I’ve played a little bit with the self-oscillation of the filters as a sound source using key tracking.
During this I have found out that while the Ladder filter sounds very stable and in tune, the Steiner filter always has some kind of wonkyness and a little detuning if I play a constant tone through all voices. I have performed a filter calibration, but it didn’t fix the issue.
This wonkyness seems to be particularly caused by the low-pass mode of the Steiner filter. If I blend the filter into high-pass or band-pass mode, the self-oscillation holds the tuning better.
Is this a normal behaviour of the Steiner filter? Has anybody else noticed this?
Yes, the Steiner is more unruly than the ladder. It’s the same on the MatrixBrute. Even so, I find that the Steiner can be usable as an oscillator over 3-4 octaves. In addition to, or instead of the key tracking you can use a little Key / Seq to Steiner Cutoff to stretch or compress the tuning for that last tweak for the range you want to play.
BTW, the PB12 manual mentions this specifically in “4.6.2. VCF 1 (Steiner)”
Thank you for the quick response. I will give this a try, but I guess it will be rather challenging to fix the tuning issues this way as the tuning is wonky between the several voice cards, not the note that is played.
Currently, I have the problem that one of the twelve voices plays the Steiner filter resonance always an octave lower than the others. I already performed several calibration runs for the filter, but unfortunately, that has not solved the problem yet.
Ah yes, there’s the voice variance. The MatrixBrute is where i used this, and it is monophonic of course. Too much in MatrixBrute mode here :-p
I just tested a little self oscillation on the PB12 (got the right focus now :)), and it was varying a little. Then I did a VCF calibration after a good warm up of the synth. After that, it’s nigh on perfect over the whole keyboard range and among the voices - better than I’ve heard before, and as playable as a filter of its kind gets. I haven’t tuned it up in a long time prior to this.
With one voice an octave out, I think you should contact support.
I have performed quite a few VCF calibrations yesterday - the synth has been on for at least an hour before. And the results varied a lot. After the first run, one filter (voice) of twelve was an octave lower. After another two runs, two filter (voices) were an octave lower Then I turned off the synthesizer and turned it back on a few minutes later. After the next VCF calibration run, all twelve filters were at least in the same octave.
I will check today how the filters react after a cold boot and maybe 30 minutes of warmup.
Okay … after testing again today after a cold start and roundabout 20 minutes of warming up, I get the following results:
All twelve self-oscillating filters (both Steiner and Ladder) play in the same octave - yeah
Ladder self-oscillation is the best in terms of tuning stability. It is close to perfect.
Steiner has more tuning variation. But it depends on the filter mode used. On my PolyBrute, the tuning in band-pass is the most stable one, followed by the high-pass mode. Both modes are close to the Ladder filter. However, the low-pass mode on the Steiner has more variation in tuning. It’s not completely off but at least a few cents between each of the twelve voices.
It seems like this is just the way the Steiner filter behaves. And I would guess that the VCF calibration cannot fix this instability because I doubt that the calibration values for the Steiner filter are specific for each filter mode.
I’m glad to hear that the Steiner filters pulled themselves together and agreed on an octave
Yes, that’s pretty much what I’m hearing. The ladder is bang on with the Steiner close behind, with a variance that makes it ever so slightly wonky. It’s not bad as far as Steiners go, because it really is the wild child. The Arturians have improved the tuning with new FW releases (there were several for the PB6) and I’m pretty happy with what we have now.
One tip for keeping the filter pitch true is to inject a very low level triangle from an oscillator into the oscillating filter. The filter will resonate sympathetically with the osc. Tune the filter as close to the osc pitch as you can, then mix in the triangle (other waveforms work too since only the fundamental comes through the oscillating filter, but it “feels more right” to use a triangle…). An osc mix level of 10-12% should do it. This way we get a filter that tracks the osc, and all we hear is the sine from the filter.
Then I turned off the synthesizer and turned it back on a few minutes later. After the next VCF calibration run, all twelve filters were at least in the same octave.
