Page 1 of 1

Dirty filter

PostPosted: Wed Feb 24, 2016 4:02 pm
by Rocko
Hi all,

Any ideas on how to get more 'dirt', analogue and oomph out of a clean sounding filter (based on IIR second order)?

What the pros and cons of just oassing the signal via a saturator (like 'tan(x)') vs. Simply using an analog emulation filter (like the moog ladder filter)?

Re: Dirty filter

PostPosted: Mon Feb 29, 2016 7:18 am
by noisenerd
Since nobody else seems to be chiming in, here's $0.02 from a DSP noob:

Rocko wrote:What the pros and cons of just oassing the signal via a saturator (like 'tan(x)') vs. Simply using an analog emulation filter (like the moog ladder filter)?


Now, I'm no expert here either, but I've done some experimenting with circuits in the analog world, including a few different filters. I did a slight modification of this OTA-based filter by Ray Wilson where I added soft clipping and adjustable gain to the feedback (resonance) amplifier. While I was playing around, I experimented to make sure that the clipping in the feedback loop sounded sufficiently different from just doing the same thing at the input or output to be worth doing. Depending on the input and settings, it can be a pretty subtle difference, but it can definitely change the character of the filter.

Point being, I think there is a definite difference between filtering a distorted sound vs. a filter with internal nonlinearities (or whatever the proper phrase would be), and that the character of those nonlinearities matters also. Not that I'm an analog purist, I just think that what's happening "inside" is a big part of the sound.

Re: Dirty filter

PostPosted: Fri Mar 04, 2016 2:57 am
by martinvicanek
Rocko wrote:Hi all,

Any ideas on how to get more 'dirt', analogue and oomph out of a clean sounding filter (based on IIR second order)?

You might get some inspiration from this Analog Modeling Tutorial at KVR Audio.
Rocko wrote:
What the pros and cons of just oassing the signal via a saturator (like 'tan(x)') vs. Simply using an analog emulation filter (like the moog ladder filter)?

As noisenerd said, it sounds different. Moreover, nonlinearities need to be inside the filter to stabilize self oscillation. As a side remark, note that the Moog ladder filter is 4th order. That's what makes it sound so warm compared to other filters.