Can't get bass frequencies from self-resonating LP filter

[jaytee]

Trying to synthesize a kick drum. Everything appears to be working as expected (osc all off, filter res all the way up, envelopes set up to create a falling pitch from the filter) except that that as I lower the cutoff (and thus the pitch), the amplitude drops out of the sound.

I have been able to create nice sounding kicks on this instrument before, which makes me suspect that I just have a setting wrong somewhere....but this is not a complex patch, so I'm also pretty certain I have it set up properly.

All other sounds appear to be working as normal, it's just that the range of the self-oscillating filter doesn't want to go into the bass frequencies.

Any ideas?

Edit: Turning up the "Brute Factor" adds some of the bass frequencies back—ie. I can get a passable kick sound—but I know for a fact this was not necessary in the past.

[stevism]

take a picture of your patch and post it, i'm curious / that will ensure it's copied exactly (or close enough)

my first guess is play with key follow

[jaytee]

More experimenting yielded the following results:

- All oscillator silders at 0
- All A sliders at 0, and all D, S and R sliders at 100
- Mod wheel is at 0, glide is off, LFO routing knobs are all at 12 o'clock (zero), arpeggiator is off
- Brute Factor is off

- For the filter section, Env Amount is at 0, KBD Tracking is at 0 (off), Resonance is at max, Mode is on LP, Envelope speed is fast

This, as expected, sounds like a sine wave whose pitch is constant across the keyboard but varies as the cutoff knob is twisted.

With the cutoff above 12 o'clock, everything sounds as normal. Ringing at slightly above C6, about 1063Hz. Move the cutoff knob one mark to the left (anti-clockwise), and we're about just above C#5, around 572Hz, with a 1dB drop in amplitude. HERE'S WHERE THINGS GO WEIRD. Moving the cutoff knob just one more mark to the left (so two total marks anti-clockwise from 12o'clock) and we lose all amplitude from the sound entirely. This seems to happen right around C#4, or between 375Hz and 380Hz.

What's going on here? I am totally stumped. I know that I used to be able to get some great deep kicks from my Minibrute, but that is impossible if the filter won't resonate below C4. As I said earlier, cranking the Brute Factor gives me some bass frequencies back, but that wasn't necessary in the past—in fact the Brute Factor was a great sound-shaping/tweaking tool once I already had a decent kick dialed in.

FWIW, haven't been mistreating or gigging with this keyboard. It's been sitting on a desk its entire life. If something happened to it physically without my knowledge, it didn't affect any other sounds. All I know is that this is the first time in a couple months that I've tried to dial in a Kick patch and the first time I've ever been completely unable to get the frequencies I need.

[jaytee]

http://imgur.com/L8jiXWo
This one should be pretty close to a kick drum patch, as described in my first post. All I get from this is the mid-range "buh" attack sound, but there is absolutely no bass to the sound. If I crank Brute Factor all the way, I get a little more bass to the sound, but not nearly what I should be getting.

http://imgur.com/FQD1kXV
This one is my "diagnostic patch" as described in my more recent post. Just to try and quantify exactly what pitch my amplitude drops out at. Results described in depth above.

[Quinthius]

I am by no means an experienced electronic musician, so feel free to call me a noob for suggesting this

But I have had some interesting results using the Band Pass filter instead of the Low Pass filter.  The BP seems to allow for a much stronger resonant spike, and used in conjunction with the filter envelope AND the Brute Factor I have come up with some interesting Kick-like sounds.  Maybe not traditional synth kick sounds, but definitely to my ears can produce useable results.  You just have to find the sweet spots for the Cutoff, Resonance, Filter Envelope, and Brute Factor.  Just be careful with that resonance.. on the BP it can go crazy real fast!

If I remember correctly the LP filter on the Minibrute is a 12 db right?  And I think a 24 db is a little better for producing those traditional kick patches?  So that might be part of the difficulty?  Either way, some very cool percussion sounds are possible with the Minibrute, I think it just takes a bit of experimenting to find those sweet spots, and also definitely worth exploring different features on the Minibrute and alternate ways of creating those types of sounds.

[fanwander]

Hello

The amount of resonance in a Sallen-Key-Filter is depending on the centerfrequency of the filter. There are various approaches to get around this circumstance, but it is known that each of these workarounds is not perfect. I don't know which one is used in the Minibrute. But for sure it is normal what you are experiencing.

If you are interested in the technical background, there is some interesting reading on the Doepfer page about the technique behind the A-101-1 module (which is very similar to the minibrute filter): http://www.doepfer.de/a1011_tec.htm

Florian

[jaytee]

Florian---

If this is normal behavior, I have to wonder where the kick drum patches I previously dialed in (and still have samples of) came from, and how I can get back the abnormal behavior that allowed for their creation.

[beefinator]

Maybe try mixing in a bit of noise or an oscillator.

Either could help the filter oscillate, giving it a bit to work with instead of having to fully self-oscillate.
You should be able to have it toned down so that the self-resonance overpowers the noise or tone except when it's in the bass range.

But it's also just the nature of analog that you might not be able to get the exact same thing (or might find it very difficult to do so) when you try it a second time.

Try stuff that you wouldn't normally think of.  When you're pushing the limits, with self-oscillation and brute factor and all, lots of unexpected stuff can happen.

[jaytee]

Beefinator---all of that *helps* to get a more acceptable kick sound, but at the same time, it's very static. Part of the reason for using the self-oscillating filter is to get a falling pitch envelope, but with the cutoff wherewewa it needs to be to be acceptably "kick"ish, you lose all of that pitch movement. Instead, all I'm getting is a mid range attack that fades away before the pitch can fall appreciably.

Adding some noise and suboscillator deepens the sound considerably, but besides that tiny bit of midrange attack, the filter has no effect on the sound (ie, twisting the resonance knob has almost no effect).

Has anyone tried the patches I posted? I distinctly recall that my filter would self-oscillate well into bass range instead of cutting out right around C4, but if others are putting in my patches and getting identical/similar results, I can let it go.

[jaytee]

Bump.

Can someone punch in the patches I posted and see if their Minibrute duplicates the behavior I am describing?

[coldwar1977]

I have just tried out your patches and indeed: I experience the same thing as you.

[beefinator]

Same here, but I don't remember whether it could or couldn't resonate that low in the past.


Probably a long shot, but I wonder if it's seasonal, that the change in temperature in the past couple months affected the filter.     (temperature I would guess would start affecting the filter first at its extremes, like self-resonating at low notes)

[matias_thinKing]

I have also tested the self oscillation, and I found it doesn't self oscillate anywhere near bass frequencies.  mine cuts out at 714hz, exactly F4.  I'd say yours is doing better than mine based on your numbers!

now I thought about this for a moment, and maybe the experts can chime in, but self oscillation at lower frequencies seems almost like an impossibility.  since bass tones have wider bands, I would think the Q of the filter would need to go into the stratosphere to grab the note's full band.  I could be totally mistaken on how this works, so take it with a grain of salt.

as I was testing this, I also found that the keyboard tracking knob is very, very sensitive, but it opened up my mind to a few more patching possibilities

edit:

a few more notes.  turning up the brute factor to right under 3 o'clock helps the filter self-oscillate down to D0 (~36hz).  this is no longer a true sine wave as there are a few harmonics, but the number of harmonics are minimal and it's producing a note without any wave forms being used.

so maybe you had the brute factor on before?  it goes in-line with what I thought about a filter self-oscillating, since I believe due to the feedback loop produced by the brute factor, the Q is emphasized.  anyway, at the very least, this should be a good place to start for your patch

all waveforms at 0
filter ADSR at 0
amp sustain all the way on and ADR at 0
filter cutoff at 10 o'clock
resonance at max
filter env at 12 o'clock
kbd tracking at 12 o'clock
brute factor at 2:50ish o'clock - if it starts to sound like an LFO, you've gone too far.

now start to work your way down the keyboard; you should be able to hit the lowest note and produce a tone without artifacts.  from here, I think you should be able to work out your filter envelope to suit the kick you're trying to patch.  I hope this helps, and happy patching

double edit:

I guess you already figured out that the brute factor allowed the self-oscillation to go into the bass region, so apologies for overlooking that!  I also found that the filter will not self-oscillate while in bandpass mode.  can anyone else confirm if that's the case in your mb as well?

[beefinator]

The bandpass definitely can self-oscillate, but not at the same parameters as the others.  Probably because the LP and HP are 12dB/octave, and the BP only 6dB/octave.

I don't know for certain, but I think the "slope" of the filters (if that's the right term) also affects the amplitude of the resonance peak.

[matias_thinKing]

what settings are you using?