You get a laggy video when you make a video with free version of FRAPS.

Actually, I can tell the difference between laggy video and input lag. Not saying it's there, just that it looks like it is by the slow deliberate movements when I get near instant response with a TIR5.

Gotcha. It could be many things. Anyway, I have Freetrack and I don't get the lag you see in the video. It is smooth and I use Logitech webcam. If there is any input lag, I don't percieve it. Also, you can use wiimote if input lag is a concern. Regarding drift, I don't get drift. I center it up once when I start playing and I'm good for the duration provided I don't shift my sitting position. A steady chair is a must. Also, I made my mouse cursor visible so I always know my vision center. But that is personal taste.

Absolute mode is the default mode in FT.

It's impossible to judge input-output lag when you can't see the person's head movement. The person might just "move like that", or they may be compensating for excessive input sensitivity, or they might have too little sensitivity, or they may just be making slower movements to create an easy to watch video.

I wouldn't be surprised if TIR has faster and smoother responses than FT (I hope so), however FT gives you a smoothing slider, which allows you to set your preferred compromise between instant response and smooth response.

FT uses a x4 multiplier for the data captured by a standard 30fps webcam to overcome the lower frame rate. In use I havent really noticed much difference in performance using a 30fps webcam with FT compared to my 120 fps TIR4 kit. A PS3 cam or a Wimote can be used with FT to have 120fps tracking (from what I have read).

I know what you mean, it is difficult to tell without a live shot of the guy's head. Maybe he's just not using quick and precise head movements in this demo, which is often a symptom of input lag which is why I brought it up.

People perceive input lag differently, like frame rates. I see lots of folks who can't consciously tell the difference between 25fps and 60fps but it shows on how slow they tend to react since it takes that extra split second to get their bearings after spinning around.

Anyway, not having used FT I wouldn't judge it's performance, I just question certain aspects of performance and precision since it lacks the dedicated hardware advantage of TIR. It may be as some say that the extra cpu load is not significant enough to make enough difference to justify the cost of a TIR.

I've never really found much use for a webcam but I may pick one up sometime to try out FT for myself and compare.

I'm actually thinking of getting one of those PS3 cams, i'm just waiting to see if CoD will support freetrack first.

I had a trackIR4 which died, i tried FT as an alternative with my standard 30 FPS webcam and the trackclipPro that i already had and i did notice it's slower to react than trackIR. I wouldn't mind getting a used PS3 cam with 120FPS if CoD supports alternative head trackers.

With a current mid-range PC the CPU load of FT is negligible. I think it's the 30FPS webcams that cause this slight delay, plus my own has a somewhat narrow field of view which further complicates things: i sometimes need to move around in weird angles to compensate for my crappy profiles and then it loses track of the dots at certain points :-P

What i'd really like in Freetrack is an interface more like the one used by trackIR for setting up the response curves, as i can't seem to get a good profile for FT no matter what i try. As for the rest, it is a very capable alternative and i guess that using a 120FPS camera would get rid of most of the slight delay between actual and in-game motion.

why don't you just get the real thing, instead of a sponge that offers less than optimal performance?

everytime somebody talks about freetrack you come in and say your idiocy...
why don't you go to hell and shut your mouth? you are only a spammer

Or you could stop taking it so personally and just not reply.

That x4 multiplier quantizes between samples taken from the lower framerate of the webcam which enhances accuracy the same as "smoothing" on a mouse or TIR. The unfortunate trade-off for using a quantizing algorithm is input lag response which is unavoidable.

With TIR for example, I keep the smoothing fairly low, around 15% because the benefit of a little more precision is negated by the lag it introduces. It's not as noticeable on newer mice because of the insane sampling rates they work at now don't really require it.

You may not think you notice it but it definitely affects the way you pivot around in-game, even by milliseconds. You can equate it to playing an online FPS shooter game at 50ms vs 200ms, not as a direct comparison but to illustrate that sub-second lag can have a big impact on your effectiveness.