Formant SoundScapes

It has been a while, got lots of other work to do… But in the meantime I’d like to give you a look inside “the lab”:

This is a realtime recording of me singing in my microphone. The Sound seperates the formants and I’m able to shift, freeze and smooth them – everything is work in progress by now. The formant extraction is really a tough thing to do. The whole thing is processed with some Granular Reverb as well.


Deeper in the Wire

In the last post I described how you can split up a spectral signal and how you can smooth the amplitudes and/or the frequencies. Now it’s time to get deeper

Johannes had a great idea in the forum: What if we had a smooth transition from original to processed signal? The Delay module’s feedback doesn’t work at sample rate so we need to build our own feedback. Thanks to Pete this is not a problem



So what’s happening here?  Read more

The Wire Between

What kind of signal is the SpectrumInRAM and/or the LiveSpectralAnalysis producing and how can you access single partials and process them seperately? This was the question I asked myself a while ago. Of course you can use the SpectrumTrackSelector or any of the Spectral Modifiers but there’s no way of splitting or summing the signal itself so you always end up with many processor-hungry OscillatorBanks. At least that’s what I thought… but after watching “The Wire Between” by Pete Johnston and with some help of the very same Pete in the Forum I worked it out

First you have to understand how the signal is presented to the OscillatorBank. To see it you have to split the signal. So just put a ChannelJoin after the SpectrumInRam and delete the delay module which will be added by default. Now you have the amplitudes on the left and the frequencies on the right. In this case I used Alien threat new h.spc which contains 252 partials. Insert an OscilloscopeDisplay to see what’s going on. Set the PulseTrain Period to 252 samp:

Read more

ADDAC502 Lissajous Curves

While searching for complex control signals Bruno showed me this module: The ADDAC502 is a quite unique module using Lissajous Curves to generate its control signals. After seeing the resulting curves I instantly wanted to have it – so I rebuild it in Kyma! I recommend downloading the Interactive User Guide (Win/Mac) and play around with it for a few minutes so you know what’s happening.

I started with 4 Oscillators, 2 for the x- and 2 for the y-axes.


To see the Lissajous Curves we connect Read more


It has been a while because I was busy supervising one of my projects at the mixing stage, but here it is: the next tutorial!

What about the dynamics of all those layered atmosphere sounds you usually build for a movie? They sound way too steady to be natural.. So what you have to do throughout the mixing is changing the volumes of the individual layers constantly to achieve natural dynamics. This is easily done by moving the faders but I thought maybe there is a way to do it on a meta-level by changing parameters and have the dynamics controlled by some kind of algorithm.

My first question was: How do I change the dynamics? Soon I realized that I need some kind of control signals e.g. LFOs and have them control the amplitudes of the samples (allthough there are maybe 100 other ways to do it – let me know your ideas :)). The boring thing about LFOs is their waveform because it’s repeating over and over again, never changing, which is unnatural. But anyway let’s start with 4 LFOs (LFOs are “Oscillators” in Kyma – no need for a dedicated LFO because it’s just an Oscillator with a very low frequency).

So set up something like this and make sure you mute your speakers when you compile it: - atmodynamics 1

Read more


Based on Douglas Murray’s guest contribution on about creating AirFill with a convolution reverb plugin (Unlimited “AIRFILL” to Fill your Every Need) I build a little Sound in Kyma.

Let’s start with the “CrossFilter_Long” prototype: Double-click it and basically all you need is right there: - airfill 1

Now replace the “source” with a “Noise (white)” prototype and replace  the sample in the “response” with a little piece of AirFill. Read more


There are certain effects you hear in about every modern movie – whooshes are one of them. But instead of using all the library stuff why don’t you create your own whooshes? Usually that’s what I do but using Pro Tools this can be quite time consuming because it’s not possible to build up a chain of processing and then choose whether to bypass certain things or not. When you want to reverse the sample at different stages of the processing things get even mor complicated. Luckily we have our Kyma Workstation

So what kind of processing do we need? I usually use the following:

– Change the Playback Rate (Varispeed)

– Pitch Shift

– Delay

– Reverb

– Reverse the Sample at different stages of the chain (e.g. reverse the reverberated sample)

Let’s begin with a sample. Changing the Playback Rate is very easy, just multiply “default” in the “Frequency” parameter field of the sample prototype with a hot value of your choice (e.g. default * !Varispeed). The next step is to build a bypassable Pitch Shifter using the “Polyphonic Pitch Shift Delayline” prototype and the “SelectableSound” prototype (just read the description, it is very easy to use – make sure to use a Toggle in the VCS). Since my sample is stereo I split the channels and process each of them. Afterwards I have to boost 6db to preserve unity gain. Have a look at the picture below: - whooshalizer 1

Read more