axiom
axiom copied to clipboard
Add separate "pre-generate" callback to allow preparations before a batch of generate calls
This should allow us to optimize some functions like delay
and the various biquad filters where changing a parameter is expensive, by only performing recalculations once before a batch.