Getting a vocoder sound without a vocoder

[mashers]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

[Loque]

In theory you need band pass filters to split the modulator, followed by envelope follower, which modulate a carrier signal, which are followed again by band pass filters. The result is summed up.

So yes, in theory you can, but you would finally build a vocoder...

In the same way you could synthesize a voice synth sound.

PS: for experimentation it would be interesting to influence the sound in each of the steps with other fx, like frequency or pitch shifter or other fx...

[mashers]

In theory you need band pass filters to split the modulator, followed by envelope follower, which modulate a carrier signal, which are followed again by band pass filters. The result is summed up.

So yes, in theory you can, but you would finally build a vocoder...

But it still needs a carrier though, right? What I meant was, can you make the vocal sound like it has gone through a vocoder but without a carrier?

PS: for experimentation it would be interesting to influence the sound in each of the steps with other fx, like frequency or pitch shifter or other fx...

That does sound interesting indeed

[ScuzzyEye]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

iZotope's Vocal Synth does just that. It figures out the pitch of the vocal and generates a synth tone to match (you can also send MIDI notes if you want, but it's optional). It's not a full AAA synth engine, but it can generate the most recognized vocoder style sounds. It also has other "robotic" sounding effects, all of which can be layered.

If you're wanting to use a completely unique synth patch. It should be possible to use Neptune to generate the note data based on vocal pitch (I think there's a Combinator involving Thor in the FSB) to trigger anything you want.

[Loque]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

iZotope's Vocal Synth does just that. It figures out the pitch of the vocal and generates a synth tone to match (you can also send MIDI notes if you want, but it's optional). It's not a full AAA synth engine, but it can generate the most recognized vocoder style sounds. It also has other "robotic" sounding effects, all of which can be layered.

If you're wanting to use a completely unique synth patch. It should be possible to use Neptune to generate the note data based on vocal pitch (I think there's a Combinator involving Thor in the FSB) to trigger anything you want.

Doesn't Neptune supports this too? A synth-mode?

[O1B]

You can get a vocoder sounds any number of ways. You do not need a vocoder.

Consider.. you are "manipulating the frequency content of sounds"

You could use a Phaser.... Set speed to taste...

Phase Vocoder
Two of the most important ways that musicians have used the phase vocoder technique are to use a sound’s Fourier representation to manipulate its length without changing its pitch and, conversely, to change its pitch without affecting its length. This is called time stretching and pitch shifting.

"The Pitch/Speed Relationship in the Digital World...."

You could use a Frequency Shifter... Set Frequency to taste.

You can manipulate the Sampling Rate....

Columbia University can help:
http://sites.music.columbia.edu/cmc/Mus ... /05_04.php
"... filtering, and a combination of spectral analysis and resynthesis."

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

[Loque]

I made such a sound by using Spectra and the spectral-envelops. It finally produced a "Yeehhaaa" voice. Sadly you cannot here the example anymore:
viewtopic.php?p=383013#p383013

[ScuzzyEye]

Doesn't Neptune supports this too? A synth-mode?

Neptune's synth mode is more more producing your voice at multiple different pitches at once. Like you can play chords, and get harmonies with one vocal track. Not really a vocoder effect.

It does have a detected (or corrected) pitch output, but it's range doesn't match the normal note CV input, so you have to use Thor (or some other CV utility) to convert it to note data.

[Boombastix]

You can get the vocal pitch out of Bitspeek as a CV signal. Send the CV to trigger a synth, maybe add Chords&Scale on top to get auto chords.
I haven't tried to see how useful this is, but in theory...

[Heigen5]

Has anyone used Predator for vocoding yet? It has a vocoder effect over the effect section.

[Loque]

Has anyone used Predator for vocoding yet? It has a vocoder effect over the effect section.

I tried several times without good results. Probably I did something wrong.

[Heigen5]

Has anyone used Predator for vocoding yet? It has a vocoder effect over the effect section.

I tried several times without good results. Probably I did something wrong.

Might try it out tomorrow myself, I'll let you know if it works out well or not.

[mashers]

Thanks guys. Lots to think about and look into here. Fun fun fun!

[zoidkirb]

My quickest/easiest method, feed the audio through Thor's formant filter and automate the x and y

[mashers]

My quickest/easiest method, feed the audio through Thor's formant filter and automate the x and y

Interesting. I use formant shift for vocals using Neptune, and in my current WIP I'm using Thor's formant filter on an electric bass. But I didn't think of trying it with the vocal. I'll give it a go. Thanks!

[sleep1979]

I tried several times without good results. Probably I did something wrong.

Might try it out tomorrow myself, I'll let you know if it works out well or not.

Did u try it ? Ive yet to try it but wonder if it works like a normal vocoder on vocals

[scratchnsnifff]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

While this isn’t an effect, you can use malström vocoder bands grain table. This will let you have that sound just using a wavetable

But if you want to make an effect. Take a delay and modulate the time. The DDL digital delay in the stock effects works perfectly for this

Take a combinator. Throw in the DDL in the matrix of the combinator set CV1 to the the milliseconds (MS) take a pulsar and hook the LFO cv out to the combinator CV 1 input. Set a slow LFO to modulate the milliseconds

After the DDL insert a band pass filter and after that an EQ. Make the bandpass filter really resonant. And on the EQ make as many peak curves as possible. Set the band pass and peak sweeps to one of the rotaries

Mess with the feedback of the delay, and for further “vocoder bands” add another delay with the same modulation but with a different range of milliseconds on one of the. Try 1-12 milliseconds and the other try 7- 20 milliseconds

After the second delay insert more peak curves and bandpass filters. The sweeping of the filters makes almost a phaser kind of sound, and the more bandpass filters the more vocoder sound

I highly recommend the air raid audio filter for this setup because there is a width knob. And messing with the width of the bandpass makes for such a cool sound

The Q knob on the mclass eq does the same thing, but in that case just use EQs

[scratchnsnifff]

Has anyone used Predator for vocoding yet? It has a vocoder effect over the effect section.

It’s decent. Very basic but it works

I wish reasons vocoder had a bandwidth knob, that’s how you can get extra vowely

[selig]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

Neptune (and others, I’m sure) can do this nicely in synth mode IMO.

[guitfnky]

I'm wondering if it is possible to process vocals so that have "that" robotic vocoder sound without actually using a vocoder. I don't like having to programme the melody into the carrier so that it matches the vocal and I can rarely get it to all match up because the vocals aren't quite to time. I know I could spend time tweaking the note data of the melody so it all matches up precisely, but there are still issues with this. So is there an effect I can apply which will make the vocal sound vocoded even if it isn't?

Neptune (and others, I’m sure) can do this nicely in synth mode IMO.

doesn't that require using MIDI to control the notes though? regardless, this is hands-down the best feature of Neptune, IMO.

[selig]

Neptune (and others, I’m sure) can do this nicely in synth mode IMO.

doesn't that require using MIDI to control the notes though? regardless, this is hands-down the best feature of Neptune, IMO.

Yea, you're right it does require a MIDI note to create the synth voice.
If it's your only option, then doing a "bounce audio to MIDI" first would get you at least part of the way towards your goal.

[mjxl]

Record voice, put in Europa. Profit.

If you set a envelope from 0 to 100 in the amount of time the sample is, you'll nearly be there already.
A couple of things to beafen it up and you should be good to go.