Step 4Modify the camera
In order for the PS3 Eye camera to work, the IR filter needs to be removed and an IR bandpass filter needs to be put in its place. The IR bandpass filter ensures that only the IR light with the utilized wavelength makes it into the system. Therefore it is important that the IR bandpass filter's centering wavelength matches that of the IR LEDs present in the IR strip used for illuminating the surface. A video guide created by Peau Productions for modifying the PS3 Eye camera can be found below.
| « Previous Step | Download PDFView All Steps | Next Step » |
























![Enable mouse and keyboard functionality on an iPod Touch or an iPhone! [Linux]](http://img.instructables.com/files/deriv/F09/VY6U/G9GQGXFB/F09VY6UG9GQGXFB.SQUARE.jpg)



















My question is, won't this shift cause the camera to be out of focus?
The wiimote is an active IR receiver and only works when IR light is shined directly into it.
Wrong. The IR camera on a Wiimote will pick up any IR light source available. Warm light-bulb reflections in glass are enough to register.
Also the wiimote's bluetooth interface will not be fast enough to get very smooth movement.
Wrong. Having used the bluetooth connection for whiteboard and head-tracking, it is more than fast enough for real-time tracking of blobs. (Think about it, it's more than fast enough to track the games, isn't it?)
For a table like this the contrast is not nearly enough to use a cheap receiver like a wiimote.
I think it could work, considering some of the very faint IR sources I've used in the past. If someone in the Willamette Valley (OR) has an FTIR, I would be happy to test this with you. (ftir-wiimote@piratesofpacifica.com)
The real limitation for using a Wiimote as your IR camera is that the Wiimote pre-processes the blobs, and is limited to tracking 4 blobs at a time. (The Wiimote simply passes 1-4 X,Y coordinate pairs, measuring blob strength from 1-6. This means:
1> You'll never get more than 4 touch points.
2> You lose all ability to calibrate the camera/blobbing algorithm.
I do not recommend trying to build a table using a Wiimote camera, but if you have one lying around already, and you don't mind taking the time to hook it up (GlovePIE?), maybe you can make it work.
You are right that a wiimote will work with any IR source since you can calibrate the sensor. Also you are right that the bluetooth interface is fast enough, since the wiimote technically only has a resolution of 164x124 or 352x288.
The problem lays in this resolution. A resolution like this is fine when you are playing a video game in 480p, but when you are trying to run a computer with 1080p resolution this is a problem. Additionally you are touching directly on the screen, whereas in normal operation of a wiimote you are at least 8 feet away from the screen.
The PS3Eye camera, which is the same price as a wiimote, gives you 640x480 resolution at 60fps. This speed and clarity allows for very accurate and fluid movements on your multi-touch surface. However this resolution is still not enough for most professional systems, and many are holding our for the new point grey usb 3.0 camera which will offer 1080p at 30fps.
Again to reiterate, a wiimote could and probably would work with some tedious calibrations, but if you are already spending upwards of $2000 to build a professional quality surface (like the one detailed in this guide) then why would you skimp on the $35 camera purchase.
The goal of the bandpass filter is to only allow the correct "color" (frequency) of IR radiation (light) through to the camera's sensor. This is especially important because we're projecting a lot of visible light at the rear projection screen, and that will totally drown out any IR touch-blobs from the FTIR (IR touch screen).