How I work

Last updated: 2020-05-05

Building Software

The core of the instrument is a combination of existing (commercial) audio software (a digital audio workstation or DAW) and completely new software developed by me. This new software is needed for detecting gestures and object movements in front of your webcam. It converts what the webcam “sees” in MIDI messages that are sent to a DAW, that can process the MIDI messages and produces the audio.
For me, to build this musical instrument, I had to experiment a lot with different DAW’s, settings, audio samples and so on. This required programming skills but also experience and sensibility for musical performance. After some time this resulted in special software functionality for experiencing a musical feeling while playing the instrument, of course in combination with the DAW used.
So building the instrument is not simply straight forward but an iterative process, cycling from programming, processing sounds, listening and managing the DAW to a lot of reading and learning. It took me several years.

Sound recording

For this instrument, I need a lot of sound samples. Until recently I made some 200 recordings on El Hierro, usually around a minute in length. With simple equipment I manage to make usable clips of 10 – 60 seconds or longer. Even of voices of people and animals from a cliff of about 1000 meters height, through the lower clouds. But also the milking of sheep, sparse drops of water in a well deep in the woods, the wind, the ocean in various forms, fierce, roaring in caves, phonolitic stones (stones that produce a tone when you knock against them with another stone. Greek: lithos = stone, phonos = sound) and of course the silence!


The recordings of stones are impressive. They are highly manipulatable too. There even is a whole lithophonic rock near the village of El Tamaduste, Roque de las Campanas. I went there, and I looked if it was possible to climb on it for making recordings, but that turned out to be quite dangerous… hands and foot work with perpendicular abysses on both sides. Not done.


But I went back a few times and I found some large stones along the path to the rock, about 70 cm long, which seemed to have fallen from above. Just tried to beat it, and … yes, a high “piiiiing”. That was easy to record. Now I can manipulate the pitch and create a kind of marimba-like instrument. These are funny finds and I gratefully use them.

I also made a few demos for people that I knew. I played the recordings for herreños (people of El Hierro) and they were very recognisable for them, often they could indicate the exact location. So for me it was affirmative that the quality of the recordings was sufficient and that no sophisticated apparatus is needed. Most DAW’s can generate interesting effects, controlable with my software.

Working on the spot


I go to interesting places of the island to make recordings on the spot. It depends highly on the weather and wind conditions where I go. There I put my iPad in Airplane Mode (to prevent network connection sparks in the recordings), connect the microphone, launch the recording app and keep it at hand. I often first try to do the student exercise of Murray Schafer: “Just stay quiet and listen for some time. What do you hear?”
When there is something interesting to record, I switch the app On and afterwards take pictures of the location.
In windy circumstances I use some self made covers for the mic.

Processing the samples

At home I transfer the photos and audio files to my computer. I administer the recordings with the photographs in a self made journal application for this.
I do some “polishing” on the recordings and try to cut and save nice parts. Important things are to keep the volume of samples rather equal (normalizing), to eliminate wind noise (equalizing) or isolate the subject from the background with sophisticated other methods like noise reduction. This requires a lot of experimenting, because not always the results of what I do is to predict. A lengthy process, although I have achieved nice effects as result.