Does anyone know how to make Reaktor blocks polyphonic?

Ofuskate
Ofuskate Member Posts: 12 Member

I’ve been learning how to build synths using the blocks in Reaktor, but I wasn’t able to figure out how to make a synth, built with blocks, polyphonic. Is it possible to do this?

«1

Answers

  • Sunborn
    Sunborn Member Posts: 2,241 Expert
  • colB
    colB Member Posts: 762 Guru

    Here's a link to an approach I developed a few years ago. The Toybox polyphonic blocks use an updated version of this method. Have a look through the structure and ask if you have any specific questions.

    https://www.native-instruments.com/en/reaktor-community/reaktor-user-library/entry/show/10343/

  • Ofuskate
    Ofuskate Member Posts: 12 Member

    I saw that, but it seemed like the consensus was that duophonic is pretty simple to accomplish, but polyphonic is more difficult to achieve.

  • ANDREW221231
    ANDREW221231 Member Posts: 295 Advisor

    duophonic is actually probably more tricky, because you have to do the voice handling yourself


    actually the easiest way to get polyphony with blocks is cut the innards out of the 'instrument' level, and paste everything into a macro. then you can set the voices as desired and combine everything with a voice combiner

    blocks are inherently mono, because the ins and outs of the 'instrument' are necessarily mono

  • colB
    colB Member Posts: 762 Guru
    edited July 2022


    duophonic is actually probably more tricky, because you have to do the voice handling yourself

    It's actually not that bad, depending on how you approach it. Here's a relatively simple way to create a somewhat limited (but fun) duophonic patch:



  • ANDREW221231
    ANDREW221231 Member Posts: 295 Advisor
    edited July 2022

    oh, yeah using all 128 voices method, that's smart!


    i went a different route, using a separate instrument with around five voices.. so like three extra 'buffer' voices such that pressing another note when two were already were held down wouldn't kill the active notes. this necessitated a total spaghetti nightmare downstream, but it did work pretty good in the end

    the behavior of mine was somewhat different to yours, as it would ignore any notes pressed after two were already active, IIRC. and i think the priority worked somewhat differently to simple lo/hi, but i believe it pretty well matched the behavior of the arp odyssey vst clone Oddity

  • ANDREW221231
    ANDREW221231 Member Posts: 295 Advisor
    edited July 2022

    ya know one thing i was wondering about? how to do a good paraphonic patch a la moog matriarch or behringer poly d. it seems it would be fairly straightforward to gate four oscillators together before a filter and envelope/vca , but what about when there's release time on the amplitude envelope?


    you know what i mean? like, you would need to hold whatever last active voices until the release had finished its cycle

  • colB
    colB Member Posts: 762 Guru
    edited July 2022

    Not sure really. It depends how para it is.. so is the envelope shared, or is the gate shared but not the envelope...

    I think the 'standard' thing is shared envelope and Gate, in which case it should be as simple as just 'hocketing' the pitch around 4 oscillators, mixing the output and sending that through an envelope driven by a monophonic gate?

    I think there are other legit approaches e.g. instead of gates using triggers and AD envelopes. So separate envelope per osc, and because it's trigger driven, those can be hocketed too because we don't need to worry about release times causing notes to overlap with each other.


    EDIT: OK, I think I see what you mean. There needs to be at least some per voice switching, so how is that handled so that a voice can be off?

  • ANDREW221231
    ANDREW221231 Member Posts: 295 Advisor

    yeah, as far as i can get trying to think of this in my head, you would need some way to switch voices off, but its all kind of fuzzy i think i need to try to build something to get a clearer picture

    I think there are other legit approaches e.g. instead of gates using triggers and AD envelopes. So separate envelope per osc, and because it's trigger driven, those can be hocketed too because we don't need to worry about release times causing notes to overlap with each other.

    this makes more sense to try i think, this way you could only send the trigger to the voices that you need, and just send one trigger to all of them? 🤔


    ultimately, i think the best thing of all might be a 'best of both worlds' scenario, each oscillator having its own vca and envelope into a single filter and envelope. it seems the favorable parts of the paraphonic sound are mostly down to sharing a single filter, though im still interested in proper paraphony in reaktor as an academic exercise

  • colB
    colB Member Posts: 762 Guru

    though im still interested in proper paraphony in reaktor as an academic exercise

    I guess the question then is - What constitutes 'proper paraphony'?

    Draw up the rules - even multiple sets of them based on whatever paraphonic vintage or modern systems there are... then it's much easier to understand what can work and what won't

  • ANDREW221231
    ANDREW221231 Member Posts: 295 Advisor
    edited July 2022

    this is somewhere in the ballpark, though as predicted it has problems with long release times and notes not shutting off like would be desired. guess it needs some kind of logic for how many voices were last active so wrong notes don't bleed in. luckily i have a paraphonic synth in the volca keys so at least theres a reference i can work towards




  • colB
    colB Member Posts: 762 Guru

     luckily i have a paraphonic synth in the volca keys so at least theres a reference i can work towards

    There's no point without a clearly defined set of rules to work towards. Paraphonic has 'problems' by definition. it is a compromise. So before building anything, one would need to make a decision about what version of paraphonic to implement, and then draw up a bunch of rules for that version. Then it will be easier to implement, and also possible to know when it's 'done' :)

    Do you know of any sources for detailed info about the behaviour of any vintage paraphonic synths?

  • colB
    colB Member Posts: 762 Guru
    edited July 2022

    So your version has polyphonic amp envelope, so just the filter and filter envelope are monophonic. That's one way to go.

    I think maybe more paraphonic hardware had both envelopes mono.

    Here's one with just a single mono envelope and mono filter - so only the oscillators are polyphonic. I think to get the most from this, you might need to learn to play the cutoff control. That way you can manually accent polyphonic playing... something like that anyway :)

    Maybe something like this needs a 'hold' button or similar, so you can manually force the envelope to stay open for the release on the envelope. Currently this version always cuts off the release. So envelopes are really just ADS...

    EDIT: note that I'm still using polyphonic gate/vca on the oscs!

  • colB
    colB Member Posts: 762 Guru
    edited July 2022

    Here's another version. This might be cheating ;)

    I'm thinking about what EEs might do, and that a poly gate signal could have a 'soft off' like a release on the gate that could be implemented cheaply with maybe just a capacitor and a resistor, so way way cheaper than a complete envelope. Might even be 'free' depending on how the poly gate is implemented? (e.g. a natural consequence of intentionally sloppy switch de-bouncing...?)

    So I hacked that in, and some logic that triggers when the mono gate is released. Any poly gates that are active are held. That way you get the release on the envelope, and there is some tolerance to the release time, so it's possible to have chords release if you are precise with your timing. The soft-off thing makes it achievable. Also makes sense in the context of component count in hardware... maybe??

    All guessing of course :-)


    ...I still prefer the duophonic thing that keeps both oscs going all the time, and just brings them to unison when only one key is held. There is a simplicity and elegance to that, and it is very musical. It would be interesting to try and extend that to more voices, although maybe not historically accurate? or maybe it is?

Back To Top