vichug

http://apod.nasa.gov/htmltest/rjn_dig.html

 there are calculated some first million digits of common irrational numbers. would it be possible, and/or useful, to try and generate some sort of noise table or pseudo random number generator in Max from this ? how complicated wou;d it be to write somethign in c that would maybe translate those digits into segments of 32 or 64 bit floating points, would it be worth it ? (worth it, in the sense of having another possible pseudo random numbers suite) Was wondering...

On this page http://apod.nasa.gov/htmltest/rjn_dig.html there are calculated some first million digits of common irrational numbers. would it be possible, and/or useful, to try and generate some sort of noise table or pseudo random number generator in Max from this ? how complicated wou;d it be to write somethign in c that would maybe translate those digits into segments of 32 or 64 bit floating points, would it be worth it ? (worth it, in the sense of having another possible pseudo random numbers suite) Was wondering...


pseudo-random-number-generator-from-those-digits

Possible, sure. Useful? If what you want is a RNG that is both efficient and has good statistical properties, not very. For MSP, I wouldn't even take bets if calculating 

 can keep up with real time. Perhaps its possible with clever optimizations, but you'll see a CPU hit that's in a completely different ballpark than any "normal" noise generator (and I've written several dozen, including a few CPU-hungry critters).

You might want to consider that a million digits buys you just over 2 sec of audio signal (44kHz, single-channel, @ 32 bits/sample).

Possible, sure. Useful? If what you want is a RNG that is both efficient and has good statistical properties, not very. For MSP, I wouldn't even take bets if calculating e can keep up with real time. Perhaps its possible with clever optimizations, but you'll see a CPU hit that's in a completely different ballpark than any "normal" noise generator (and I've written several dozen, including a few CPU-hungry critters). 

You might want to consider that a million digits buys you just over 2 sec of audio signal (44kHz, single-channel, @ 32 bits/sample).


pseudo random number generator from those digits --