naturerecordists
[Top] [All Lists]

Re: Sampling rate conversion

Subject: Re: Sampling rate conversion
From: Walter Knapp <>
Date: Fri, 10 Dec 2004 12:58:22 -0500
From: "ANDREW CARTER" <>

> 
> At the risk of being totally incorrect (as an ignorant newbie) - in 
> converting from 48 to 44 is there not the risk of of an interference pattern 
> (the two rates being fairly close together) - or am I wrongly understanding 
> the processes involved?
> Andrew

This particular conversion has turned up in some soundcards. There are 
lots of cheap 48khz chips around. The windows computer market is driven 
by cheap prices so it's a way of cutting corners for the manufacturer. 
It has, in that situation, resulted in a noticeable loss of quality for 
at least some of the soundcards involved. But I don't think anybody has 
reported that as a interference pattern. In probably all cases the 
conversion method avoids that.

In general, even the best methods of conversion are not perfect. It is 
better to minimize the number of times the sound is converted, and never 
is the minimum.

A great lot of this jumping on higher sampling rates is hype. I see 
folks using low level mics spending sums of money chasing a number, when 
they would improve their sound far more spending it on mics. Even in the 
high end few mics exist that are designed and optimized for the higher 
sampling rates. And as we have seen with the MKH-800, compromises had to 
be made in those.

As has been pointed out, the place where the higher rates can help a 
little is in sound processing. They help to avoid round off errors in 
the math. For most all the better sound editors the use of higher 
sampling where needed is almost automatic. You can run your entire sound 
path with 44k files, and not worry about it in that case. Certainly 48k 
is not a significant change in sample rate as far as sound quality is 
concerned.

Walt






________________________________________________________________________
________________________________________________________________________

<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU