaster (admin)
Class : Webmaster (admin)
Posts : 780
Registered on : 13/07/2007

Off line

Deimos @ 11/05/2008 - 23h34 a dit:


Accelleration Sensors and a gyros stabilisator could not be used,
they don't give u a chance to get a relative position of your head to
the electronic.


Yes, they can - it's true that their raw output is measured linear and angular acceleration, but if you differentiate their output, you'll get their angular and linear positions. And numerical differentiation is not that hard, especially if done on PC side, or with a DSP (digital signal processing) microcontroller.



Measuring position absolutely is always going to be better because there is no chance of centre drift.
WhiteWolf #7 12/05/2008 - 05h19

Class : Apprenti
Posts : 6
Registered on : 21/04/2008

Off line

I was thinking about this yesterday myself. One device which could be used is an electronic compass but this would only give you side to side movment. The ultimate harware solution would need to incorporate several seperate devices. head tilt could be achieved using electronic levels possibly, one for horizontal and one for vertical. Getting full 6dof would be very difficult to achieve and also vey expensive to build.
Deimos #8 12/05/2008 - 14h12

Class : Beta Tester
Posts : 120
Registered on : 07/11/2007

Off line

Kestrel @ 12/05/2008 - 03h16 a dit:


Measuring position absolutely is always going to be better because there is no chance of centre drift.


True. But still, sometimes relative positioning systems are being used with success:
http://www.youtube.com/watch?v=a2rM5Zy5T5Y
And a bit more high-end solution:
http://www.youtube.com/watch?v=JeGflcAW_-g - full motion capture with no IR markers, no cameras needed, just inertial sensors :)
Anyway, such pre-built solutions are usually well out of financial capabilities of average user, and building them yourself does need quite a lot of knowledge so let's all  be grateful to FT developers for a simple low-cost alternative :)
Kestrel #9 12/05/2008 - 15h56

Webmaster (admin)
Class : Webmaster (admin)
Posts : 780
Registered on : 13/07/2007

Off line

All thanks to micro-tech. Can't wait for it to go nano. Exciting stuff.

 >  Fast reply

Message

 >  Stats

1 user(s) connected during the last 10 minutes (0 member(s) and 1 guest(s)).