At 07:09 PM 8/23/95 +0100, Andrew Spring wrote:
In the simplest case, where the microphone is not attached, the signal consists of long runs of '0x80's alternating with '0x7f's. Now, I have no problem transforming this into uniformly distributed RN's : just hash the buffer with MD5.
If you know that's what the signal looks like, you can improve it a lot by run-length encoding before doing the MD5, e.g. crunch the output down to a series of count1 value1 count2 value2 count3 value3 (if you stick to runs of 255, you can use 1 byte for each.) That gives you a much shorter input to the MD5, and a more realistic view of how much random data you have. (I suppose it may make it harder to do things like Fourier transforms on it...) #--- # Thanks; Bill # Bill Stewart, Freelance Information Architect, stewarts@ix.netcom.com # Phone +1-510-247-0664 Pager/Voicemail 1-408-787-1281 #---