DJ Saint-Hubert

  Quote

I work with algorithmic generation of music and accompanying album art, and I figured out this constant

 

log(16)/log(10)*44100=53101.69123512628283570354102940056992190869508991594668715940414282

18845880084834734939672736480407479767417765684275848723170542319535

46458345727665508299441378322894439716229545001132566759233608840471

23870197366604313342044428377682526099537021443387163839212660644834

16912288684074669450634614544823410701487747461644155291035473134166

36297409637144159357340560282690647093652799638604413248007640102256

44888384442310107374338826213461196353935977511277009709552887107517

793526171854047156519408958400

 

to the first 500 decimal places.

 

it is the number that when you check if (beat_position/log(16)/log(10)*44100) > (old_beat_position/log(16)/log(10)*44100) and refresh the midi sequence by my special midi shuffling algorithm, produces a time signature that relates the sample rate to the duration in seconds to number of beats in the sequence, or a sense of 4/4 but not in 4/4 at the same time. I am also going to experiment with taking the piece of music and oversampling it to an arbitrary floating point amount equal to that constant which i think will create a sense of an idealized sampling rate for the audio. I don't know how to explain it in better terms than that, here is a sample of the musical output

 

http://picosong.com/wRcua/

 

and here is the resampled version

 

http://picosong.com/wRcJ4/

 

and just for fun here are the two tracks overlaid with each other which i think produces a scary good psuedoanalog feel

 

http://picosong.com/wRcJa

 

anyway i think this constant somehow links math and aesthetics in a meaningful way

 

oh

 

important note: the beat_position vs. old beat position is checked and conditionally refreshed every SAMPLE, not after an arbitrary amount of samples.

 

and also i would die to hear the sample rate natively done on a soundcard, but it's an odd thing to propose to a company

 

the version i posted is resampled to the nearest integer amount, this is important because I eventually want to resample to a floating point amount which apparently is possible to do on a computer/is computable