I think it needs to a hardware filter, or physical filter as you said. As I said before, I just used the end of an exposed bit of film. Had to dig through the old photos seem evrything is digital these days.
And remember that this setup works best with infra red led's, not just red ones. Mice and remotes use the infra red spectrum.
I had no drama's with any amount of light due to the ir filter. In direct sunlight it may make it not so perfect but inside a house, lights on or off, it worked well.
This technology has been around for so long. You can buy cheap sensors (accelerometers) to do this. I bought some for work for under $40 australian dollars that measures 3 dimensions.
But the problem is these types of motion detection (acceleromter type and the wii controller) work on acceleration so a slow movement won't be detected, only quick movements and would still think you're in the same location. The freetrack setup senses absolute position rather than acceleration so if you move slowly or it misses some info, then when you return to the centre freetrack still realises you are in the centre.
If you already have a webcam then chances are it'll work. The exposed film is a ir filter and filters everything out except infrared. Infra red led's are cheap, even use a mouse to test it seem it gets power from the usb. I pulled apart a spare mouse and stuck the led on a baseball cap's peak to test it. One LED worked surpisingly well for me. Cost me nothing except for time to pull the mouse apart.
As I said, worked really well with hardly any setup. Can't see the reason to buy a track ir if you already have a webcam. My webcam lense is quite wide so it covered a fair range of head movement. As I said LFS worked straight away and detected the freetrack for looking around.
You can also use a normal remote in front of the camera with the ir filter on the camera's lense just to see how it works. Problem here is the ir beam can be narrow and needs to point at the webcam and it also flashes so not ideal but easy to see if it works.
For those asking questions, read the freetrack site. Works on the same principle as the trackIR. Not exactly new technology.
Look at a program called freetrack. I used it with my standard webcam with exposed camera film over it and it worked really well with no setup needed in lfs.
Was pretty cool and easy to setup......But then got over it pretty quickly though.
I used an optiical mouse as the infrared source just to test it out. Only need one infra red led to be able to look left, right, up and down. Read the freetrack website for more info on degree of freedoms.
I have now found that the wheel lock and compensation seem to make a difference on the turn axis when adjusting the options while driving but for some reason they don't show change on the axis when in the main menu settings.
I am now experimening with a lower steering lock. Cheers
Well I have really only raced the fxo gtr mainly around blackwood gp and can get 1.10's and consistent 1.12's without too much difficulty but I feel like I need to be extra carefull on the stick for just about every corner. I have a lock of 20 degrees, this may not help by the sound of it.
When I change the wheel turn compensation from 1.00 to 0.00 it doesn't seem to alter the steering bar at the bottom of the screen nor does adjusting the steering wheel turn from 90 to 540. Should these show on the bar in the control setup? I have tried clicking ok and coming back in but ti doesn't appear to be any different. LFS reports the controller as xbcd xbox 360 gamepad.
lil chris: Yes the mouse pointer centres when you centre the stick so the steering works well in pinnacle. Don't need to push the other way to get centre again. Just make sure you map the stick to mouse with spring back. Let me know how you go with acc and brake if there is a work around the digital limitation when steering with mouse.
BYN:
Yes I have both analog smooth and compensation at the extremes. My steering locks may be high but sometimes I feel you need the lock to get out of some situations.
I too thought the lower deadzone value would make it less sensitive. In a way it does but the xbox 360 controller I have has a small amount of area where the resistance (physical) is low. Making the deadzone bigger means when the stick gets out of the deadzone there is a fair amount of resistance (physical) to movement.
I thought I would give pinnacle a go after using xbcd. First impressions seemed good. Xbcd doesn't have sensitivity but pinnacle seemes to. Sensitivity only appears to work in lfs when the joystick is assigned as a mouse. Using "spring back" the mouse centres when the stick centres. So all looks good except..........
LFS will only let you assign buttons or keyboard for acc and brake when using mouse steer so...... Damn it, only digital. Pinnacle lets you assign controller commands to keyboard or mouse but this doesn't get around the digital part... Back to xbcd.
I have recently made my dead zones a little bit larger and now the controller is easier to use due to going into the stiffer part of the thumb stick before it starts to steer much. I was thinking if I set it really low then it would be easier to control but that didn't suit me.
Still not ideal but seems the best for corded controller. Hopefuly xbcd will come out with a non linear response curve for the sticks but I doubt that will happen in a hurry.
Use xbcd and you can assign each trigger to a seperate axis. Rumble support works although lfs uses the force feedback inputs. Works perfectly on my computer with xp and xbox 360 controller. Only problem is the sticks are are still a bit sensitive with steering comp on 0.95. You get used to it though. I agree a wheel would be better due to more precise movement.
There is a good section somewhere in this forum, just search for it, that will help for xbox 360 controllers but here is the link for what works for me.
I tried the webcam with IR lamp. Works surprising well with no tweaking using freetrack.
Used a standard webcam with the dark bit of film at the end over it from an exposed film.
Pulled apart a spare IR dell mouse as I have a few of them. Just plugs into the usb port for power. Just gotta disable the mouse in control panel otherwise the mouse still works. Will be removing the lamp and connecting it to a usb lead so the whole board doesn't need to be there.
Don't know how much I will use it but I was surprised how wqell it worked with things I already had at home.
I have a dell d630 I am going to play LFS on. It has a nvidia quadro 135m graphics card with 2 gig of ram and a T7700 processor.
Now with all the settings for graphics turned up in LFS with resolution of 1440 x 900 I get hot lap frame rates that peak at 59.7 fps and pretty much stay above 50 the whole time. Now I have tried changing the aa settings and all the other settings mentioned here and the fps nor the picture quality change. I have tried using the nvidia control and also the Nhancer control. I have tried turning the antialiasing on 16xQ and anistropic on 16 but it doesn't make any difference in fps or image to when they are turned off.
I am running the latest dell drivers.
Any idea why the settings don't change when in game? I have tried changing global settings and also program specific.