naturerecordists
[Top] [All Lists]

Re: Sample Rate Conversion

Subject: Re: Sample Rate Conversion
From: Walter Knapp <>
Date: Wed, 12 May 2004 14:03:37 -0400
From: "Michael McInnis" <>

> 
> Yeah...what was I thinking?
> 
> My software, the little known Saw Studio, does realtime sample rate
> conversion, so I can use this method for comparison purposes on the
> different settings within the one program...unfortunately, this doesn't work
> for the task at hand.

Spark XL resamples on the fly when playing a soundfile, converting and 
sending out at the rate setting we have set for the output. It in fact 
flashes a warning when it's resampling. So, for instance I can listen to 
the original 48khz file and the 44khz derivatives back and forth. And 
watch the sonogram too. Spark XL can run virtually all it's filtering 
processes, even combinations in realtime. Then when we are happy with 
the results it runs more elaborate math to produce the final file. 
Sometimes that final file sounds slightly better than expected, most of 
the time it sounds the same as the realtime listening.

> What one could find out is how damaging the round trip would be...48k to 44k
> to 48k again.  It might help in evaluating where each software is "putting"
> the errors.  This might have some meaning for software only conversions.

It is one way to evaluate the conversion software. As long as we 
remember we are evaluating both directions at once. Personally I depend 
primarily on listening to the converted file. And doing sonograms in detail.

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