View Single Post
Old 12-07-2018, 11:55 AM   #113
geraintluff
Human being with feelings
 
geraintluff's Avatar
 
Join Date: Nov 2009
Location: mostly inside my own head
Posts: 346
Default

Quote:
Originally Posted by nitsuj View Post
The SVF filters are extremely good - much better than the RBJ cookbook ones that a lot of EQs used to use. Clean, stable, low noise and transparent.
Could you expand a bit more on this, please?

My understanding was that the ZDF/digital-SVF filters were equivalent to a traditional biquad calculation for steady-state, and that the (rather cool) advantages only came either for fast modulation, or when modelling distortion in the analogue circuit.

It's possible I've made a mistake, but I think the maths comes out that for pure filtering (no modulation, no distortion) the ZDF is equivalent to a biquad structure.

(Discussions of ZDF stuff often bring up frequency-warping for the critical frequency, but the RBJ cookbook includes compensation for this aspect of the bilinear transform, which is standard. JSFX also uses 64-bit floating-point values, so I'd be very surprised if there were any difference in the noise levels.)

For illustration, here's an image comparing ReaEQ with ReEQ with the same frequency/bandwidth (oversampling turned off). Since both filters are minimum-phase, an identical amplitude response means an identical phase response as well:



So unless you're modulating the bands quite fast (or modelling non-linear components), they should be the same, right? Not that they aren't neat, just that I don't think we're getting the benefit from them in this situation.

Geraint
Attached Images
File Type: png freq warping.png (9.9 KB, 2636 views)
__________________
JSFX set | Bandcamp/SoundCloud/Spotify

Last edited by geraintluff; 12-07-2018 at 12:04 PM.
geraintluff is offline   Reply With Quote