Additive Syntheis Synth using waveforms other than sine Pt. 2

[rlared]

With wavetable synthesis. say you have 8 points in your wavetable. Only one point sounds at any given time. I'm looking to sound all 8 points at a given time. With additive synthesis you can obviously do that, but then you're restricted to all sine waves.

How is this different than what you can do in a synth with a lot of oscillators like Avenger or Synthmaster or Arturia Moog Modular? Also if literally all you're doing is stacking waveforms at the same frequency it's just going to give you a different static waveform, right? You'd have to do something like detune them, or modulate the phase or the frequency content of the individual waves. Throw in a filter and you've arrived at subtractive synthesis (with wavetables if changing the partial's frequency/phase content). Or do you just want to be able to play the oscillators at weird inharmonic intervals?

Is it really that difficult to understand what it is I'd like to do?

Okay, let's try an example. We'll make it simple or I'll be here all day.

Let's take a very simple additive oscillator. Just 4 partials.

Partial 1 - Sine Wave
Partial 2 - Sine Wave
Partial 3 - Complex Wave (doesn't matter what)
Partial 4 - Sine Wave

That's what I want to do. Is this really so hard to comprehend? And with the complex wave NOT being at the fundamental, it won't overpower the other waves, especially if you can control the volume.

In fact, in looking at this, I can't believe this hasn't been done. This seems so basic to me. It almost boggles my mind.

Doesn't synthmaster let you choose more complex waveforms for the partials? And Zynaddsubfx definitely lets you do this.

[felis]

Sounds like you just want multiple oscillators - 16, 32, 512, or whatever.

Partials are sine waves. Your partial #3 Complex wave is not a partial.

Basically, it's just nit-picky semantics, unless someone wants to use the correct terminology.

[synzh]

exactly - you need an oscillator for each wave - be that sine or complex wave (each "point" of a wave table is a wave in itself)

[Ah_Dziz]

This is the kind of thing reaktor and such is for. There are some neat tools for this if you want to do math. The additive section in Alchemy will allow for the waves used for partials to be swapped for non- sine shapes. This comes in handy under some resynthesis circumstances. The high frequency “mess” causes by using triangles or softened saws adds a little something something to percussive sounds or drumloops. I know mathematically it’s all just sine waves, but this is a shortcut that can add some extra high partials when you have a limited number of additive oscillators. It may be fun to be able to swap waves for only a certain part of the spectrum.

[wagtunes]

This is the kind of thing reaktor and such is for. There are some neat tools for this if you want to do math. The additive section in Alchemy will allow for the waves used for partials to be swapped for non- sine shapes. This comes in handy under some resynthesis circumstances. The high frequency “mess” causes by using triangles or softened saws adds a little something something to percussive sounds or drumloops. I know mathematically it’s all just sine waves, but this is a shortcut that can add some extra high partials when you have a limited number of additive oscillators. It may be fun to be able to swap waves for only a certain part of the spectrum.

So Alchemy CAN do this. Problem is, I'm not on MAC and we all know what happened to PC Alchemy. But yes, that's essentially what I'm looking for.

[Greenstorm33]

With wavetable synthesis. say you have 8 points in your wavetable. Only one point sounds at any given time. I'm looking to sound all 8 points at a given time. With additive synthesis you can obviously do that, but then you're restricted to all sine waves.

How is this different than what you can do in a synth with a lot of oscillators like Avenger or Synthmaster or Arturia Moog Modular? Also if literally all you're doing is stacking waveforms at the same frequency it's just going to give you a different static waveform, right? You'd have to do something like detune them, or modulate the phase or the frequency content of the individual waves. Throw in a filter and you've arrived at subtractive synthesis (with wavetables if changing the partial's frequency/phase content). Or do you just want to be able to play the oscillators at weird inharmonic intervals?

Is it really that difficult to understand what it is I'd like to do?

Okay, let's try an example. We'll make it simple or I'll be here all day.

Let's take a very simple additive oscillator. Just 4 partials.

Partial 1 - Sine Wave
Partial 2 - Sine Wave
Partial 3 - Complex Wave (doesn't matter what)
Partial 4 - Sine Wave

That's what I want to do. Is this really so hard to comprehend? And with the complex wave NOT being at the fundamental, it won't overpower the other waves, especially if you can control the volume.

In fact, in looking at this, I can't believe this hasn't been done. This seems so basic to me. It almost boggles my mind.

So basically it's just a matter of being able to properly tune an oscillator to have its fundamental at a frequency within the harmonic spectrum of a different oscillator? I'm pretty sure the (sine) partials of the complex waveform are going to be inharmonic to the base waveform, which wouldn't sound very good to most people and is probably why it hasn't really been done. To get a feel for how it sounds to have complex partials you can use the harmonic editor in Sytrus after creating a harmonic-rich wave with the waveshapers. Every partial in the harmonic editor will actually be the base waveform played at the pitch of that point in the harmonic series of the base frequency.

[Ah_Dziz]

This is the kind of thing reaktor and such is for. There are some neat tools for this if you want to do math. The additive section in Alchemy will allow for the waves used for partials to be swapped for non- sine shapes. This comes in handy under some resynthesis circumstances. The high frequency “mess” causes by using triangles or softened saws adds a little something something to percussive sounds or drumloops. I know mathematically it’s all just sine waves, but this is a shortcut that can add some extra high partials when you have a limited number of additive oscillators. It may be fun to be able to swap waves for only a certain part of the spectrum.

So Alchemy CAN do this. Problem is, I'm not on MAC and we all know what happened to PC Alchemy. But yes, that's essentially what I'm looking for.

Yes it can. I still use the pc version quite a bit. It’s a fun trick that works sometimes and sometimes just makes things sound terrible. I don’t know if the updated version had any more features along this line added since I’m a PC user as well.

JJ

[felis]

....I'm pretty sure the (sine) partials of the complex waveform are going to be inharmonic to the base waveform,.....

Partials are part of the harmonic overtone series. Non harmonic overtones are not.

All partials are overtones, but not all overtones are harmonic partials.

[Greenstorm33]

....I'm pretty sure the (sine) partials of the complex waveform are going to be inharmonic to the base waveform,.....

Partials are part of the harmonic overtone series. Non harmonic overtones are not.

All partials are overtones, but not all overtones are harmonic partials.

Say you have a saw wave at C3, and you play another sawtooth with its base frequency at frequency 13 of the C3 harmonic series. I believe many of the partials of that wave are not going to be at positions within the harmonic series of the base saw wave/beginning at C3, despite the fact that they are by definition within the harmonic overtone series of a note at whatever pitch harmonic 13 of C3 is. So, the process wagtunes is suggesting will yield inharmonic overtones with respect to the base frequency. I could be wrong.

[felis]

Say you have a saw wave at C3, and you play another sawtooth with its base frequency at frequency 13 of the C3 harmonic series. I believe many of the partials of that wave are not going to be at positions within the harmonic series of the base saw wave/beginning at C3, despite the fact that they are by definition within the harmonic overtone series of a note at whatever pitch harmonic 13 of C3 is. So, the process wagtunes is suggesting will yield inharmonic overtones with respect to the base frequency. I could be wrong.

Yes - they are independent. Some overtones might match up, I haven't mapped them out.

But basically, you are just using two sawtooth oscillators at different frequencies.

You could do the same with any number of oscillators if they were available. It could get awfully messy though.

[Chandlerhimself]

Synthmaster does this and I’m sure some other synths do also.

[yellowmix]

Wavetable synths can generally do this, at least the ones that let you work with partials. Vaporizer, Hive, Serum, Avenger, all do. Serum you could write formulas, Hive an UHM script. Vaporizer has some GUI tools that would help. Avenger would be a little tedious. Zebra 2, for that matter, in Spectromorph or Spectroblend.

Here's a freebie:

Put the formula below in Serum's formula parser and run it, it will create a "Complex Wave" (a sawtooth) at the 3rd partial. Then hit the Wave to FFT button. In the partial bin on the top, add your 1st, 2nd, and 4th Sine Waves. Congratulations! You have a 3rd octave sawtooth plus whatever you put in those bins.

Code: Select all

(x*3<-1?(x*3+2):0) + (x*3<-1 || x*3>1?0:x*3) + (x*3>1?x*3-2:0)

If you're not good with maths, then simply create or add your complex waveform to Oscillator B, and set Octave to whatever partial you want it to start at. Unfortunately it only goes up 4 octaves, but then again, much higher and it's at the high end of human hearing. Then you can work with the partial bin in Oscillator A as your heart desires.

[shonky]

....I'm pretty sure the (sine) partials of the complex waveform are going to be inharmonic to the base waveform,.....

Partials are part of the harmonic overtone series. Non harmonic overtones are not.

All partials are overtones, but not all overtones are harmonic partials.

Say you have a saw wave at C3, and you play another sawtooth with its base frequency at frequency 13 of the C3 harmonic series. I believe many of the partials of that wave are not going to be at positions within the harmonic series of the base saw wave/beginning at C3, despite the fact that they are by definition within the harmonic overtone series of a note at whatever pitch harmonic 13 of C3 is. So, the process wagtunes is suggesting will yield inharmonic overtones with respect to the base frequency. I could be wrong.

I was trying to write something similar last night, but thankfully you've explained it better than I could. By way of analogy, this is a similar reason as to why complex harmonies and distortion can be problematic. A root-fifth powerchord potentially introduces 9ths, major 3rds and flat 7ths as more harmonics are introduced, which generally work within many keys, but once you introduce more complex harmonies, you start introducing a lot of competing frequencies that cause beating, phase cancellation and obscure the notes in the chord (although this has been use by Glenn Branca, Sonic Youth, My Bloody Valentine, et al. to good artistic effect throughout the years).

[Beachboy]

Let's take a very simple additive oscillator. Just 4 partials.

Partial 1 - Sine Wave
Partial 2 - Sine Wave
Partial 3 - Complex Wave (doesn't matter what)
Partial 4 - Sine Wave

Just an idea, what if you create partials 1,2 and 4 with an additive osc and the partial 3 with a second "normal" osc which runs through a frequency shifter. The frequency offset has to be calculated. These 2 oscs would combine harmonic and inharmonic tones.

[faxinadu]

why do u need a frequency shifter or anything funky. as stated a dozen times here, "partial 3" is just a normal synth oscillator detuned. there is zero point in having it in the additive sequence.