Just in case anyone wants the short version: it starts off sounding like a fan slowly speeding up with something stuck to one of the blades and then (not very quickly) transitions to train/car horn noises.
Of course, the exact same data could sound completely different if different choices were made as to how and what to assign the digits to.
For instance, instead of using "8 sawtooth oscillators and each one is assigned 50 notes from the 400 digits", the composer could have use 2 samples of babies crying and 3 of babies laughing, with some more digits assigned to amplitude and duration of each sample.
The result would sound completely different, and yet hold just as valid a claim to being "The Sounds of Pi".
Or maybe you could assign certain combinations of digits to samples of instruments from an orchestra, or a didgeridoo, or samples of people reading from Shakespeare or a comic, to sounds of car crashes, or directly to frequency and amplitude variation, or any combination of the above. These assignments could be "random", or chosen by Pi itself, or through some algorithm, or by hand by the composer.
The above doesn't even touch on filtering or effects that could be applied... which themselves could be controlled and/or assigned by digits of Pi.
There are an infinite number of ways you could represent any given sequence of numbers as sound, which are really only limited by the composer's imagination.
So what does Pi really sound like? It doesn't sound like anything, or it sounds like anything, depending on what choices one makes.
I think everyone uses decimal (base 10) because it’s what we’re accustomed to. The vast majority of programs that output digits of pi do some in decimal.[a] But 10 doesn’t fit into our (well, Western) base 7 (common scale) or base 12 (chromatic scale) note system. Using pi encoded in base 7 or base 12 would make a lot more sense.
[a]: There does exist an formula to output any digit of pi (in base 16) without calculating any of the preceding digits: the Bailey-Borwein-Plouffe formula[0], but that’s not base 7 or 12. I do remember reading about someone generating the first trillion digits of pi using said formula and converting to decimal as it went along, so one could adapt that program to do base 7 (or 12) if they wanted.
Your point is well-made, but I think there are more appropriate sonifications of pi - this one [0] for instance, done in SuperCollider [1].
I think the main difference between this one and the posted one is not just in how arbitrary the "pi" utilization is, but also in how basic and derivative it is. This is something I commonly see from users of Puckette-style languages like Max or pd.
Very nice to see this expressed, thank you! This is a trap so many folks fall into when starting out with generative work IMHO. There's this kind of natural instinct to anthropomorphize the data -- where the assumption is that some platonic representation exists that can be expressed with "the right choices" or some neutral choice -- while little attention is paid to the impact of the choices themselves on the representation.
FWIW Andrea Polli nicely shines a light on this during her Sonic Antarctica album (on the track "A Model is a Cartoon") where she superimposes her own sonifications & field recordings with an interview exploring the limitations of modeling data in a scientific context... https://www.gruenrekorder.de/?page_id=342
I did an experiment years ago where I wrote a program that plays notes from a pentatonic scale, with each digit of pi choosing the note (0 being the deepest and 9 being the highest). It sounded pretty much the same as when I used random numbers instead of pi.
Well, pi is theorized to be “normal”[0]. If it is, there’s a equal distribution of digits[a] which a perfect random number generator (working digit by digit) would exhibit as well.
Key work though is theorized; it hasn’t been proven.
[a]: I.E. Each of the 10 possible digits occurs the same amount of times (10%).
8 comments
[ 3.2 ms ] story [ 31.7 ms ] threadFor instance, instead of using "8 sawtooth oscillators and each one is assigned 50 notes from the 400 digits", the composer could have use 2 samples of babies crying and 3 of babies laughing, with some more digits assigned to amplitude and duration of each sample.
The result would sound completely different, and yet hold just as valid a claim to being "The Sounds of Pi".
Or maybe you could assign certain combinations of digits to samples of instruments from an orchestra, or a didgeridoo, or samples of people reading from Shakespeare or a comic, to sounds of car crashes, or directly to frequency and amplitude variation, or any combination of the above. These assignments could be "random", or chosen by Pi itself, or through some algorithm, or by hand by the composer.
The above doesn't even touch on filtering or effects that could be applied... which themselves could be controlled and/or assigned by digits of Pi.
There are an infinite number of ways you could represent any given sequence of numbers as sound, which are really only limited by the composer's imagination.
So what does Pi really sound like? It doesn't sound like anything, or it sounds like anything, depending on what choices one makes.
[a]: There does exist an formula to output any digit of pi (in base 16) without calculating any of the preceding digits: the Bailey-Borwein-Plouffe formula[0], but that’s not base 7 or 12. I do remember reading about someone generating the first trillion digits of pi using said formula and converting to decimal as it went along, so one could adapt that program to do base 7 (or 12) if they wanted.
[0]: https://en.wikipedia.org/wiki/Bailey%E2%80%93Borwein%E2%80%9...
I think the main difference between this one and the posted one is not just in how arbitrary the "pi" utilization is, but also in how basic and derivative it is. This is something I commonly see from users of Puckette-style languages like Max or pd.
[0] https://sccode.org/1-50w [1] https://supercollider.github.io/
FWIW Andrea Polli nicely shines a light on this during her Sonic Antarctica album (on the track "A Model is a Cartoon") where she superimposes her own sonifications & field recordings with an interview exploring the limitations of modeling data in a scientific context... https://www.gruenrekorder.de/?page_id=342
Key work though is theorized; it hasn’t been proven.
[a]: I.E. Each of the 10 possible digits occurs the same amount of times (10%).
[0]: https://en.wikipedia.org/wiki/Normal_number