Rhythm Analysis Series
– 3 approaches to rhythm visualization
– ArtScience Interfaculty
– 2020-2022
During this project, I was engaged in a sustained investigation of everyday rhythms, such as passing cars, arm swings, footsteps, rain drops, charcoal cracking or metal clicking as it expands.
My goal was to reveal hidden patterns within these phenomena and I developed many strategies and software tools to do this.
On this page, I present the 3 most successful approaches, and the rhythmic phenomena which they were developed for.
– 3 approaches to rhythm visualization
– ArtScience Interfaculty
– 2020-2022
During this project, I was engaged in a sustained investigation of everyday rhythms, such as passing cars, arm swings, footsteps, rain drops, charcoal cracking or metal clicking as it expands.
My goal was to reveal hidden patterns within these phenomena and I developed many strategies and software tools to do this.
On this page, I present the 3 most successful approaches, and the rhythmic phenomena which they were developed for.
1
2
3
The title comes from Henri Lefebvre's book 'Rhythmanalysis: Space, Time and Everyday Life', which provides a beautiful cross-section of different approaches to rhythms and everyday life.
Rhythm Analysis 1: Arm Swing
Fig. 1This software extracts movement patterns from a video and visualizes them as a spectrogram. I will use the following arm swinging experiment as an illustration:
In the experiment, I filmed myself swinging one arm for 20 minutes, trying to maintain the same speed (fig. 1). From this video, I extracted the amount of movement in each frame, giving me a graph showing the speed of my arm over time (fig. 2). I broke the graph down into its individual frequency components using an FFT algorithm, which produced the final frequency/time graph (fig. 3). In this image, you can see that my arm swung at a fundamental frequency of around 1.4hz, which decreased slowely to rest at around 1.2hz in the second half.
Fig. 2
Fig. 3
Rhythm Analysis 2: Spiral
Spiral was developed specifically for visualizing rhythms in music. The audio waveform is twisted into a spiral, allowing the user to tune in to different frequencies by adjusting the circumference and number of twists.
The idea came from the paper "An Auditory Illusion of Infinite Tempo Change Based on Multiple Temporal Levels" (Madison, 2009), where the author used spiral diagrams to visualize internally repeating rhythms.
The experience of using Spiral is similar to that of beatmatching on a turntable. Rather than defining a specific tempo, the DJ gradually tunes in and adjusts to what they hear. This makes Spiral perfect for analysing polyrhythmic music or sound where the rhythms are implied rather than explicit.
![]()
Fig. 5
Guy Madison, 'Implementation of the multiple temporal levels in a 96-event example pattern.'
The idea came from the paper "An Auditory Illusion of Infinite Tempo Change Based on Multiple Temporal Levels" (Madison, 2009), where the author used spiral diagrams to visualize internally repeating rhythms.
The experience of using Spiral is similar to that of beatmatching on a turntable. Rather than defining a specific tempo, the DJ gradually tunes in and adjusts to what they hear. This makes Spiral perfect for analysing polyrhythmic music or sound where the rhythms are implied rather than explicit.
Fig. 5
Guy Madison, 'Implementation of the multiple temporal levels in a 96-event example pattern.'
Fig. 4
Hand rubbing + metronome
4 mins, 90 bpm
In this spiral diagram, you can see the human error as the hands go in and out of sync with the metronome, as well as other sounds (cars, people) from the urban environment.
Fig. 6
Daft Punk - Rock n Roll (1997)
Rhythm Analysis 3: Firewood
In this project, I was looking for a way to visualize sequences of rhythmic events, rather than continuous movements. The software takes a sequence of durations as input and then displays them as a graph, separated into individual frequencies. It works best between 0.5-10hz.
To test it, I lit a fire in my parents' wood burning stove and used the software to visualize its crackling/sputtering patterns. The output was mostly noise, with faint shapes and traces in parts of the image, representing some kind of rhythmic regularity.
To me, these unexplained traces invite speculation: maybe the logs where chopped in a rhythmically regular way? Did they grow from a regular tree? Was this exact pile of logs synchrony with itself? Or, maybe the regularity was purely chance, and can never be reproduced?
To test it, I lit a fire in my parents' wood burning stove and used the software to visualize its crackling/sputtering patterns. The output was mostly noise, with faint shapes and traces in parts of the image, representing some kind of rhythmic regularity.
To me, these unexplained traces invite speculation: maybe the logs where chopped in a rhythmically regular way? Did they grow from a regular tree? Was this exact pile of logs synchrony with itself? Or, maybe the regularity was purely chance, and can never be reproduced?
Fig. 8My parents' wood burning stove.
Fig. 7
Visualization of fire crackling
Audio recording of fire crackling + metal creaking (used to generate Fig. 7).