Is there a way to take a duplicated output and adjust the phase and pitch of it?

notkingoli2
notkingoli2 Member Posts: 3 Member
edited February 2023 in Building With Reaktor

I'm trying to create a unison pannel for an FM synth Im making, and I was create a manner that both saves CPU and also creates more stable FM by not just duplicating oscillators but rather duplicating the output and adjusting the values from there. Is there any way to do this? I'm still new to core and have been looking in there but I havent found any useful way to do it.

Thanks

Comments

  • colB
    colB Member Posts: 761 Guru

    Chorus and phaser FX are implementations of what you are describing. There are implementations of both in the factory library

  • notkingoli2
    notkingoli2 Member Posts: 3 Member

    They're definitely similar effects, but chorus implements delay not starting phase change, and also I am asking because I kinda just wanna learn how to make my own to tune it a certain way. If i can make a way where i have control over specific voices n stuff i think i can make something really cool. So if you know how to design the DSP let me know!

    Cheers

  • colB
    colB Member Posts: 761 Guru

    Chorus uses a delay line to implement a pitch shift that gives the effect of two or more voices... the implementation causes audible artefacts that give it it's characteristic recognisable 'sound'.

    The reason these FX are a thing is because what you are asking isn't possible without some sort of audible artefacts. You could implement a traditional chorus or flanger or phaser. Or you could try a more modern pitch shifter type of approach maybe using FFT or some granular, but whatever you do it will have some recognisable characteristic of that approach.

    You really can't make an 'effect' that sounds exactly as though you are using twice as many oscillators to implement unison.

    There are approaches that work well in very specific situations, like using a wavefolder type approach to 'multiply' saw waves... works great for saw and sounds very much like unison, but only with saw waves - not going to be the same when applied to FM based sounds.

    There are probably other tricks that work well in other limited situations. But I think if you are doing FM stuff and you want it to sound exactly like multi voice unison, with no artefacts or 'colour', then you need to implement it by adding more oscillators. Shouldn't be too bad though, FM is fairly cheap. Reaktor has a built in unison feature which might help in terms of cpu

  • notkingoli2
    notkingoli2 Member Posts: 3 Member

    Wow, that was actually a really informative reply thank you for that.

    Just one question tho, because I was under the assumption that stacking oscillators in FM would actually mess up the clarity of the FM. Like if 2 voices stack and starting phasing eachother out, wouldn't that mess up the FM amount. Thinking that this was a problem is what originally lead me down trying to find a way of just duplicating the output as opposed to making multiple oscillators.

    Cheers

  • colB
    colB Member Posts: 761 Guru

    Stacking oscillators in unison (detuned otherwise there's no point), will always reduce clarity. That's kinda the point, thicken up the sound add texture...

    Its somewhat confusing... the distinction between oscillators and operators. One way to look at it is that an FM oscillator might contain multiple operators... or maybe better to call is a voice...

    It shouldn't effect the operation of the FM process, you would have the whole FM 'voice' repeated, so if it's a 4op voice, you repeat all 4 operators as a block. Not the individual operators... So the FM between the operators within the voice will not be effected.

    Easiest option would be to use Reaktor's built-in Unison feature. That should just automatically work assuming the Synth works :)

  • Studiowaves
    Studiowaves Member Posts: 451 Advisor

    Sounds like you are trying to fatten up a sound. You might use an existing envelope generator for one of the operators and feed it into a delay. Feed the output into the delay then balance between the two. How many operators are you using? You could also use a dedicated envelope generator connected to the delay. There are several envelope generators to choose from. I'm sure you'd get some interesting sounds like that. The envelope generators aren't cpu hogs Adding a dedicated one might not affect cpu much with a low voice count.

  • Michael O'Hagan
    Michael O'Hagan Member Posts: 93 Helper

    So I decided to give this a go using a frequency tracked single cycle waveform record and playback system, I got some pretty good results.


    Here...



    There are three snapshots in the ensemble.


    Snap 1, no arp, playing individual notes sounds great, like true unison.


    Snap 2, with arp, gets some artifacts/pops/clicks, it doesn't like rapid pitch changes.


    Snap 3, Arp with glide, lots of artifacts/pops/clicks, it hates constant pitch changes.


    If anyone's got any ideas on how to eliminate the problems with constant rapid pitch changes, please, let me know.


    Feel free to use this for your synth if it fits your needs.

  • Michael O'Hagan
    Michael O'Hagan Member Posts: 93 Helper

    I figured out how to make one that handles pitch changes and glides properly, use this one.



Back To Top