This short article presents the recorded audio track entitled Resonate which was broadcast 19/12/2020. This thirty minute sound scape is part of a series of works that explores machine sentiment analysis as a means for creating drone music. The sentiment data is coming from news articles, specifically from The Guardian newspaper and for this piece I’m using a single article as source. The machine analyses text and spits out sentiment data - i.e. how positive, negative or neutral the article is. It is also picking out keywords from the article and assigning emotional scores for each of the following : joy, sadness, anger, disgust, fear. There is quite a bit of data coming out of the program I wrote for this procedure, more than enough to play with. Play being the operative word here because it would be meaningless to say that the machine is making the music. Rather, I’m making these sounds with the machine’s help and the main artistic decisions are being composed by myself. That said, this is not some random plotting of data to music. Conceptually and technically there are some complex ideas being pursued.

So, what is actually going on here under the hood? Discussing the intricacies of composing music with images would be a good start although the history is dense and beyond the scope of this article. However, it should be said that the ideas are very similar and I’ve no doubt been somewhat influenced by my work with the visual. When talking about sound and image, there are common themes that bring the two together. Tone, timbre, colour, rhythm being some of the main concepts that have influenced the visual landscape since modern times. The pioneering works of Walter Ruttman, Oskar Fischinger, Len Lye and Normal McLaren are all good starting points for learning more about those special relationships between the two mediums. My approach then is taking inspiration from this and finding ways to translate data for making music. Perhaps another way of thinking about this conceptually would be the idea of mapping data to sound. Now, you could say that any kind of data could be easily mapped to sound and of course that is quite true. However, what I find particularly interesting with this data is its link with emotion and language. I’ll drop that just here for now, but there is lots to be said about my intentions and choice of data - more of that in future articles. For the moment, let me divert your attention to the sonic element and remark that I’m concentrating my efforts on the essentials of sound and to do this I’ve decided on the particular style, or rather mode of music, we call drone.

What makes drone music so appealing musically is that it draws one’s focus to the raw yet ever so essential parts of what makes sound - harmonic oscillation of wave forms that have the characteristics of tone, timbre, amplitude and tuning or perhaps frequency would be a better term. Traditional oscillators, as a sound source, have a limited scope in terms of timbre, ranging from the smooth round and wholesome sound of the sine wave to the harsh, resonant qualities of the saw. Complex sounds however come from combining these different timbres and eventually modulating some parts such as pitch and amplitude or syncing one oscillator to another. Interesting timbres develop from slight modifications of these parameters. For this to happen though, one must be attentive to the quality of the sound. Offset phasing and detuning with a little LFO modulation can breath a lot of movement into sound, making it evolve, sometimes in chaotic manners when mixed with other elements of a traditional analog synth setup. The sound can be shaped with further processing, often this comes from a mix of envelopes and filtering before being piped through some delay or reverb unit to add spacial qualities. As a process, making and configuring synths, modelled mainly on the old style analog modular types, is a fascinating one. Personally, I get a lot of joy and satisfaction from listening attentively to such minimal sound because it teaches me to pay attention to the finer details and puts me in a state of slow processing. It quite simply chills me out. 

Building the sound source is of course only part of the parcel here because once I have put together a modular synth, then I also need to work out how I feed it data and more precisely which parameters of the synth does that data modify. My approach for the moment is split into two main parts. Firstly, feeding sentiment data that is positive, negative, neutral as an overall ratio which is then used to calculate the main oscillators’ tunings. We can see this therefore as setting the tone if you like of the piece. The second part consists in feeding each of the emotional datas to specific parameters of the synth to create slight modulations of the sound over time. The data is read in a linear manner, as like the reading of the text. New data is pulled in at fixed times and the new settings are set using varying interpolative durations. This makes for constantly evolving sound scapes, longer interpolations giving rise to smoother motion in the sound.

Rhythm is an important aspect in music and especially on an emotional level. I would say that rhythm is perhaps one of the more primitive aspects of our musical history with the potential to affect our bodies in a very direct manner. Faster rhythms for example raise our heartbeat and that alone is sufficient to spark certain feelings and emotions. Rhythm then is a vital part of the equation, yet drone music has no beat per se, so why have I not chosen a mode of music that is more in line with that knowledge. Well, again there is a very good reason for me making that choice. As I mention in an earlier post, I’m looking to induce a certain state with the listener, me included of course. That state being one of active listening to the particulars. Simply put, drone music solicits attention if one wants to get anything from the experience. Either you pay attention to the sound that evolves and it takes you on a journey, one that is personal, introspective and meditative. Either you are elsewhere. Attention therefore is required and although the drone appears to manifest a rather monotone, mono-rhythmic sound on first listening, I find that, on the contrary, a vast scope of rhythmic possibilities can be created and perceived. It is however in the finer and more granular aspects of the timbre where rhythm is expressed.

I realise that I am at the beginning of a long process ahead. A process that requires active listening and spending time tweaking the system so that I can settle on something that is both aesthetically agreeable to the senses and coherent in the reasoning I’m developing conceptually. This is exactly what I am looking for though. A project that enables a personal quest for meditation on a subject. I’m happy with the result of this first piece and excited about the potential for future developments. Resonate was broadcast for the first time 19/12/2020 on the community driven independent UK radio, Threads and in partnership with SubText.

Listen to the piece at this link on Mixcloud.

https://threadsradio.com/

https://www.subtextradio.net/m-nifes

The Drone Machine is an evolving project that is equally part of satellite works such as my DroneWorks published on Bandcamp.