you have email

my original patch has been updated to keep cutoff roughly the same when changing the upsampling factor

Thanx Seb!

@seb-harmonik.ar

Thanks. I did that in some other filters I made for VCV, from the same site, it did make the filter sound the same on all oversample rates

@jaffasplaffa The oscillators themselves do much of the work in getting those odd waveforms. The sawtooth's ramp portion is not linear and it has some overshoot at the peak, since all other waveforms are derived from the sawtooth they also show its faults. Here is a quick and dirty vintage style oscillator, the saw and triangle are not bad and in the realm of reality as too what vintage synths generated, the pulse is on the rough side, but it shows the logic of how the mini did it (if memory serves), as I said, quick and dirty, you will have to tune it to perfection.
waves.pd

@oid

Thanks man that's really cool. Those looks nice

Will try it out tommorow, when I have more time.

For now, MERRY Christmas!

Hey oid, your OSCs sounds so nice "imperfect" ... Imperfect in sense vintage style Thanx!
Seb and Jaffasplaffa, is something wrong with the HP mode? When i use the LP mode i can cut the freqencies of the sound to zero(no sound output). In HP mode it fails?! Is this intended?

@tungee said:

Hey oid, your OSCs sounds so nice "imperfect" ... Imperfect in sense vintage style Thanx!
Seb and Jaffasplaffa, is something wrong with the HP mode? When i use the LP mode i can cut the freqencies of the sound to zero(no sound output). In HP mode it fails?! Is this intended?

There is no hipass in the original algo. This is my own improvised hp filter.

Sorry not sure how to do that, feel free to play around. Don’t have time today for experimenting.

Hey no problem jaffasplaffa, it was just an observation by me. Happy Christmas

@jaffasplaffa and @tungee You should be able to do pole mixing with this filter to turn it into a multimode beast like Oberheim Matrix 12/Xpander 15 mode VCF. http://electronotes.netfirms.com/EN85VCF.pdf

@oid

Yes, I am sure it's possible, just did not try to make the bandpass yet. Hipass probably could be better also, but to be honest it my case, the filter is for bass, so not sure i actually need the other filter types a lot. But yeah, still gonna try making them, now it has become a matter of principles

I checked the oscs you made, the tri and saw sounds fine, but as you mention the pulse osc sounds a bit weird. But thanks for the starting point, I need to look more into them and see what I can come up with

@jaffasplaffa Yeah, it is overkill for a bass synth, just threw it out for the info. Craig Anderton wrote an article on putting the theory of the Hutchins article to practice, he tends to be more practical and less theory than Hutchins and his article may give you the exact mixing ratios for the various modes if you want to avoid the math, a search for "Craig Anderton Multiple Identity Filter" should find it. There is also some information in the Matrix 12 and Xpander service manuals. I do not recall how into depth any of these go, but they are certainly more terse than Hutchins.

The pulse really does need to be reworked, I was heading out the door and threw it together quickly in a way that would clearly show how it was done, the [send~] and [recieve~] is a little silly, a couple [*~] would do better and and make PWM much simpler, using the triangle to trigger the threshold might be better, or skipping the triangle and just using a copy of the saw which is 180 degrees out of phase to get the other half. It was done lots of ways in those days. These early waveforms were not completely accident or a failing of the technology, Bob could have gotten perfect waveforms if he wanted them, but perfect waveforms are less useful when it comes to subtractive synthesis, having some extra harmonics to play with is good. While the triangle has a fairly saw like sound to it, once you run it through that filter of yours you will easily be able to get a great sounding triangle when you need it, and unlike a more pure wave you can easily add in more harmonics, even with a very simple signal chain like you have in the mini, just turn up the cutoff frequency or use the filters envelop to do it for you if you want a more dynamic sound, or turn up the triangles volume on the mixer to drive the filter harder and get some distortion, no need for waveshapers or more waveforms. Bob really knew what he was doing, unfortunately others said "our waveforms are more pure!," the market followed. The mini is probably the most nuanced synth ever designed.

Nice explanation oid. Desired imperfections results in perfection

I stripped off the HP-calculation of the Karlsen filter abstraction and got ca.35% CPU Saving.!
Only used the clipping -1 +1 , no tanh calculation

27% with 8xOS

21% with 8xOS without HP-calc

My PD uses 7% in neutral mode(not running an abstraction)
Bildschirmfoto 2020-12-27 um 17.30.58.png

It would be more elegant to use a switch in the HP pd... but nevertheless for your moog you need only the LP-part of the Karlsen Filter?!
karlsen-ladder-24-LPonly 3.pd