Can a software oscillator ever be called analogue?


I hope someone here can help settle a debate (or at least chime in).

I’m having a lively discussion with the designer of a little microcontroller based synth. He claims that it uses analogue pulse waves which are then filtered with a digital filter. After questioning this, it turns out that what’s actually happening is that the microcontroller is generating naïve pulse waves at a relatively high (but not that high) sample rate of 125kHz. So the edges are on exact sample boundaries, making it a bit like an oscillator based on dividing a master clock (but with much lower pitch accuracy). This then going straight into an 8-bit digital filter running at the same rate, all within the microcontroller.

I don’t think it’s right to call it analogue at all, but maybe I’m entirely wrong. Apologies if this is the wrong section, seemed like the right place for an academic sort of question.


I think the answer to your question is “no”, but that’s not to make a value judgement about whether analogue is better, which I think is the subtext of the discussion you’re having. The sample rate may be high enough that the ears can’t tell the difference, but that doesn’t make it analogue. But if the ears can’t tell, who cares?


I assure you, you’ve completely misread the subtext. I love digital synths just as much as analogue, I just think accuracy is important. This is a purely technical discussion as to the classification of a particular technique.

It’s a cool sounding synth in its own way and I’m considering buying a DIY kit. I dislike false claims though and wanted to check if anyone else felt there was a basis for calling it analogue.


I think I know who this is and I think you should not give this person your money. :slight_smile:


Haha, it’s not our friend “Mr 8-bit Solina Emulation”! Although the situation reminds me of him a bit. I’d like to get a couple more opinions on it an then I’ll reveal what the product is.

EDIT: I want to be fair, I don’t think the person I’m discussing this with is disingenuously pretending their oscillator is analogue, I just think they have a faulty definition. But I wanted to get second, third, fourth opinions to make sure I’m not being unreasonable. I hope pichenettes sees this and weighs in, even if to tell me I’m wrong.

The developer in question says their oscillator is no different from a master clock based divider organ or a DCO like in the Juno 106.


Totally agree. For example it bothers me a lot when people call switching PSUs digital, because there’s nothing inherently digital in them (there’s no numbers being stored or processed nor is any sort of software involved) but people use the term because switchers use PWM to modulate the voltage (they more or less transform the current into a square wave switching it on and off) and that looks like zeros and ones.

With synthesis the digital/analogue is more complex though, since you get pure digital synths (software outputting the signal through a DAC), you get various sorts of digitally controlled analogue stuff, then pure analogue. At its very core any digital circuit is analogue as well and you can make totally analogue circuits by using digital chips in an analogue way (i.e. they are not used to process data, but just as square wave generators).
This said, if software of some sort is involved I’d call things digital.


Not analogue.



In case anyone was wondering, it’s the Ploytec PL2. For some reason the designer insists that his oscillator is analogue. :confused:


Not analog.

The PSU example is one of semantics again, but some would say that digital is used to describe systems that break down information into binary numbers which can be represented with “state on” and “state off”, or zeros and ones, and by that definition the switching power supply is certainly “digital”, since it is simply switching between the two states. I feel the same way about the micro controller-based pulses, no matter how fast the clock is running; if it is a program creating pulses switching between only two states then it is digital.

Back when I used to build Serge modular synths we were fond of saying “but what about all the numbers between zero and one?” THAT is analog, in my humble.


He indeed calls it analog on his site.
But as i understand. There is some digital output running an oscillator. Which goes from the analog domain into an 8 bit ADC into the digital filter. Is this what’s happening?
That isn’t really analog but i get where the confusion is coming from. But i have no idea why you would build a synth like that.


Isn’t it more a matter of whether the signal is continuous or discrete time sampled rather than whether software is used in controlling the signal?


I don’t think so, at least from my understanding the oscillator code and the digital filter are all within the same microcontroller and it never leaves the software realm before that.


Yes it of course is, but it’s usually easier to look at things being done with software or not.
Analogue basically means that there is an analogy between signal A and signal B. For example the orientation of molecules on a magnetic tape or the “jaggedness” of the vinyl grooves is analogue to the waveform in shape or in the way the molecules are oriented. In the digital domain there is no analogy between two signals. This of course all gets much more complex when you start to look at synthesis, where you generate something from nothing, so there’s no analogy that could be identified to start with. In this case there usually is some sort of software involved, or there isn’t and most of the time that will be a good indication about if you can call something analogue or digital.


An analog signal is any continuous signal for which the time varying feature (variable) of the signal is a representation of some other time varying quantity, i.e., analogous to another time varying signal. […] It differs from a digital signal, in which the continuous quantity is a representation of a sequence of discrete values which can only take on one of a finite number of values. —

That’s how it’s defined.


Isn’t that what I’ve said? :joy:


Correct. But the original statement “if software of some sort is involved [then it is] digital” would suggest that any instrument using MIDI is digital (i.e. some software is necessary to interpret the data received via the serial communication connection and translate it to a sound). It might be more correct to say that if software of some sort is involved in generating the waveform and the waveform is discrete time sampled, then it is digital.


Yes, you’re totally right!


I have one of those PL2 units, the only thing that’s analog is the 2nd filter, which is more like a tone knob on a guitar, not resonant, just there to smooth things out. The oscillators are extremely digital/lofi, and a bit harsh. The sound is different from all my other synths, and I still havent used it on a song yet.


To be honest, I’m most interested in it for the alternative voice synthesis firmware. Have you tried that at all?


I have so many synths that this little thing is velcroed on the back of one of my rack systems. I just have the normal 2.x firmware and havent tried it’s voice mode yet. The reason I got it was actually because I had 2 square inches of free unoccupied space in my rack and felt compelled to put something there to fill in the hole. :grin: