Additive oscillators

  • I'm loving the crisp, tight sound of the Virus - it's making me realise what I've been missing with a lot of soft synths. However, there seems to be a growing fashion for additive synthesis - and the sounds available using an additive synthesis engine are interesting, diverse and powerful. Software synths like Razor and Harmor have shown that the seemingly bewildering concept can be broken down into an easy-to-use, familiar format.


    It would be really interesting to have the ability to swap out a couple of oscillators for additive oscillators (with a selection of pre-created additive waveforms) and plug in a pair of additive processors while still within the additive engine, before it drops back to the subtractive engine to use the classic filter and effects sections. I think this 'best of both worlds' approach would give the Virus a unique edge.


    I do, however, accept that the amount of development time and processing power required may well be prohibitive.

  • I'm not sure I'd call Harmor easy to use except for just fiddling. Bear in mind you do have FM and RM, also, which are additive processes, although the Virus doesn't feature some of the frequency domain manipulation tools that Harmour has although I wonder if these are just gimmicks or if they do actually offer something unique...

  • I didn't think of FM and RM as additive, more as waveform modulation techniques. (I see an additive engine as being one where the final output waveform is constructed from a number of sine waves of different pitches, phases and amplitudes. FM and RM take existing waveforms, sometimes already complex, and multiply them.)


    I think the uniqueness of the operations within the additive engine come from the way that the resultant sound is utterly, unashamedly digital. The fact that the algorithm is playing with sine waves rather than complex waveforms makes for a particular kind of precision. Also, it's possible to think up additive tools which work in ways no analogue-style filter or effect could. Detuning selected harmonics is one. The 'filters' and effects you find on Razor and Harmor can make some remarkable sounds.


    I'd love to hear how all that works as a stage before the rich analogue emulation of the Virus. But like I said, it's a lot of work - and would probably take a hell of a lot of DSP power.

  • Subtractive synthesis is where you start with a base oscillator and its harmonics, and every stage thereafter only alters the amplitude of the frequency components you have.


    Additive synthesis takes base oscillators and creates new harmonics with its processing. FM and RM do this.


    Also, frequency domain manipulation is not so intensive, thats how digital vocoders work. You can quite comfortably take the FFT (convert to frequency domain) of an incoming signal, filter it in the frequency domain by multiplication and then convert it back in overlapping blocks (COLA) on even the most lightweight of DSP systems. 'Frequency remapping' (i guess you'd call it) probably isnt a huge amount more intensive, depending on how complex you go and if you can think of clever tricks to make it more efficient.

  • > Additive synthesis takes base oscillators and creates new harmonics with its processing. FM and RM do this.
    OK, if you like. We're getting hung up on semantics though. FM and RM simply aren't what I'm talking about. They work in a completely different way and they produce very different sounds.



    Based on all I've read regarding different synthesis techniques to date, I think of 'FM' and 'additive synthesis' as different classes of synthesis. What I call 'additive synthesis' uses the addition of many partials - all sine waves and all individually malleable - to create the final waveform. FM, RM and I guess PM are all carried out using the multiplication of audio frequency waveforms, often complex ones.


    What would you specifically call what I refer to as 'additive synthesis', just so we're using the same language? If, for example, you didn't want to lump it in with FM and RM?


    > Also, frequency domain manipulation is not so intensive, thats how digital vocoders work. You can quite comfortably take the FFT ...



    But rough FFT approximations aren't how additive engines work. They model the individual sine waves in order to give that characteristically crisp, digital sound. I do know that even a simple additive patch on Harmor or Razor will eat CPU like it's going out of fashion. Could be bad programming, but I suspect the calculations are intense.


    Interestingly, given your description of subtractive, both Harmor and Razor would fit the bill. Looking at the interface, they use complex oscillators at the start of the signal chain and everything afterwards alters the frequency components. The difference is the engine and the new sonic options allowed to you thanks to the ability to play with individual sine waves.


    The Virus isn't an FM synth. It gives a limited FM option so you can add an FM flavour to sounds. I'm suggesting a similar thing when it comes to additive synthesis. The ability to add an additive engine component to the signal path to add some of its flavour to the synthesis.

  • I think I'm going to retire discussing anything to do with DSP on these forums as it is a major source of stress...


    Quote

    But rough FFT approximations aren't how additive engines work.


    Not exclusively, no, but FFT is not an approximation, it is a very specific and unique calculation based on a section of a signal - the precision is limited to the bit depth of the system calculating it. But you could also keep track of them as 'partials' which may be more efficient for different systems, but since Harmour can use audio files as an input it mustn't work entirely this way without some spectrum -> set of partials going on. In general, partials just refers to the harmonics on a waveform (although some synthesizers like the Roland D50 have massaged this meaning).


    Quote

    What would you specifically call what I refer to as 'additive synthesis', just so we're using the same language? If, for example, you didn't want to lump it in with FM and RM?


    I guess, if you are talking about some of the things that Harmor does, then the term 'spectral remapping' i coined earlier would probably fit...


    Quote

    Interestingly, given your description of subtractive, both Harmor and Razor would fit the bill. Looking at the interface, they use complex oscillators at the start of the signal chain and everything afterwards alters the frequency components. The difference is the engine and the new sonic options allowed to you thanks to the ability to play with sine individual waves.



    No, subtractive synthesis exclusively manipulates the amplitude of frequency components and not their frequency. An example is a filter or amplitude envelope, in fact every subtractive operation can be represented as a filter.


    Additive can create components, or 'partials' at different frequencies based on operations on the already existing ones - such as FM which varies the frequency of the partials based on the value of an oscillator, or RM(AM) which creates partials above and below the partials based on the sum and difference of the frequency of the partials. The prism function on Harmor is actually pretty similar to AM - but I haven't tinkered with it enough to know exactly what it is doing.

  • This is all fascinating stuff, but come on, *you know what I mean*. Both Harmor and Razor show you a graphic representation of the partials as you manipulate the processors and oscillators. They can do this because each partial is modelled individually. Unlike with FM or RM, where such modelling isn't necessary and everything can be carried out using simple analogue components if desired.


    (I believe audio imports are converted into their corresponding internal partial representations before hitting the synthesis engine in Harmor.)


    So basically, I've found this particular partial modelling technique to produce an interesting and powerful sonic palette. If you want to create the name 'spectral remapping' for it then fine... Although I don't think the phrase quite does it justice and... well, 'additive synthesis' perfectly describes what I mean already, according to Wikipedia for one:


    : Additive synthesis is a technique of sound synthesis that creates musical timbre by explicitly adding sinusoidal overtones together.
    : The timbre of an instrument is composed of multiple harmonic or inharmonic partials (individual sine waves), of different frequencies and amplitudes, that change over time. Additive synthesis
    : allows the emulation of a given tone or sound by giving control over the frequency and amplitude of each individual harmonic or partial. In general, each harmonic generator has its own
    : customizable amplitude envelope and instantaneous frequency function, creating a realistic, dynamic sound that changes over time.



    Specifically this:


    :Additive synthesis allows the emulation of a given tone or sound by giving control over the frequency and amplitude of each individual harmonic or partial. In general, each harmonic generator [partial] has its own customizable amplitude envelope and instantaneous frequency function



    Have I done enough yet to narrow down what I mean to you?


    And rather than splitting hairs with terminology or using inappropriately broad categorisations to imply that the Virus does this in some form already(!), what do you think of the idea of the feature? Do you think that the Virus can already do everything that it could if it had an additive engine component?


    By the way, where did you get your definition of additive from? As I said, everything I've ever read about additive synthesis agrees with the Wikipedia entry, and I've been interested in synthesis for a fair while now. You obviously know your stuff, so what's your source?

  • Thanks! :D


    That article sounds oddly familiar, especially the crack about the Hammond Tonewheel Additive Synthesizer. It could well be one of the very articles I read about different types of synthesis many years back when I had a SoS sub.


    Again, it seems to reinforce my understanding of additive synthesis and in particular the architecture of Razor and Harmor. They appear to have cracked the idea of using a single osc for each harmonic, each independently controllable.


    The more I think about it though, the more I think that an additive engine would utterly swamp the resources available to the Virus.


  • Yes it is, of course, possible to have a bank of oscillators, each modelling each partial - I have no idea if that is how Harmour works (you could just work in frequency domain) - but in the case of the Virus, it has 80 voices, so you can guess how quickly you would run out of voices...


    Be comforted by the fact that using wavetable synthesis and the variants on the Virus you can get a lot of the harmonic modulation effects you might want from there and using the formant shift you can get a lot of great sounds which are easy to keep in check.


    Personally I'm not that impressed with Harmor, it was going cheap and I wasn't persuaded, but I parted with over £1000 for the Virus, so what does that tell you?


  • Yes it is, of course, possible to have a bank of oscillators, each modelling each partial - I have no idea if that is how Harmour works (you could just work in frequency domain)


    Unless the press release and instruction manual are lying then yes, that is how it works. They could be lying of course, I haven't decompiled the code or anything. But they call it an additive synth, describe it as using an oscillator per partial and, indeed, additive synthesis is defined as the use of an oscillator per partial (unless you can show me a source more reputable than SoS and Wikipedia which supports your definition. Perfectly possible to do of course, I just haven't seen it yet.)

    - but in the case of the Virus, it has 80 voices, so you can guess how quickly you would run out of voices...


    Unless, maybe, it's possible to optimise a simple sine wave oscillator in a way that you can't with, say, an oscillator capable of pulse width, wave shape, triangle, saw, square etc.

    Be comforted by the fact that using wavetable synthesis and the variants on the Virus you can get a lot of the harmonic modulation effects you might want from there


    Having a wavetable oscillator isn't like having an additive engine. It's lovely to have and there may be ways to get some similar sounds (depending on the wavetables you're given) but it's not the same.

    and using the formant shift you can get a lot of great sounds which are easy to keep in check.


    All of the Virus features are capable of giving you a lot of great sounds. The virus is a great synth, that is why I shelled out a possibly unwise amount of money on it. But this thread is about feature requests, surely, not listing existing features and saying how great they are?


    If I want a sound that an additive synth will do well, I'll go to Razor or Harmor.

    Personally I'm not that impressed with Harmor, it was going cheap and I wasn't persuaded, but I parted with over £1000 for the Virus, so what does that tell you?

    Horses and courses. I love 'em both and I ended up buying it.

  • I wonder if I'm too used to using subtractive synthesizers to get on well with things like Sytrus and Harmour. It may also be I'm too used to reproducing sounds made on subtractive synths or making real sounds which I find easier on a subtractive synth - a WWII air raid siren and screaming crowds was my last effort - wheras additive specialises in the unreal... I know people go on about the fact that pianos can be done in additive synths and not subtractive, but I haven't heard a convincing effort from either - aside from the Roland V Piano and I have no idea how that works, and whether it is sample based or not.

  • I wonder if I'm too used to using subtractive synthesizers to get on well with things like Sytrus and Harmour.

    Have you tried Razor yet?


    I'm not familiar with the other two you've mentioned but I just love the sounds that come out of Razor. I haven't delved into it too deeply as yet, but up to now there's been no need. I like to go through the different banks & so often I'll find myself thinking - "Now that is a lush sound" & then just using it as is. No tweaking required.


    It doesn't come with any piano patches though so I don't know how well it would create them, but, if the other patches are anything to go by I bet it would be able to do them really well.

  • Why not just create a multi patch using 16 parts (16x3oscs = 48 oscs!) to generate the required partials > record > send back into Virus to filter/apply fx etc?


    A little convoluted but still seems doable.

  • I guess your limitations would be that (a) under 400Hz you need more than 48 oscillators to cover the audible range (although at the area just below 20kHz the ear is not distinctive enough and you could just put white noise in there and no one would notice); and (b) 80 voices max divided by 16 channels at the same time means you are downgraded to 5 voices.
    Another thing you should pay attention to, is that semitones alone will not give you the accurate sub-harmonic, and changing the Virus' tuning from equal temperament to just intonation will scramble everything an you will have to re-tune.

  • Why not just create a multi patch using 16 parts (16x3oscs = 48 oscs!) to generate the required partials > record > send back into Virus to filter/apply fx etc?


    A little convoluted but still seems doable.


    Awesome. After a quick scan through the manual I found the feature you're talking about, 'Input Mode'. In theory that means I can use any synth I like to generate a waveform to be shaped by the Virus filter section (and ring modulator by the looks of it). For monophonic sounds that could work - and no recording would be required. (I could pipe a soft synth through a spare output and into the Virus.)


    Nice. Now if I can sort out the hideous MIDI timing issues I get when using my sound card, I shall play.

  • I wonder how 'additive' the hypersaw synth engine is - I guess it has to be on some level... Which means what you speak of may be possible as a modification of that - you don't lose too much polyphony by using the hypersaw module...

  • I see. So a modification of the hypersaw to turn the oscs into sine waves and snap them to the harmonics. (Or a knob to pull them there over time, why not?) A dial to change the harmonic components used (saw to square to other interesting waves) then a few knobs to control some tasty frequency domain/phase effects.


    Great thinking. Only problem I can see is that the hypersaw only has 9 voices IIRC. 18 with the sub oscillator, but I have a feeling that's a clever optimisation. If you disabled the second oscillator I suppose you could get 32. Halve the polyphony from an already severely limited range and you get 64, which is beginning to get there...


    But I think Razor has 320.


    I guess it would have to be a balls-out 'this feature will be monophonic, alright?' type setting for arps, leads, basses and the like, but on reflection I think it would probably consume the whole machine.

  • Fair enough. Maybe it it possible, and that would be gorgeous. Depends if the programmers feel that the sonic palette is worth it.


    Knowing my luck the Virus TI3 will come out next month with four times the DSP, USB 2/Firewire, the ability to offload work to the host CPU, additive, 6 operator FM and physical modelling. ;^)