Welcome to Mats Claesson`s course in Vokator.
I’m not in any way associated, employed or have gotten any form of payment from Native instrument. I have payed for all programs that I use including those from Native-Instruments.


Vokator is a program made by Native Instruments, the same guys that makes, Reaktor, Kontakt, FM7 and Absynt (to name a few).

Vokator is a pretty unique program but let me start with a "warning", it NOT as impressive sound vice as all the programs above. What it does and how it dose it, is impressive, but the sounds that Vokator can make is not always as "flashy" as lets say Absynth.


Vocator is a Vocoder :-)

A vocoder is a "thing" that morfs sound qualities from one sounds in to an other sound. To put it simply, it takes the timbre from one sound and the volume from an other and makes a whole new sound! Well, its not that easy but not far from. I will tell you more about how it works later.


Vocoders have been used in popular music since the mid 1970.
They have mostly been used to produce this "synthesizer that sings" kind of sound. Listen for example to Herbie Hancock from that period (sorry cant give you any musical example,s if I did I would probably get sued....-(


In this course you will learn some of the basics "behind" Vocator called FFT, Fast Fourier Transformation, and all (most) about the basic programming of Vokator. I will also try to give you some my own personal view of what Vocator is best at.

But as always, don't trust anything but your own personal taste :-). But you could listen (read) to what the teacher says (that's me).





MIDI keyboard

To follow this course you have to have a MIDI keyboard.

To use Vokator you have to play on the keyboard.


Start Vocator.

You can use Vokator as a standalone program or as an VST pluggin.

All the function and programming are the same.

In this course most of the examples are done by using the standalone version of the program.

Listen to the presets.

Its always a good idea to spend some time with the presets. Lets start with the preset that really explains it all, "Classic Vocoder". It will be used a lott in this course.

Look at the video to the right. It shows you how to open a Preset in Vocator. This is the one of the two way to open a preset in Vokator.


Read the" what to do" below and look at the video

(use the pause button if I'm going to fast....).



If everything when well you can see the name "Classic Vocoder just below the IMPORT button.



Sound and MIDI preference.

You have now loaded a preset. Lets check the settings for the sound card and MIDI interface. Once programmed, they could be left pretty much untouched.....


Look at the video (and use the pause button if I'm going to fast).







Play a key on your MIDI keyboard.

You should hear something like what you hear in the video example to the right (yes this time the video example has sound :-)


Look at the start of the video. The cursor shows you two areas. The first one shows a little blink every time you press a MIDI key. The second area shows the output of Vocator. If things doesent work (no sound) these two places could be a god thing to check out first. Is Vocator receiving MIDI is there sound out of Vocator?



Classic Vocoder is a good example off what a vocoder does best.

I guess you have heard this type of sound before. This is the most typical Vocator/vocoder effect, and I definitely think this is what a vocoder is best at, put a harmonic (synthesiser) timbre "into" something we all recognize clearly, like a voice.


In the preset Classical Vocoder the human voice is a recording, that is played back by Vokator's own sampler.

Lets listen to another preset.

Look at the video below. Open "Diamonds are 4 ever".


Try "Diamonds are 4 ever" .

In this preset  its impossible to hear if there are any recorded sound like the voice you heard in Classic vocoder.


Modulation wheel.

Vokator has something the "Native guys" calls Morph.

In the example "Diamonds are 4 ever" and most of the preset the modulation wheel controls the morph function.


Play the move and listen to the sound,......nice isn't it? In the video i move the modulation wheel very slowly.


Lets import the G Preset.

You will find it in the Preset folder > Synth-Sampler > FX .

On the first page I told you that Vocator was a vocoder and that a vocoder " merged" caracters from two sounds together.... But there is more, Vokator can also take only one sound and mess it up. The G preset is one example of that.


Try it out

In the video you first hear the processed sound and then (I just pushed one button:-) the unprocessed sound.


-That's all I have to say about presets.

Its a god idea to look around yourself. There are a lot of preset to try, but wait with the Live input folder, we will come back to that one later in this tutorial.

And for what its`s worth I don't think you will get overly surprised, but maybe you find something you like. Vokator dont`t have a lot f programming possibilities but don’t expect anything in the league of Reaktor or Absynth.

What does Vocator do?

Vocator is basically a vocoder, its the most advanced dedicated vocoder commercially available (that I know of) and it does a number of things other than straight vocoding.


So what is vocoding?

A vocoder uses two sounds, usually one is a "live" source, like someone talking in to a microphone, and an other synthesiser sound, like a sawtoth.


The idea is the one sound (the live sound) "controls" the volume of the new sound and the other sound (the syntesisersound) controls the timbre.

Volume is not just the overall volume. In a vocoder the frequency range is divided into a several frequency bands (or filters). And the olume changes inside these frequency bands (can be anywhere fom 8 to hundreds) ar tracked.

I wont`t tell you how a vocoder worked in the old days but the principle is almost the same in the modern version, Vokator. The main difference is that Vokator can produce an effect with much higher resolution compared to the old days..


Vokator does vocoding a little different than in the old analogue days.

Just to warn you, the following descriptions on how Vocator and fft works is far from complete, sometimes i even not fully correct :-(

I have tried to make this as simple as possible, yet it will give you most, if not all, information you need to understand and program Vokator.

If you want to know how this works in details you have to read my fft course (go to the homepage).


In Vokator both sound are analyzed.

The analyze part of Vocator (done by fft ) split the sound from lowest to highest frequency in filter bands. Depending on the programming, it could be from 63 to 511 filter bands.


Lets say you analyze the two sounds with 255 filters.

If we use the same resolution as on a CD (44100 samples per second ) we have a frequency range of aprox 20-22.000Hz. Our 255 filter bands is divided equally over the frequency rang, so every filter has a frequency range of aprox. 86 Hz.


The fft analyse finds the strongest frequency in each of the 255 filter bands, and the volume of that frequency. When the analyze is complete we have 255 "filters" that have an individual frequency and volume.


Lets see how this works.

Listen to our"live" sound in the video A. "live" source a human voice.

If you watch carefully you can see that there are barographs in the display that shows the volume on each filter band.


The other sound, the synthesiser sound a sawtooth waveform, is analyzed in the same way. Listen to the sound in the video B. Sawtooth synth


We can then use the the volume envelope (called spectral envelope) and the timbre (called phase) from the two analyzed sound to make a new sound. Listen to the two combinations in the video:


Vocoded sound B*A.


Vocoded sound A*B.


Vocoded sound B*A is the classical Vocoder sound, a really "classic" sound. Here the volume changes are controlled by A, the spoken sentence, and the timbre is from the synth sound, the sawtooth


The first letter is the spectral envelope and the second letter the timbre.


Does all this sound complicated to you?

You didn't get it? Then you have to spend some time on the few lines above. Its important that you understan th basics principle, Spectral envelop(volume in filterbands) and timbre.

A. "live" source.
B. Sawtooth synth
Vocoded sound B*A.
Vocoded sound A*B.


Time to look under the hood.

The thing that make all this work is the fft (FastFourierTransformation) a program inside Vokator.

This is something I could talk a lot about :-) but here I will give you the short version.


The fft algorithm analyses the sounds spectrum , the sounds timbre. It does that by "cutting up" the sound into small windows, small parts, and then it analyses these windows one after another. The size of the window (the length of the sound/window) is measured in samples (on a CD there are 44100 samples every second....). We have 4 different fft window sizes in Vokator. 1024, 512, 256 and 128 samples. Look at the video and you will se how you change the window size.














So does changing size mean something?

I'm glad you asked :-) Lets listen to how window size affects the sound.


Open the Classic Vocoder preset, look at the video, it shows you a faster (better) way to open a preset.




















A window that's 1024 samples long.







Listen to the first video. Here you can hear how a Vocoder effect, with a window that's 1024 samples long, sounds. You can hear what the synthesiser"says" butt the words are somewhat blurred, the articulation is not perfect, its difficult to hear the words at the beginning of the example. The timbre is god there are no artifacts, no noise, just a smooth sound.


A window that's 512 samples long.







Its definitely easier to hear all of the words with a window size like this one.

The timbre still sounds quite smooth, but maybe a little bit "harder" ...?


A window that's 258 samples long.







The articulation of the words is getting even better in this example. You can even hear the first word "computer" clearly . The timbre, the sound quality is getting a bit ruff. You can hear noises and strange artifacts that was not there in the examples above.


A window that's 128 samples long.







The articulation is very god. You can hear the text very clearly. The timbre is very distorted with a large noise part.


There`s always this relation in an fft program, smooth (god) timbre or detailed time resolution.



matsc 2004 rev. 2009