gmeter sample and result resolution?

[kcleung]

hello

I am interested in gmeter but would like to find out how often does it samples the accelometer and how much detail can I see from the output (and is there a limit to the duration of a session?

[billy]

Sessions are unlimited. It samples at 25Hz and can talk at about 1Hz. Given on a track or while driving fast, you need your eyes on the road :), the 1Hz talk rate is more useful. It only talks when the G number changes by .1G which is a compromise on talk speed and useful information.

I'd recommend getting a Griffin Window seat windscreen mount and then hook the iPhone/iPod Touch up to the radio so you can hear it better.

[kcleung]

hi billy

thank you for the info. my runs typically will last for 5-10mins but what I really am interested is how smooth my driving is. so sample rate of 25hz should do the job fine but will I be able to see any jerky movement in my driving with gmeter? and as you might imgaine, I would be more interested in post session analysis as well as on the spot. the speaking function is definitely useful. appreciate your input on this.

[billy]

It seems like you're looking for a chart of your driving during the 'session'. That capability isn't in the G-Meter application but we're working on something right now which should give you exactly what you are asking for. That said, I'll see if we can add a chart of the session to the G-Meter

[kcleung]

thats great, I look forward to the new features then.

[shurtz]

Is there any chance that you will could generate a tone that varies with the g-force. The voice is so slow. It seems like it might be possible to update a tone as often as the accelerometers updates. It would be nice to hear how smoothly the g-force changes through a turn.

[Eric]

That's a good idea. Do you think there should me a tone every time the reading changes by a tenth of so, or that there should be a regular patern of sound and the tone simply changes with h reading?

[shurtz]

I think that it should sound like a continuous tone, but maybe it is silent below a threshold. Making this work well will be a very subjective thing. I don't know what will work best. It will require some experimentation. I would think that as the g-reading increases the pitch should increase, but it may work well with an inverse g to pitch relationship. I think care should taken to find a pitch range for the tone that is not annoying. You may be able to put a couple user options that help your experimentation.

[kbrew8991]

and if you hear the pitch rising and falling rapidly that may help someone realize how jerky they're being with an input...

thats a pretty cool idea actually, I could see that being useful in an instructing situation :)

[Warpedcow]

I would recommend doing something like a one octave pitch change going from 0G to 1G. Or maybe one octave for 0G to 0.5G and another octave for 0.5G to 1G, etc. Let the user define the 0G pitch and work from there.

Remember that pitch is exponential (concert A = 440hz, one octave up is 880hz, another octave up is 1760hz, etc).