Vocoding for the uninitiated

[dakta]

This isn't a reason specific question so I've thrown it in da kitchen.

I've written (almost) a fast paced electronic song which is very 80's sounding, according to friends it is very 'PSB /Chris Lowe'

whilst its not a pop song I am putting some words over it, which will be vocoded (partly to mask my awful voice but also for electronic effect).

I have bought a mic, and I am getting some interesting results with the vocoder in reason, just wondering if anyone has any general tips to get the most from it - because at the moment I don't really know if I should be concerning myself with pitch when recording (should I be speaking, singing, or paying any attention to pitch) when recording, given I'm going for a deadpan spoken Neil tennant/Chris Lowe sort of effect

Any thoughts?

[selig]

Speak as clearly as you need to speak to get the desired effect. I’ve found I often need to over annunciate, sometimes a LOT, to get clear vocoding results. Pitch is totally ignored, BUT you may be able to get clarity with some pitches over others. So pitch matters in some cases, but doesn’t matter with regards to the final pitches used. In some cases you can get great results doing a sort of scream/whisper.

You’ll need to set everything up and listen while singing different ways to ‘test’ the results and find the sound you’re looking for. It does take technique, just not traditional technique, so it may take a little practice to get ideal results. I say all this from a total non-singer approach (me). It’s almost like the difference between playing acoustic vs electric guitar: playing one doesn’t mean you can automatically play the other…

[dakta]

Hi, and thanks Mr Selig, that's brilliant starting point/tips - I definitely think you have a point with the annunciation, I have a bit of a drawl I guess for want of a better phrase and letters seem to smudge into each other, getting that clarity is haaaaaaaaard

[crimsonwarlock]

Another tip: make sure your carrier (the synthesizer) has enough high frequencies to fill in that part of the speech spectrum. Even mixing in white noise with the synth sound can help a lot. Or distorting the synth before it hits the vocoder, which also introduces higher harmonics.

I used an analog vocoder back in the eighties, and I discovered (then) that using my DX7 as the carrier gave me the option to add loads of higher harmonics, that made the result so much more pronounced.