performance limit in a sports game?
Posted: Tue Dec 31, 2002 6:50 pm
Hi, I am considering purchase of trackIR for sports game use (golf, namely; but also want to use it for any game where I can swing a stick), which has a consierably higher cursor speed rate. In particular, the sample rate and latency become problems. I have questions on "playability":
Q1: what is the expected sample rate range when the system is not hogged by some other apps (namely the game itself)?
For example, I wrote a test code with an open source image processing library (OpenCV) to track the brightest spot my $10 webcam could see on 800 MHz PIII system, and I got 13~23 frames per second with in debug mode, with real-time rendering. The webcam itself can only do up to 30 fps, so in this case, I would rate my test software at 20 fps, to be conservative.
Q2: what is the latency range?
That is, how many milliseconds will elapse between the time when say your wand was at some point, till the time when the mouse move is posted? Since both throughput and latency depend on system, I would be most obliged just to see numbers on your system.
Q3: do you have any plan to put 1 more LED on the other end of a stick, or sell the LEDs separately so that I can put 3 of them on a stick, and your software can track the 3 points and resolve it as a line? That would give me so much more information than just the current 2-dimensional information. With the 3 points (or even just 2), I would think trackIR can be used in a much value added stick-oriented-game.
Q4: related to Q3, but do you have any plan to put a wide-angle lense on trackIR so that I don't have to stand 20 ft from the camera for it to see my entire golf swing?
Thank you very much for your help.
Henry Choi
Q1: what is the expected sample rate range when the system is not hogged by some other apps (namely the game itself)?
For example, I wrote a test code with an open source image processing library (OpenCV) to track the brightest spot my $10 webcam could see on 800 MHz PIII system, and I got 13~23 frames per second with in debug mode, with real-time rendering. The webcam itself can only do up to 30 fps, so in this case, I would rate my test software at 20 fps, to be conservative.
Q2: what is the latency range?
That is, how many milliseconds will elapse between the time when say your wand was at some point, till the time when the mouse move is posted? Since both throughput and latency depend on system, I would be most obliged just to see numbers on your system.
Q3: do you have any plan to put 1 more LED on the other end of a stick, or sell the LEDs separately so that I can put 3 of them on a stick, and your software can track the 3 points and resolve it as a line? That would give me so much more information than just the current 2-dimensional information. With the 3 points (or even just 2), I would think trackIR can be used in a much value added stick-oriented-game.
Q4: related to Q3, but do you have any plan to put a wide-angle lense on trackIR so that I don't have to stand 20 ft from the camera for it to see my entire golf swing?
Thank you very much for your help.
Henry Choi