[chuck-users] R: ResonZ default freq

Kassen signal.automatique at gmail.com
Tue May 22 14:53:01 EDT 2007


Resonant filters can be very loud, silence is a safe default.

Yes, yes, but at a Q of 1 (it's default) it isn't especially unruly or loud.
ResonZ is a bi-quad, not a 303 set to "accent" ;-), especially at it's
default Q of "1".

Moreover, defaulting to any common (or uncommon) note could hide a missing
> initialization, possibly much later ("Hey, my complicated filter should be
> tuneable and it isn't! It's a bug! Help me!").

I suppose so, on some level, but if you're building a complicated filter
yourself you likely aren't accidentally adding in a ResonZ set to zero Hz
either, nor does filters set to 0Hz make tuning anything much easier. If
you're doing that you'll likely using adding BiQuad anyway.

Third, a filter is expected to be set to an user-specified frequency and no
> default is reliably appropriate; so why bother second-guessing what the
> filter inputs would be?

With all due respect to your thoughts above, this point seems the core of
the discussion to me. My answer, and the reason why I brought it up at all,
is that oscilators *are* instantiated at arbitary frequencies. This is in
fact quite usefull if you are livecoding or for some other reason setting
something up realy quickly and want to get a sound imediately to be
fine-tuned later.

My point is that many things if left (temporarily) untouched default to
reasonable asumptions, for example everything defaulting to unity gain.
Unity gain can be speaker-damaging loud but it's a sensible default. There's
some merrit to your line of thought, we could have oscilators default to 0Hz
at a gain of 0 and make people turn it up themselves. That's safe in a way
but very inconvenient in practice.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.cs.princeton.edu/pipermail/chuck-users/attachments/20070522/406cae8d/attachment.htm 

More information about the chuck-users mailing list