HOWTO: Home Made Head Tracker
Moderators: Board of Directors, Command
HOWTO: Home Made Head Tracker
Hi everyone..
I’d like to share with you my attempt at making a head-tracking device that utilises the Free Track 2.2 software.
When I first joined ACG I was using a piece of software called Face Track NoIR. This is quite a good piece of software considering it’s downloaded free on a donation basis. It works by using a standard web cam to track your facial features primarily eyes and nose, which is then interpreted by a software filter to produce head movement in the game.
The main issue I found with it was in some lighting conditions or with quick head movements this would cause the tracker to freeze, usually when being pursued by one of our black-crossed colleagues!! Not a very satisfactory state of affairs!
I looked a purchasing a Track IR but then read several articles about simmers making their own head trackers using some pretty standard items found around the house and a bit of ingenuity.
The main article that I used as reference can be found here:
http://www.therevolutiongaming.com/foru ... /post/last
I didn’t really like the look of the Lego head set the author made so I set out to produce something I liked but using the same wiring that was proven to work ok.
Bits I used were:
1) An old metal coat hanger
2) Ridged plastic tube 5mm internal diameter: B&Q, however an old Bic pen would work just fine.
3) Bell wire or something similar: Maplins
4) Soldering iron and solder: lead based; the other stuff is crap!
5) 3mm Heat shrink rubber: Maplins
6) Fine sand – budgie sand from a pet shop is ideal grade
7) 5mm internal diameter flexible silicone tube: borrowed off the good ladies new handbag. They were used as zip protectors in the packaging
8) 8-10mm cable clip with ratchet to allow the device to be removed off my head phones: local electrical shop
9) Insulation tape
10) Snipe nosed pliers and standard pliers
11) No More Nails adhesive tape
The finished tracker looks like this. I’ve put a picture at the start of the narrative to make the construction process a bit easier to understand.
Making the tracker frame:
Just like in the link I made a template from card with the correct spacing of where the LEDs need to be positioned. Once I was happy where the LEDs were going to fit I began to shape the coat hanger.
This is simply done by firstly cutting out the hanger section that you normally put on the clothes rail to leave a long section of wire. You can then use the snipe nosed pliers and pliers to form the headset frame based upon the template you have previously drawn. With a bit of luck you should end up with a form that looks something like this:
Making the plastic tube housing:
This is made from plastic tube with an interior diameter of 5mm. I got some from B&Q. if you want to keep the costs to a minimum then it will be possible to use a Bic pen instead.
This is the type of product I used:
The length of the plastic tube required is dependant on your individual headset design due to where you need to install the mounting clip. Mine worked out at about 75mm. This measurement is not critical but if you make it too short it will be more difficult to install the resistor inside the tube later.
In order to make a bend in the plastic pipe please see the videos on YouTube similar to this:
https://www.youtube.com/watch?v=-BYq6hENHsk
Its very easy to do. I made a small funnel from aluminium baking foil to pour the sand into the pipe. The end of the pipe can be blocked off with a second piece of aluminium foil as in the photo: and no, you can’t smoke it!!
I gently heated the sand filled pipe over a low gas ring until the plastic could be bent to an angle of 90 degrees. When the plastic has set and cooled you simply pour the sand out.
Wiring:
The head tracker is wired as per the link to the original article. The author signposts the reader to purchase the LEDs and resistor from mouser.com. I found the postage from this retailer to be extortionate so I took the shopping list to a local electrical dealer who got me the components directly and saved me a fortune.
The wiring is soldered directly onto the LEDs in series with the resistor soldered onto the live feed and contained within the plastic pipe to prevent damage. The positive and negative of the LEDs are positioned either side of the coat hanger wire and are finished with heat shrink insulation to prevent a short circuit like this:
The LED is then held in place with a piece of the 5mm silicone tubing. Like this:
Before joining the USB cable to the LED connecting wires I fitted another small piece of silicone tube to the plastic pipe to give it good purchase in the mounting clip like this:
With a bit of luck you should have completed all the complicated construction. The only remaining job is to securely bind the plastic tube where the USB cable enters and the wire frame exits giving you something that resembles this:
Mounting Clip:
The retaining clip for the tracker is secured with a small piece of No More Nails double-sided tape. You need to take care when doing this, as it’s important to stick it at the right angle of about 90 degrees to the vertical centre line of your head so the tracker is held at the correct angle:
When mounted to the head set clip the tracker looks like this:
Setting up with a web cam:
This was the longest and most frustrating part of the operation. In a nutshell it all boils down to individual lighting conditions in your room. I have a Microsoft HD web cam and didn’t want to purchase another as suggested in the article. I tried cutting out a filter from and old floppy disk as suggested but this wouldn’t work at all well with my set up. The results were incredibly shaky and prone to stop in any position despite changing the Free Track threshold adjuster.
I then tried a piece of old negative film cut to fit the lens of the camera. This set up let in too much light with the result the Free Track skull looked like it was head banging to Motorhead!!
The solution was to insert a second piece of negative film on top of the first piece. This worked very well and allowed me to be quite accurate with the threshold settings. There are quite a few videos on how to set up Free Track the one I used can be found here:
https://www.youtube.com/watch?v=jixqi4ev10o
Be prepared to be patient to get free track to work correctly. Make small adjustments and note the changes and keep a log so you can revert to the original setting if required. On some P/Cs when starting Free Track 2.2 you may get and error message that prevents the application from running. If so try the following:
1. Go into the Desktop of your P/C
2. Go into personalise
3. Go into display
4. Go into “set custom text size (DPI)"
5. Make sure the DPI setting is at 100% and not 125%
6. Ok the change
Free track should start ok. If not re-boot and check the DPI setting again and ensure you save the setting.
Conclusion:
There is a wealth of information out there on how to make tracking devices. I recommend good research before starting the project. Pay particular attention to correct wiring if you are going to use a different scheme to that in the article. If you are concerned about using a USB cable it would be possible and safer to substitute this for a battery pack. There are many articles to help you calculate the power required.
Here are some of the reference articles I used for the project:
http://www.free-track.net/forum/index.p ... 466&page=1
http://www.free-track.net/english/
http://facetracknoir.sourceforge.net/home/default.htm
If you have any questions I’d be happy to try and answer them either on Team Speak or the forum.
I would like to thank anyone I linked to in the narrative for some excellent work that has benefited the entire gaming community. Now.. If we can just make a home made Oculus Rift…
Regards
Kev
I’d like to share with you my attempt at making a head-tracking device that utilises the Free Track 2.2 software.
When I first joined ACG I was using a piece of software called Face Track NoIR. This is quite a good piece of software considering it’s downloaded free on a donation basis. It works by using a standard web cam to track your facial features primarily eyes and nose, which is then interpreted by a software filter to produce head movement in the game.
The main issue I found with it was in some lighting conditions or with quick head movements this would cause the tracker to freeze, usually when being pursued by one of our black-crossed colleagues!! Not a very satisfactory state of affairs!
I looked a purchasing a Track IR but then read several articles about simmers making their own head trackers using some pretty standard items found around the house and a bit of ingenuity.
The main article that I used as reference can be found here:
http://www.therevolutiongaming.com/foru ... /post/last
I didn’t really like the look of the Lego head set the author made so I set out to produce something I liked but using the same wiring that was proven to work ok.
Bits I used were:
1) An old metal coat hanger
2) Ridged plastic tube 5mm internal diameter: B&Q, however an old Bic pen would work just fine.
3) Bell wire or something similar: Maplins
4) Soldering iron and solder: lead based; the other stuff is crap!
5) 3mm Heat shrink rubber: Maplins
6) Fine sand – budgie sand from a pet shop is ideal grade
7) 5mm internal diameter flexible silicone tube: borrowed off the good ladies new handbag. They were used as zip protectors in the packaging
8) 8-10mm cable clip with ratchet to allow the device to be removed off my head phones: local electrical shop
9) Insulation tape
10) Snipe nosed pliers and standard pliers
11) No More Nails adhesive tape
The finished tracker looks like this. I’ve put a picture at the start of the narrative to make the construction process a bit easier to understand.
Making the tracker frame:
Just like in the link I made a template from card with the correct spacing of where the LEDs need to be positioned. Once I was happy where the LEDs were going to fit I began to shape the coat hanger.
This is simply done by firstly cutting out the hanger section that you normally put on the clothes rail to leave a long section of wire. You can then use the snipe nosed pliers and pliers to form the headset frame based upon the template you have previously drawn. With a bit of luck you should end up with a form that looks something like this:
Making the plastic tube housing:
This is made from plastic tube with an interior diameter of 5mm. I got some from B&Q. if you want to keep the costs to a minimum then it will be possible to use a Bic pen instead.
This is the type of product I used:
The length of the plastic tube required is dependant on your individual headset design due to where you need to install the mounting clip. Mine worked out at about 75mm. This measurement is not critical but if you make it too short it will be more difficult to install the resistor inside the tube later.
In order to make a bend in the plastic pipe please see the videos on YouTube similar to this:
https://www.youtube.com/watch?v=-BYq6hENHsk
Its very easy to do. I made a small funnel from aluminium baking foil to pour the sand into the pipe. The end of the pipe can be blocked off with a second piece of aluminium foil as in the photo: and no, you can’t smoke it!!
I gently heated the sand filled pipe over a low gas ring until the plastic could be bent to an angle of 90 degrees. When the plastic has set and cooled you simply pour the sand out.
Wiring:
The head tracker is wired as per the link to the original article. The author signposts the reader to purchase the LEDs and resistor from mouser.com. I found the postage from this retailer to be extortionate so I took the shopping list to a local electrical dealer who got me the components directly and saved me a fortune.
The wiring is soldered directly onto the LEDs in series with the resistor soldered onto the live feed and contained within the plastic pipe to prevent damage. The positive and negative of the LEDs are positioned either side of the coat hanger wire and are finished with heat shrink insulation to prevent a short circuit like this:
The LED is then held in place with a piece of the 5mm silicone tubing. Like this:
Before joining the USB cable to the LED connecting wires I fitted another small piece of silicone tube to the plastic pipe to give it good purchase in the mounting clip like this:
With a bit of luck you should have completed all the complicated construction. The only remaining job is to securely bind the plastic tube where the USB cable enters and the wire frame exits giving you something that resembles this:
Mounting Clip:
The retaining clip for the tracker is secured with a small piece of No More Nails double-sided tape. You need to take care when doing this, as it’s important to stick it at the right angle of about 90 degrees to the vertical centre line of your head so the tracker is held at the correct angle:
When mounted to the head set clip the tracker looks like this:
Setting up with a web cam:
This was the longest and most frustrating part of the operation. In a nutshell it all boils down to individual lighting conditions in your room. I have a Microsoft HD web cam and didn’t want to purchase another as suggested in the article. I tried cutting out a filter from and old floppy disk as suggested but this wouldn’t work at all well with my set up. The results were incredibly shaky and prone to stop in any position despite changing the Free Track threshold adjuster.
I then tried a piece of old negative film cut to fit the lens of the camera. This set up let in too much light with the result the Free Track skull looked like it was head banging to Motorhead!!
The solution was to insert a second piece of negative film on top of the first piece. This worked very well and allowed me to be quite accurate with the threshold settings. There are quite a few videos on how to set up Free Track the one I used can be found here:
https://www.youtube.com/watch?v=jixqi4ev10o
Be prepared to be patient to get free track to work correctly. Make small adjustments and note the changes and keep a log so you can revert to the original setting if required. On some P/Cs when starting Free Track 2.2 you may get and error message that prevents the application from running. If so try the following:
1. Go into the Desktop of your P/C
2. Go into personalise
3. Go into display
4. Go into “set custom text size (DPI)"
5. Make sure the DPI setting is at 100% and not 125%
6. Ok the change
Free track should start ok. If not re-boot and check the DPI setting again and ensure you save the setting.
Conclusion:
There is a wealth of information out there on how to make tracking devices. I recommend good research before starting the project. Pay particular attention to correct wiring if you are going to use a different scheme to that in the article. If you are concerned about using a USB cable it would be possible and safer to substitute this for a battery pack. There are many articles to help you calculate the power required.
Here are some of the reference articles I used for the project:
http://www.free-track.net/forum/index.p ... 466&page=1
http://www.free-track.net/english/
http://facetracknoir.sourceforge.net/home/default.htm
If you have any questions I’d be happy to try and answer them either on Team Speak or the forum.
I would like to thank anyone I linked to in the narrative for some excellent work that has benefited the entire gaming community. Now.. If we can just make a home made Oculus Rift…
Regards
Kev
Re: HOWTO: Home Made Head Tracker
Great tutorial Kev !!!
Just one question for you. How is it working when you look at the wing direction? Is it smooth ?
I'm using FT for years and my first clip was similar like yours. But I had problems because when I looked over the wings,
FT just couldn't follow well because at that angle even if you moving your head a camera just don't "see" enough movement.
I solved that with remodeling my clip a bit to look exactly like TrackIR Clippro so the LEDs won't be in line.
That was much better, but I finally made another model like cap( geometrically) and got better resolution on yaw movement.
Another thing, how to choose the right dimension for a model ?
That depends on your camera lens angle and on your distance from it.
The best results you'll get if the distance between the dots on the FT window are around 1/3 of the window.
Just one question for you. How is it working when you look at the wing direction? Is it smooth ?
I'm using FT for years and my first clip was similar like yours. But I had problems because when I looked over the wings,
FT just couldn't follow well because at that angle even if you moving your head a camera just don't "see" enough movement.
I solved that with remodeling my clip a bit to look exactly like TrackIR Clippro so the LEDs won't be in line.
That was much better, but I finally made another model like cap( geometrically) and got better resolution on yaw movement.
Another thing, how to choose the right dimension for a model ?
That depends on your camera lens angle and on your distance from it.
The best results you'll get if the distance between the dots on the FT window are around 1/3 of the window.
Re: HOWTO: Home Made Head Tracker
Hi Vranac,
Thanks for your thoughts and comments…
That’s a really good idea about offsetting the IR’s. I don’t really get an issue looking in any vector but I can see the benefit of the offset. Occasionally I get a slight freeze if I’m flailing around trying to look behind but other than that its very accurate. I might try bending the top two arms slightly to give the same effect and report beck.
Regards
Thanks for your thoughts and comments…
That’s a really good idea about offsetting the IR’s. I don’t really get an issue looking in any vector but I can see the benefit of the offset. Occasionally I get a slight freeze if I’m flailing around trying to look behind but other than that its very accurate. I might try bending the top two arms slightly to give the same effect and report beck.
Regards
Re: HOWTO: Home Made Head Tracker
Very cool stuff Kev
and Vranac
and Vranac
Re: HOWTO: Home Made Head Tracker
Great post! Hopefully I'll make one soon!
Re: HOWTO: Home Made Head Tracker
HI,
Sadly, the link to the TRG forums is no longer useful as a cross reference.
After joining this site, a search through the forums yielded no references to TrackIR or Head Trackers in general.
Do you have an alternate link available or a circuit diagram which details component technical specifications and wiring?
Thanks in advance
Archangel
Sadly, the link to the TRG forums is no longer useful as a cross reference.
After joining this site, a search through the forums yielded no references to TrackIR or Head Trackers in general.
Do you have an alternate link available or a circuit diagram which details component technical specifications and wiring?
Thanks in advance
Archangel
"Never argue with an idiot - they only drag you down to their level and then beat you on experience" (Unknown)
Re: HOWTO: Home Made Head Tracker
Hi!
I've made it work with FreeTrack, it works pretty well at first. But when I switch into the game it changes my camera settings so that it doesn't catch the LEDs...
Any ideas?
Thanks
I've made it work with FreeTrack, it works pretty well at first. But when I switch into the game it changes my camera settings so that it doesn't catch the LEDs...
Any ideas?
Thanks
Re: HOWTO: Home Made Head Tracker
Hi Woop, I don't think that the game is changing your camera settings. I don't know which camera you're using but I suspect that exposure is set to auto in the settings, try to put it the lowest you can and than switch off auto.
Re: HOWTO: Home Made Head Tracker
Thing is my camera properties window that you picture (not the camera software) only appears blank for some reason. No idea why...Vranac wrote:Hi Woop, I don't think that the game is changing your camera settings. I don't know which camera you're using but I suspect that exposure is set to auto in the settings, try to put it the lowest you can and than switch off auto.
Re: HOWTO: Home Made Head Tracker
Well, it seems that Logitech cams aren't the best choice for FT. Try this workaround in post #8 from chubbyrain.
http://www.free-track.net/forum/index.p ... topic=2947
http://www.free-track.net/forum/index.p ... topic=2947