rwelsh24

I am working on building filters inside of gen (like many people seem to be). Using the cascade with the filtergraph~ object, it seems like the gen biquads do not respond as expected. I've posted a patch here. Any thoughts from anyone?

I am working on building filters inside of gen (like many people seem to be).  Using the cascade with the filtergraph~ object, it seems like the gen biquads do not respond as expected.  I've posted a patch here.  Any thoughts from anyone?


cascade-with-gen-biquads

you had named all your histories with exactly the same name. see attached patch for fix.

in GenExpr you must name your histories, but in gen~ patching you don't actually have to at all - they would automatically have different names. in the original example patch where it comes from, they are only named to explicitly show the algorithm, and also so that they can be reset (zeroed) with a message externally.

what you are doing is expensive - it is a lot cheaper to use the cascade~ object, or even stacked biquad~s, in msp. but presumably you have your reasons, us lot usually do ;-)

Thanks so much. In my copy-paste frenzy, I missed their names. Is there a way to make this less expensive?

Also, the reason I'm doing this may not be the best reason, but I don't quite understand something. Is there a difference in sound quality (sample per sample calculation vs vectorized calculations) between building a filter in Gen and using an msp object? Is there a difference in sound quality if you build a complex filter and feedback function within the same Gen patch, or does it not matter if audio is routed through various Gen patches and traditional vector based objects like *~?

Thanks so much.  In my copy-paste frenzy, I missed their names.  Is there a way to make this less expensive?

Also, the reason I'm doing this may not be the best reason, but I don't quite understand something.  Is there a difference in sound quality (sample per sample calculation vs vectorized calculations) between building a filter in Gen and using an msp object?  Is there a difference in sound quality if you build a complex filter and feedback function within the same Gen patch, or does it not matter if audio is routed through various Gen patches and traditional vector based objects like *~?


Cascade~ with Gen Biquads?