Wonder no more! Give this a listen:
CMS’s very own Piotr Traczyk put together this piece and the video, and he explains his motivations and methods below.
Guest post by Piotr Traczyk
Last year I was part of the team that made a video for CERN’s 60th anniversary called “LHChamber music”. It was a performance of a piece composed by Domenico Vicinanza and performed by LHC physicists and engineers in the experimental caverns. The musical piece was composed based on a “sonification” of data from the four big LHC detectors, transforming data into musical notes using some relationship chosen by the composer, representing the data as sound instead of the plots and graphs we physicists are so used to. The data from the experiments were mapped to a major scale and arranged such that they could be all played simultaneously in harmony, as a metaphor of scientific collaboration.
Shortly after finishing work on that video I had the idea of doing something similar, just as a fun experiment for myself. I wanted to take some CMS data plots and make a short musical piece that would “play” the plots, minimising the input from the composer. I decided to take two plots from the Higgs discovery seminar of 4 July 2012: the Higgs-search mass plots in the gamma-gamma and 4-lepton channels. They both have about 20 bins in the histograms, so taking the first 16 bins and assigning each bin to a sixteenth note I could have a single bar of music representing each plot. Or two bars if I use eighth notes, and so on.
The way I created the melodies is more apparent in the case of the 4-lepton plot:
The data (the points with error bars in the histogram) have low numbers of entries in the bins. Some bins are empty: an empty bin would correspond to a pause. A bin with X entries would correspond to a note X semitones above C# (I chose C# as the base note). I preferred to assign the notes of the chromatic scale (semitones) – and not, for example, a major scale – to give the data freedom to yield some crazy melodies. So the rhythm and the harmonic structure are also derived from the data. I figured that the resulting melody, if played low, could have been a guitar riff in a wacky heavy-metal song. A perfect match for the CMS guitar! So this was the guitar I used for the 4-lepton riff.
The gamma-gamma mass plot clearly has to be a fast descending run:
I again mapped the points in the histogram to a chromatic scale, but I couldn't use the numbers of entries from the histogram directly, since they go from 1500 to 500 and the range of the electric guitar is something like 50 semitones. I therefore compressed the histogram range and used the logarithm of the number of entries (the relation between note frequency and musical scale is also logarithmic) to get something reasonable. When I mapped the numbers to the notes, I chose the starting note such that the melody would fit the one from the other plot.
The resulting “song” is written for two guitars, one plays the gamma-gamma run, and then the other plays the 4-lepton riff a few times. Each guitar also plays something else when the other guitar plays the sonification; I took the liberty of writing these parts to make the whole thing blend seamlessly. I added a drum and bass part – it is a metal tune after all! The only thing that remained to be done was to film the video and CERN’s main auditorium was the obvious choice of location: this is where the original discovery was announced.
Now I’m waiting for a discovery in the next LHC run, so I can produce another track.