Vocoder

I’m in the bed right now but, it does legato (or portamento) between notes, but don’t want chords. And I didn’t find a parameter to change this behavior (… Yet ?)

Nice that the vocoder is now working for you. The parameter for the portamento strength is under Ctrls#5

2 Likes

Get those zss files flying!!.

In this particular case the original .zz file seems to 500 the snapshot section of webconf…even after a reboot

We really gain the more we exchange zynthian data at that level…

Looks rather accusingly at the zynthian & 1820 and the conveniently placed microphone…

This is the MIDI & Audio example…

So things have moved on such that a vocoder seems to now include components to make the carrier ( or is it source…?) waveform rather than process an external input. I’m sure that’s probably out there but I haven’t’ found it yet.

Will the recorder audio for workflows do this automatically to protect anonymity, in this security conscious world?

1 Like

I can confirm that this:

Is all it takes.
In my case my external USB audio in/out was set to loopback, which was burying the vocoder in microphone source and hiding the output.
It works great, polyphony and all.

5 Likes

Yep! I played a bit yesterday and polyphonic vocoding sounds cool. Shame it has limited sounds. It would be nice to vocode a synth with more sonic configuration.

3 Likes

Wouldn’t Zynthian be able to put the Polyphonic Vocoder after any synth in a chain? You could even drive the synth and vocoder from separate midi controllers or one, he said, without actually trying it.

1 Like

I was also thinking about that today. Tal Vocoder can work as a pure vocoder filter. But the audio routing could be a challenge and great differences are not to be expected, as a vocoder, by design, works best with very overtone-rich sustained sounds.

2 Likes

I am still not home but have regained remote access to my Zynthian so am trying some workflows.

I see that Tal-Vocoder-II has sound crafting in the form of oscillator waveform mixing, tuning, portamento, etc. I don’t see filtering of the carrier though. I am not sure how much of this will allow substantial customisation of the built-in sound generator.

There is an option to set the input mode to use the left input for the carrier (sound source) and right input as the modulation. (I observe that toggle inputs like input mode, polyphonic, etc. are not recognised as toggles in zynthian. That needs fixing.) Using this input mode it should be possible to vocode another synth voice but… zynthians audio routing makes this a little challenging, specifically its automatic mono / stereo routing - but there is a way to do it…

  • Add a synth chain with the sound you want to modulate (the carrier).
  • Add an audio chain.
  • Add a MIDI+Audio chain with Tal-Vocoder-II. (It is in the “Pitch” group.)
  • Use chain options to rename the audio chain, “Mic” (optional but useful).
  • Use chain options to rename the MIDI+Audio chain, “Vocoder” (optional but useful).
  • Use chain options to disconnect synth chain from “Main mixbus” and connect to “Vocoder” chain.
  • Use chain options to disconnect audio chain from “Main mixbus” and connect to “Vocoder” chain.
  • Use chain options to disconnect unused audio input to audio chain, leaving just the micrphone input.
  • Pan the synth chain hard left.
  • Pan the audio chain hard right.
  • Set vocoder Input Mode to full value (enables split input mode).

Simples! This should work. I will try when I get home.

[Edit] Yes - it works. We may want to simplify this workflow but for now, you can do it.

10 Likes

I love this thread.

What “carrier” is fun to play with ? Should we stick to hard synth or a guitar could work ?

You tell us. Try it and see.

Don´t get it full. You setup three chains:
1st synth chain - carrier, f.e. OBx
2nd audio chain:
3rd MIDI+audio-chain: TAL-Vocoder
You rename the first chain “Mic”, hm not the 2nd…

1 Like

I have modified the instructions - thanks!

Okay…
It can be fun, but right now it sound bad, mostly because of me I guess. I’ve put two audio chain instead of a synth and an audio one, but yeah : renaming is useful. One is “mic” and the other “guitar”.
So : Vocoder with a Guitar as a Carrier : :face_with_monocle:


(i’m not proud… but it is to thank you for let us know how to do that)

Is


there a electronic splat on them…?

I have no clue on how to translate “electronic splat” so… maybe ? :no_mouth:
021-vocoder carrier test.zss (10.6 KB)

Vocoder_FluidSynth_FluidPipes_Bottle_Chiff

Vocoder_FluidSynth_FluidEnsemble_Synth_Voice

It looks like the WAV file has a chunk of data that probably isn’t of audio origin. The two thick blue lines in the WAV file picture dump.

It could be an artifact of a problem in the generation…

Here are the actual samples…

It appears that the data is hitting some kind of endstop and behaing badly…

Not your fault, possibly a problem in the vocoder…

I see.
No. This is “real”, it happen when i talk to loud in the mic or hit to loud my guitar strings. One or the other hit a “peak”. Sorry, it hurt the hears :star_struck:

No problem. I’ve spent a little time listening to audio data and have a slightly over reactive approach to this sort of thing, it does look like an interesting one to get the measure of. Have you got a source that you can gradually increase the volume of…and use that as a source… ( or carrier, don’t expect me to ever remember which is which, its a symmetry and I’m really bad at those…)

Ha! That is a funny effect. The vocoder is wrapping the level so that excessive values are offset (kinda inverted). This should be reported upstream.

I’ve put the zss up above. But then I have to many problems to help us finding the “why” it happen :

  • I’m a simple user
  • With “a” mic (a good enough one, no brand, no known quality (… Well, looking back at it : it’s a Philips SBC MD650)
  • With a Zynthian V4
  • With some kind of input amplification that feel good to me -audio level gain set to “20db” for both jacks, that work with my mic, guitar, bass, other bass, kazoo-
  • With some kind of amplification into the vocoder TAL-Volume set to “72” and i have no real clue of what it does

Knowing what every parameter from the vocoder does should help to prevent the peak, but i’m not ready yet.