On Mon, 2012-04-02 at 11:36 -0700, Jonathan Wilkes wrote: > >________________________________ > > From: Martin Peach <[email protected]> > >To: Jonathan Wilkes <[email protected]> > >Cc: "[email protected]" <[email protected]> > >Sent: Monday, April 2, 2012 2:04 PM > >Subject: Re: [PD] pddp keywords > > > >On 2012-04-02 13:45, Jonathan Wilkes wrote: > >... > > > >> One example is the keyword "bandlimited": the definition currently is, > >> "object that describes itself as > >> bandlimited"! Thus, I only tagged those objects which called themselves > >> band limited in their description, > >> and this obviously has left out many objects. So what is a decent, short > >> definition for bandlimited? > > > >All frequency components are within a specified bandwidth. > > > But with something like [creb/blosc~] you don't specify the bandwidth. Is a > > different definition being used there?
I guess not. "Band-limited oscillator" in this context means that it creates a waveform that does not contain any partials (multiples of the base frequency) higher than the Nyquist frequency, which is half the sampling rate. It assumes a implicit band-limitation within O Hz and sr/2 (22050 Hz, if your sr is 44100). A plain (non-band-limited) [phasor~] doesn't care about the Nyquist theorem and happily plays partials well above the Nyquist frequency, which are reflected on said border and appear as aliases in the audible spectrum. So to say, a non-band-limited [phasor~ ] suffers from aliasing effects, which a band-limited or anti-aliased sawtooth generator does not. Roman _______________________________________________ [email protected] mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
