Hi Steven. That's what I usually do, just set ksmps, but I have always instinctively tried to make it a factor of sr. No longer! -Chuckk On 10/20/07, Steven Yi wrote: > > Hi Chuckk, > > Just to note, you don't need to set kr and I recommend not bothering > with it anymore as it really just is used to find ksmps and if you set > ksmps already then you're causing more work to make sure ksmps and kr > match. Better just to set ksmps and not set kr and let csound do it > for you. > > steven > > On 10/20/07, Chuckk Hubbard wrote: > > Just a little more clarification on this... > > If I understand correctly, > > > > ksmps must be an integer (html/ksmps.html) > > -b should be a power of two (html/UsingRealTime.html) > > -b should be a power-of-two multiple of ksmps > (html/UsingOptimizing.html), > > or at least, for duplex audio, a multiple of ksmps ( > > htmlCommandFlags.html#FlagsMinusLowerB) > > > > Two questions- > > Are all of these statements up-to-date? and, > > Is it okay for KR to be non-integer? > > > > I'm guessing non-integer KR is okay, since the docs don't say that it's > NOT > > okay; and since a second is kind of an arbitrary measurement for things > that > > happen that fast. That is, I don't know of any inherent need for audio > > buffers to line up with seconds. But with sr=44100, 4 is the highest > > possible value of ksmps that a) satisfies all of the musts and shoulds > > above, and b) equates to an integer KR. > > > > -Chuckk > > > > -- > > http://www.badmuthahubbard.com > -- > Send bugs reports to this list. > To unsubscribe, send email to csound-unsubscribe@lists.bath.ac.uk > -- http://www.badmuthahubbard.com