Fractal Beat: MIDI Sequencer
The default mode produces a fast-paced irregular but consistent (fractal) beat / note pattern, which is self-similar across different time scales (that is, the type of variability observed in a 4-second interval will be similar to the variability observed in a 40-second interval — small changes that have a few, but significant number of bigger ones). The synth can be used to create a melody or rhythm. You can leave it running on repeat or use it as a background for your 4/4 creations, to bring more fractal variability into your music and expression.
Try it on FractalBeat.Live
Types of Fractal Variability
This algorithm used in this tool is based on our open source DFA fractal signal analysis tool and fractal signal generator. Its principle of operation is based on generating a time series with variable distances between the signals. Variability of those distances can belong to four distinct states: uniform, regular, fractal, and complex.
Uniform variability is when the distance between the signals is more or less equal with small deviations and with a clear average value (normal distribution). In DFA algorithm this is the value where alpha (closely correlated with the Hurst component) is below 0.6. Practically, any 4/4 beat, techno beat, or repetitive action will have this type of variability. It is associated with regenerative states.
Regular variability is when there is more variation between the signals, but there’s still a clear average value. This is a less repetitive state but the average distance between the signals is still predictable. In DFA algorithm this is the value where alpha is between 0.6 and 0.9.
Fractal variability is when the time series is undergoing a critical shift. There is no more average value, so it’s hard to predict the next impulse, but there is interaction across different time scales. As a result, variability that is observed at very short periods (e.g. 4 seconds) is similar to variability observed at longer periods (e.g. 16 seconds and 64 seconds). As a result, the signal has highly organic nature and feel to it, the beat sounds like fire crackling, water drops, or heart beat. In DFA algorithm the value of alpha associated with this state is between 0.9 and 1.1.
Complex variability is when there’s no single pronounced state that is defined by specific parameters, so the dynamics may change completely during the observation time frame. This is associated with quick pattern shifts and prolonged repetitive signals that last only for a limited period of time. In DFA algorithm the value of alpha associated with this state is above 1.1
Fractal Beat produces variability not only in time series but also in the type of signals that it produces. As it can produce 4 different types of signals (i.e. MIDI notes), it encodes one of the 4 states of variability in the note sequence as well. Therefore, the MIDI notes are constantly changing according to the type of variability selected (uniform / regular / fractal / complex). This creates an additional layer on top of the time series dynamics that can be used to build a melody and explore variability from multiple perspectives.
