Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!elroy.jpl.nasa.gov!grian!heim!scott
From: scott@heim.glendale.ca.us
Newsgroups: comp.music
Subject: Re: Sample pitch shifting reference
Summary: What you mean is undersampling - low pass and sample
Message-ID: <1990Mar16.053022.5335@heim.glendale.ca.us>
Date: 16 Mar 90 05:30:22 GMT
References: <8608@chaph.usc.edu>
Reply-To: scott@heim.glendale.ca.us
Organization: Watson Residence, Glendale CA.
Lines: 26

In article <8608@chaph.usc.edu> alves@nunki.usc.edu (William Alves) writes:
>In article <2441@rodan.acs.syr.edu> jmwobus@rodan.acs.syr.edu (John Wobus) writ\
>es:
>>I am having fun writing a C program that manipulates files of sampled
>>audio, but have been mulling over the best way to take data sampled
>>at a high rate (say 55000 samples per second) and producing data
>>sampled at a lower rate (say 11000 samples per second) with a minimum
>>of distortion.  Can anyone point me to any simple and/or efficient
>>algorithms or strategies to do this?
>>
>The intuitive response would be to take every fifth sample. However, doing
>that without some kind of filtering leaves you open for all kinds of aliasing
>problems.

Thats right - that's why the the device which did the original sampling probably
had a lowpass "anti-aliasing" filter infront of the A/D.  You can simulate this
same procedure in software (for an 11K sample rate, you probably want to 
pass 5K and below), however I don't have a lowpass code fragment for you.

Maybe some kind comp.music reader can help out - otherwise you might want to
address this to comp.dsp.

-- 
Scott Watson - "Inane little message goes here" 
    uucp: {rutgers,ames}!elroy!grian!heim!scott
Internet: scott@heim.glendale.ca.us