FreeTrack Forum

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

RSS >  Mounted tracker on HMD, Mounted tracker on HMD
robt #1 06/02/2008 - 03h30

Class : Apprenti
Posts : 6
Registered on : 06/02/2008

Off line

Hi,

I tried to post before, but it didn't post. I can't read french, all the buttons are in french, so hopefully i do this one right.

Thanks for a great program. I was about to sell my eyetrek glasses since i had no need to use them. I had actually posted the for sale at at craigslist when, about an hour later, just by chance, i found your site. I took the posting down right after reading a bit..

Anyways, the tracker works great. I have a few questions, if anyone knows, some feedback would be great.

I used 3 IR transmitters that i bought from radio shack. I sanded them down and mounted them.

I wanted to know, does the surface that was sanded need to be polished to get optimal signal level from the transmitters?

Also, does the sanded surface need to be flat? I just noticed that two of them are at a slight angle, and one of them actually has two "facets" from sanding them at slightly different angles.

Thanks,

Robert Trujillo

P.s.

Annuler must mean cancel...
didja #2 06/02/2008 - 05h59

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

Off line Www

Hello ! :)

You can choose English as your prefered language : "Données personnelles" -> "Parmaètres d'affichages".

Beside, english is now the default language.

Rémy
Posted Image
tristan68 #3 06/02/2008 - 08h09

Electronicien en chef
Class : Electronicien en chef
Posts : 651
Registered on : 10/07/2007

Off line

Hello Robt, and welcome to the fun.


1)
Using a forum written in other languages shouldn't be an issue today.
There are enough online translators, allowing you to process the entire webpage.
The results are often sorta funny, but still readable  :dry: .

So you shouldn't have had probs to use the interface, even with french labelled buttons.

2)
This issue is solved today , since the whole forum interface is avail in english.

3)
Yes : annuler means cancel  ^^

4)
Glad to hear your setup works fine, we are eager to have some feedback from people using freetrack and video goggles or glasses.

5)
If one of your led has 2 facets , chances are you'll get an "artifact" at some angles (the webcam will see 2 leds very close eachother).
It's better to sand the leds really flat. But this has to be adapted, based on your personnal experience. If you don't have this artifact problem, so better not change your setup.

6)
IMO, it's better if the surface is still a bit rough, no need to polish it.
Since you'll loose forward brightness, the lateral brightness could become an issue.
If that's the case, your can paint the border of your leds or wrap them in electrical tape, so only the flat top of the led can emit IR.

Hope this helps.


Cya.
robt #4 08/02/2008 - 06h06

Class : Apprenti
Posts : 6
Registered on : 06/02/2008

Off line

Thanks for the info. I took most of that into account, i've been playing with different 3 point mountings for a few days with those points in mind.

One thing that i keep running into- the software expects the center (top) LED to be further from the camera than the side LEDs. I noticed that in the setup, there is no way to specify a negative number (so that i could mount the top, or center, LED closer to the camera than the side LEDs).

The only reason i mention it..  it would be simpler both physically and anatomically to mount the center LED closer to the camera and the side LEDs slightly further back. I don't know anything about the math involved, so i'm sure there might be valid reasons...

It just seems like it would be easier to mount the two lower LEDs further away being on the sides of the head , since mounting the top LED further away from the camera means going on top the head, which is (at least imo) an awkward place to get to. If you don't mind a hat, it's not that big of a deal. Most of the mounts i've been trying out are modeled on glasses, so the framework exists on both sides of the head to mount LEDs to..

Would it work mathematically to allow negative numbers in the setup, so that having the top LED mounted further toward the camera than the side LEDs?

Anyways, a minor observation. The more i use the program the better it seems..

Thanks!

Rob T.

BTW suddenly everything is in english...
tristan68 #5 08/02/2008 - 07h34

Electronicien en chef
Class : Electronicien en chef
Posts : 651
Registered on : 10/07/2007

Off line

Ouch !

questions about the setup geometry and program algorithms.....  :pinch:

Thats a job for Kestrel or The_Target

(Tristan running away  :p )

Sorry i'm no help for you for those questions, but i'm sure you'll get an answer as soon as Kestrel or Target see this post.

cya
Kestrel #6 08/02/2008 - 13h18

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

Off line

robt @ 08/02/2008 - 07h06 a dit:

the software expects the center (top) LED to be further from the camera than the side LEDs. I noticed that in the setup, there is no way to specify a negative number (so that i could mount the top, or center, LED closer to the camera than the side LEDs).



If the model's center point is forward of the other two, it only changes the yaw and pitch direction, which can be accounted for in the software by inverting those particular axes. Note that this configuration is less ideal because instead of two points moving a lot with respect to one reference point that moves little, you have two points that don't move much and one point that moves a lot. This means less information is available because you're relying on the movement accuracy of one point instead of two, all while subject to the ubiquitous camera noise.
robt #7 11/02/2008 - 04h42

Class : Apprenti
Posts : 6
Registered on : 06/02/2008

Off line

Thanks for the reply, that actually helps me with some other questions i had about optimum placement for the LEDs.

If there is one led near the top of the head, it's going to be pretty near the Y axis (up/down) center of rotation. Since the two leds on the sides have to be placed further forward and lower than the center LED, they're not so close to a center of rotation and so move more, which means the computer has more data about what's happening. That makes perfect sense, if i have it right.

Thanks,

Rob Trujillo

 >  Fast reply

Message

 >  Stats

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