(from the "Getting started" forum) - an "old man" trying to vocode via free MULAB plugins

[CesareBrizio]

Dear All,
greetings from Italy - first time here, please excuse both my suboptimal English and my deep ignorance. Last time I tried to make music with a computer it was 1990, and I used a Voyetra sequencer under DOS...
Now nearing 59 years of age, after 25+ years I am trying to understand whether or not I can set up sort of a vocoder for free, using only freeware modules.
I read all that I could, and it seemed to me that the quickest way was using the simplest VST plugin available for MULab (I had to learn very quickly and very superficially some MULab fundamental too).
I'm very clumsy and uncertain if my moves, and I'm very far from obtaining my objective.

In a few words, I would like to:
1) Record beforehand an audio track in which I talk or sort-of-sing (in the attached illustration it's in "Rack 3") - only the right channel is kept because the left is reserved for the carrier generated by "MDA vocinput". This existing audio track would be the raw material that I would like to vocode.
2) Activate the VST effects "MDA vocoder" and its accompanying "MDA vocinput"
3) Use the virtual midi keyboard and the basic synth (in one of its cleanest possible settings) to generate a different carrier frequency depending on the midi key pressed (I use the F-keys of the computer keyboard)
4) Feed the carrier in the "MDA vocinput" setting the "tracking" knob to FREE or QUANT so that (I believe...) it will be the input from the synth to determine the carrier frequency
5) Feed both the audio track in "rack 3" and the carrier form "MDA Vocinput" into "MDA Vocoder"
6) Get my voice vocoded according to the MIDI keys I press when I play the audio

Something in fact happens, in the general direction that I supposed, yet it's all very unpredictable.

Just to say one, with my voice muted I would expect that the 12 function keys (that in fact generate 12 successive tones for as many successive white keyboards) would generate a carrier in their respective frequencies. If I output "MDA Vocoder" to the master output, i don't get anything similar to a clean carrier, as its frequency seem more or less random (e.g. C4 get an higher pitch then A5).
Another thing that escapes me is that my "original voice" - the audio in Rack 3 - remains more or less clearly audible and is not completely obliterated by its vocoded version.

It's obvious that I'm missing something fundamental: I hope that from the attached illustration you can help me with some general hint without wasting too much time.

Hints may include an even simpler (yet free...) vocoder or relevant links to tutorials.

Many thanks for you attention, warm regards and all the best to everybody.

[pljones]

From my reading of the MDA VocInput help, that plugin is meant to generate "voice-like input" from another source. If you already have "voice-like input" - Rack 3 - I don't think you want it.

You should just have the modulator (your voice) and the synth (whichever synth you want to vocode) feeding into the L and R channels of MDA Vocoder.

Obviously this means the two inputs need to be mono rather than stereo. A quick way to get this is the Mixer Strip set to 0 stereo width on each of the modulator and synth (if there's a quicker way, I'd like to know). Then route these to the two inputs of a stereo combiner and feed that into the vocoder.

I've not actually tried it...

(I also notice in your screenshot you have Basic Synth Rack disabled - was that just at the time you took the screenshot?)

[CesareBrizio]

thanks pljones.
Yes, the synth was disabled at the time of the screenshot.
Regarding MDA VocInput, it generates a steady carrier when in OFF mode. And the short help page that you kindly linked states that it operates in three modes, two of which require an input (FREE- Track input pitch, QUANT- Track input pitch to nearest semitone). Hence I supposed that some pitch should be input in MDA VocInput.
In any case I'll try to follow your suggestions. One of the problems may be that, rather than preparing a mono audio file (my voice in rack 3), I made it stereo with an empty channel. So, before anything else I'll have to convert the voice to mono. Then, I'll try the trick that you suggest (Mixer Strip to 0 stereo width, plus stereo combiner) may work.
During the weekend I'll test your suggestions, again thank you.

[pljones]

Hi again,

I just tried my idea (heh, really I should do this first...) and... it worked!

The project, along with some sample vocals, is here:
http://peter.drealm.info/media/MuProjec ... 123613.zip (22M)