BAASnotes.com – Bay Area Audiophile Society

On Bandwidth (and Sampling Rate)

For our technically-minded members, I’m sure that you’re aware of Fourier Analysis and the Fast Fourier Transform (FFT) that revolutionized signal processing. For others, the core idea is that any signal can be efficiently decomposed into a number of sine waves. The more sine waves, the better the approximation.

The graphic at left shows a number of such approximations to a square wave. Engineers (even audio engineers) love square waves because of their infinitely rich harmonic structure. Listen to this sound clip (sin-sq-440-128k.mp3) to hear the A above middle C (440 hz) – first as a pure tone for 10 seconds, then 10 seconds as a square wave.

Sounds a bit like clipping, doesn’t it? It should – as the illustration below clearly shows.

Getting back to the first graphic (above left), we clearly need a lot of sine waves to approximate harmonically rich sound signatures like a square wave (or a bowed string). And those sine waves must be of increasing frequency. Thus, we need a lot of (analog) bandwidth. If we’re in the digital domain, this translates to a fast sampling frequency.

Hopefully this analysis helps to illustrate why many audiophiles desire 96+ khz  sampling (e.g., SACD) and/or ultrasonic analog system bandwidths (e.g., ribbon tweeters or supertweeters).

But the jury’s still out on whether these characteristics are needed to render a truly accurate musical event.

No comments yet.

Leave a comment

You must be logged in to post a comment.