Serial Sound Design?

Question

I'm looking for a way to create sounds in a serial manner, so that Sound1 feeds into Sound2 which then feeds into Sound3 etc. I noticed when using Operator (in Ableton) today that 4 simple sound waves played in parallel, just sound like 4 simple sound waves, but if sounds are played in a serial manner the sound it creates is completely different. So I'm looking for a way to do this but with audio files and other sounds, I feel that layering sounds doesn't 'infuse' the different sounds enough and they can still end up sounding like different sounds that are just sitting on top of each other.

I hope this makes sense :)

Cheers!

Answer 1

Not sure what you mean here by 'feeds into'? Beyond summing audio together there's convolution, or using one source to modulate another like FM or AM. You could check out Soundhack for convolution which can work with any sounds.

Answer 2

You are talking about a fundamental concept in audio synthesis: control versus audio signals.

Control signals are not meant to be heard. Audio signals are. When you are using simple waveforms such as sine, square, triangle waves, you can use them as either control or audio signals with predictable results.

In Operator you have 4 oscillators, and those oscillators can be connected in different ways. If you connect the 4 oscillators in parallel, then you hear a mixture of simple waveforms; all four are acting as audio signals. If you start connecting oscillators in series, you are able to hear the result of, say, a sine wave controlling the frequency of the next oscillator. Things start to sound interesting when you connect oscillators in series, as some of the oscillators are now control signals.

From the Live PDF manual, page 408: "[Operator] utilizes four multi-waveform oscillators that can modulate each other’s frequencies, creating very complex timbres from a limited number of objects."

The final thing to be aware of is the frequency rates: suppose that you connect 2 oscillators in series and listen to the output of the last one. As you control the frequency of first (control) oscillator, you will notice that low Hertz values correspond to gradual changes in the resulting sound's frequency. This is also known as vibrato. As you speed up the vibrato, there is a point at which the control oscillator is going fast enough that it ceases to be perceived as rhythm, and is rather perceived as altering the frequencies within the resulting sound. These sounds are called sidebands.

All this is covered in a basic synthesis course, so I suggest you do some reading.

Answer 3

I don't know of any VST/Softsynth platforms for doing what you are after but I can recommend a different approach.

The Composer's Desktop Project (CDP) is a (free) collection of commandline tools for processing audio files. I won't go into detail on the tools themselves as there are hundreds of them. A good number of the tools allow you to use an infile as a control source for the output, in addition to your starting signal. Since the operations are performed in commandline serial manipulation is very easy. You perform an operation on your starting file, then use the outfile as an infile on your next operation.

You could take this a step further by creating batch files that perform a series of operations in one go.