>Well, I suppose you´re right. But for equal distribution, you could just >sample an ordinary triangle oscillator, that runs with much higher >frequency than the sample rate (and out of sync of course). Wouldn´t a saw be better ? -Michael Buchstaller