Showing posts with label music about data. Show all posts
Showing posts with label music about data. Show all posts

Tuesday, March 14, 2017

Pi in the sky


Today we celebrate π day, because (non-metric) Americans write the date 3/14, like the first three digits of the digital expansion of the ratio of the circumference of a circle to its diameter. Enjoy with some mathy Kate Bush and yet another incredible math-art work about pi by Martin Krzywinski. This year he's translated the 12,000,000 digits of Pi into star charts (by taking blocks of 12 digits and using them as latitude, longitude and azimuth). Then he's selected 80 constellations from these imagined stars and named them after extinct plants and animals. Find more here!

Martin Krzywinski's 2017 Pi Day Star Chart Carree Projection

 

Monday, October 13, 2014

Music about Data

Gafurius's Practica musice, 1496 showing Apollo,
the Muses, the planetary spheres and musical ratios.
Science and music, like other arts, have a longstanding, close connection. Music can be described in terms of physics; notes translate to waveforms at a certain frequency, or equivalently certain pitch. Acoustics, tempo, rhythm, tones and overtones, harmonies and more can be explained in terms of physics. We can likewise discuss our physical world in terms of music.


In ancient Greece, Pythagoras and his followers placed a mystical meaning on his discovery of the mathematical underpinnings of music; he found that the length of a plucked string determined its pitch and that   simple (rational) ratios of a given length produced harmonies. They turned this idea on its head and apparently concluded that other fundamental patterns in nature were due not so much to mathematics, but that there was a musical underpinning to the known universe. Hence, the idea of the 'music of the spheres' and the hypothesis that planetary motions obeyed mathematical equations corresponding to musical notes and that the whole solar system together played its own symphony.





Kepler's musical notation for planetary motion and the range of sound
he ascribed to Saturn, Jupiter, Mars, Earth, Venus and Mercury
The idea was so persistent that when Johannes Kepler (1571- 1630) was developing the best model of our solar system to fit the beautiful dataset gathered by his mentor Tycho Brahe (1546-1601), one of the first notations he used was not mathematical, but musical. In fact, the idea was pervalent, and Kepler ended up embroiled in a priority dispute with Robert Fludd (1574-1637), whose own harmonic theory had been recently published in De Musica Mundana. While we tend to think of Kepler with his rational, more precise elliptical version of a Copernican heliocentric solar system as one of the first, modern scientists, he progressed from his musical notation, to a model based on a rather mystical appreciation for the Platonic Solids. That is, rather than explaning planetary motion in terms of his laws, as we know then today, he tried to make a model spacing of the planets from the sun based on the relative size of a nested spheres just large enough to coat a  series of special shapes called the Platonic Solids: the tetrahedron, the cube, the octahedron, the dodecahedron and icosahedron. He progressed from there, in his Harmonices Mundi (literally, harmonies of the worlds) to describe planetary motions in musical terms. He found that the difference between the maximum and minimum angular speeds of a planet in its orbit was very close to a harmonic proportion. For instance Earth's maximal angular speed relative to the sun varies by about a semitone (a ratio of 16:15), from mi to fa, between aphelion (the furthest point from the sun on its elliptical orbit) and perihelion (its closest point to the sun). In his words, "The Earth sings Mi, Fa, Mi", and he built up a choir of similarly singing planets. He found that all but one of the ratios of the maximum and minimum speeds of planets on neighboring orbits approximate musical harmonies within a margin of error of less than a diesis (a 25:24 interval) - to use a musical term.

Today we would attribute these patterns to the underlying mathematics of planetary motion, or the physics of music, rather than a music of the spheres underlying everything. Nonetheless this trick of Kepler's, of mapping observed patterns onto music, or of writing data as music still has its place. I recall a professor extolling the virtues of plotting data as it was collected, because we are wired to see patterns and would for instance, recognize a friend's face in a crowd with much greater ease than their phone number from a list of 7-digit numbers. The same can be said of sound; we are wired to recognize musical patterns. We can both appreciate the beauty of regular data mapped onto sounds we can hear, or use what we hear to recognize patterns.

Galileo Galilei (1564-1642) was the son of a famous lutenist, composer, and music theorist, which may have primed him to be observant of the measure of time, rhythm and periodic patterns. In Galileo's Daughter, author Dava Sobel argues that in the absence of accurate time pieces, music likely played an important role in his experiments. Many experiments involved timing repeated observations as precisely as possible and it is likely that he may have used song as his yardstick of time.

A couple of contemporary examples of expressing experimental data musically have been in the news of late.




The European CERN particle physics lab in Switzerland celebrated its 60th birthday with this delightful composition by physicist and musician Domenico Vicinanza, which turns data from four detectors at the Large Hadron Collider into LHChamber Music. Performed by CERN scientists and engineers, the result is surprisingly musical, like Baroque chamber music. Vicinanza has 'sonified' data before (including the satelitte Voyager I's magnetometer data), employing an algorithm to assign a musical note to each measurement created by experiments, so that the same data is presented as a musical score, much like Kepler did.



Sonifying data also allows scientists to hear patterns, to cope with massive datasets and find complexity which may otherwise have escaped them. Above, cicada calls are replaced with notes. The University of Uppsala team explains their sonification and visualization of the data:
The circles represent recording stations in the Australian bush that pick up the calls of cicadas. The intensity of the circle’s colour and its size is proportional to volume of sound in that area of the forest at that time (the videos is 15 x real time).
They could also add the sound of the cicadas themselves (speed up 15 times), but in the words of researcher James Herbert-Read, "that would be horrific". Instead they decided to translate cicada calls into music.
Each one of the four different coloured block of recorders also plays a different chord (we chose the standard I–V–vi–IV progression in the key of C major). By doing this, you can now not only see, but hear when cicadas in different areas of the forest start to sing, when other cease singing, and listen to the additive effect of all individuals singing together across large swathes of the forest.
The video is the cicada 'morning chorus' beginning at 5:30 am when light strikes the right hand side of the area shown, where the  first cicadas call. You see and hear other cicadas join, the early oscillations in volume and then the crescendo to full volume for the remainder of the chorus.

Locals had noted waves of cicada song moving through the forest and the researchers wondered whether they could prove the cicadas were in fact synchronized. They found quantifiable waves did in fact move through the forest. Though, they theorize that this is an emergent pattern, where each cicada follows his own rules and does not consciously try to synchronize with his neighbours.

Thursday, October 24, 2013

Smell, Taste and Hear the Data

Some of our contemporary computational and graphics tools make data visualization an exciting, rapidly advancing field. Sometimes scientists presenting data are not visual thinkers and may not be communicating their results transparently. Sometimes graphic designers put aesthetics ahead of the essence of what we could be communicating with a given dataset. But, when we get it right there is some real innovation occurring. Today I want to show you some 'visualizations' which are not in fact visual at all; I want to show some playful and intriguing forays into expressing data for senses other than vision.



Listen to wikipedia is a wonderful multimedia audiovisual exploration of an unexpected dataset: the recent changes to wikipedia feed. It is in fact surprisingly musical. Data are shown as circles in white (for edits by registered users), green (for unregistered users) and purple (for automated bots) on the Payne's grey field, like in the screenshot above. Text labels appear briefly to show which articles are being edited, the frequency of edits (84 edits per minute at the bottom left) and to post notices when new users register. Data make different sounds: bells indicate additions and subtractions sound like plucked strings. Pitch is proportional to the size of the edit; larger edits result in lower notes. While visualizations can communicate a lot of specific information compactly, this audio display of information certainly communicates a great deal while still allowing an observer to do something else; I am listening to it as I write. It is quite easy to pick up on the frequency of edits, their nature (additions or subtractions) and with a little more attention to the pitch and hence size of edits. Listen to Wikipedia was written by Stephen LaPorte and Mahmoud Hashemi, and is open-source. It was inspired by Listen to Bitcoin which plays tones with pitch scaled to the value of Bitcoin transactions.

BevLab image via Spark blog
You might recall magpie&whiskeyjack wrote about Kate McLean's SensoryMaps which plot scents on imaginative maps of places like 'Auld Reekie', Edinburgh itself. But smell and taste themselves can be used to express datasets. The CBC radio show Spark episode 227 profiled i & j ideations and their BevLab project which translates data into different flavoured beverages and invited the public to 'taste the data' of real-time tweets about food. The words they recorded were mapped onto flavour profiles which controlled proportions of ingredients (for instance blueberry juice for sweet, lemon juice for sour, and ginger juice for spicy, standing in for salty) in a beverage. A beverage can be produced at any given moment which is thus intended to represent the food words posted on twitter at that given moment. I am not certain of our ability to identify this three dimensional sweet/sour/spicy beverage and comprehend how the data is changing in time, by drinking a series of beverages, but the idea is playful and intriguing.



Artists are also playing with the use of data in their art - and these projects too sometimes involve more than just sight. Artist Charlotte Jarvis collaborated with the Netherlands Proteomics Centre on a fascinating project called 'Blighted By Kenning', described in the video below. The Universal Declaration of Human Rights was encoded* into DNA of bio-engineered bacteria which was sprayed onto the surface of apples grown near the International Court of Justice at The Hague. The concept was that The Hague was 'contaminated' with the message of the Universal Declaration of Human Rights on these contemporary 'forbidden fruit'.  Then they sent the apples were sent to Genomics laboratories across the world, which were asked to sequence the DNA and to find the message hidden within. Finally, the scientists who sought the hidden message were also invited to eat the fruit. So this is a very different way one can taste data.



*There is an established means of encoding letters in DNA by mapping each letter onto a codon, a tri-nucleotide unit consisting of a specific combination of Adenine (A), Thymine (T), Guanine (G) and Cytosine (C). Read more on Charlotte Jarvis' Blighted by Kenning site.

MODEL OF THE UNIVERSAL DECLARATION OF HUMAN RIGHTS EXPRESSED AS A PROTEIN

Blighted by Kenning installation (Photo by James Read)
 

LinkWithin

Related Posts with Thumbnails