Before I embark on working on a varaible resonant filter to use in conjunction with the Audio Library, I was wondering if anyone has looked at alternatives to the simple SVF already included? I'm interested trying to get much 'fatter' and aggressive sounding filters along the lines of classic analog filters like Moog and ARP, or even some of their software emulations. Cascading multiple sections of the AudioLib SVF (and adjusting Q to taste) still gives a fairly clinical result to my ears.
Custom resonant variable filters
- Thread starter rvh
- Start date
- Status
thedreiskre
Active member
There was a thread a little while back about porting moog ladder filter models to teensy, though I don't think there was ever a simple drag and drop moog filter object. https://forum.pjrc.com/threads/60488-Porting-moog-ladder-filters-to-audio-objects
On my synth projects I have experimented using the waveshaper effect to create more of an "analog" sound by using soft clipping curves and slightly overdriving the signal. There is a good list of waveshaping curves here.
https://forum.pjrc.com/threads/44061-Teensy-based-multistage-distortion-modeling-pedal/page2
If you're interested in looking at making your own filter model or implementing an existing algorithm, there is a lot of really great stuff on musicdsp.org https://www.musicdsp.org/en/latest/Filters/index.html
I hope this helps.
On my synth projects I have experimented using the waveshaper effect to create more of an "analog" sound by using soft clipping curves and slightly overdriving the signal. There is a good list of waveshaping curves here.
https://forum.pjrc.com/threads/44061-Teensy-based-multistage-distortion-modeling-pedal/page2
If you're interested in looking at making your own filter model or implementing an existing algorithm, there is a lot of really great stuff on musicdsp.org https://www.musicdsp.org/en/latest/Filters/index.html
I hope this helps.
Thanks very much for your reply. I did indeed find the first thread you refer to, but as you say there's no 'drag and drop' filter ready as an Audio Library object (yet?).
But there are about half a dozen different models described in that thread, and example code is also available e.g. at https://github.com/houtson/Teensy-4-Test-Area/tree/master/src. I've today managed to just about get one of them (2012 Stefano D'Angelo) translated into a preliminary custom audio object, but my intial test suggests that unless I have a bug there are stability limitations with it (e.g. instability at high freq, and still noticeable LPF when the cutoff is set very high). But the drive function and self resonance sound good - much nicer than the default SVF filter in the Audio library.
It's not clear to me if there was a consensus as to which of the models was considered 'best'. Maybe I'll just have to implement several to find one htat works for me. But I'm inspired by my initial results. I actually implemented a digital moog ladder on a DSP almost 20 years ago, and do recall that there was quite a bit of non-linear 'tweaking' required to get it to behave. I was hoping that after two decades it might be easier to find a good solution
But there are about half a dozen different models described in that thread, and example code is also available e.g. at https://github.com/houtson/Teensy-4-Test-Area/tree/master/src. I've today managed to just about get one of them (2012 Stefano D'Angelo) translated into a preliminary custom audio object, but my intial test suggests that unless I have a bug there are stability limitations with it (e.g. instability at high freq, and still noticeable LPF when the cutoff is set very high). But the drive function and self resonance sound good - much nicer than the default SVF filter in the Audio library.
It's not clear to me if there was a consensus as to which of the models was considered 'best'. Maybe I'll just have to implement several to find one htat works for me. But I'm inspired by my initial results. I actually implemented a digital moog ladder on a DSP almost 20 years ago, and do recall that there was quite a bit of non-linear 'tweaking' required to get it to behave. I was hoping that after two decades it might be easier to find a good solution
- Status