Support |
> > I don't have my patch in front of me now, but I think it was 100hz. I > actually asked Jeff this, and > he said it made no difference, based on testing with different > frequencies. a higher frequency would be easier to filter out. ..but too high and you'd want to interpolate the peaks. I'd go into the kilohertz range. ....but as it works anyway ;-) > All you need is a > wave...square, saw, sine...doesn't matter...just something to generate a > signal and level that can > be measured by max. He's been using this method for years and has wrote > papers on its > efficiency. I have not percieved any delay in the signal. well, if you didn't perceive a modulation of the signal at 100Hz there must be some delay, as to filter out the 100Hz component would certainly cause one. mind you only around 10mS or so. In the same way that using a compressor with a fast setting on a bass note can cause distortion there's a play-off. Interesting technique, could be really handy for someone with a minimal setup. ..and at least as good as Midi with it's 128 steps which also need filtering. > The > conversation process from signal level > range to parameter range is so unbelievably simple and requires very > little code in max. take absolute value and lo-pass it. ( do-able in Bidule, probably others too ) ...or something a bit more complex with hold times and stuff to really get the fastest response. andy butler