[ProAudio] The High-Resolution Challenge

[Mike]

On 2/14/2020 Dan Lavry<dan at lavryengineering.com> wrote:

> The focus on sampling rate is somewhat backwards. The question is what is the needed bandwidth. In theory, and not too far from practice, all one needs is to sample at twice the bandwidth.? And the real question is the pluses and minuses of including high frequency energy that is above audibility. . . . . . .  Make sure that you include what you hear, and don't include anything at higher frequencies.

Several years ago, there was a company that was making large electronic 
organ systems based on samples. The cheap version used 48 kHz samples, 
but if you wanted the really accurate thing, they had a set of 96 kHz 
samples, and that's what they were pushing. At the time, they needed a 
rack of 6 or 8 PCs in order to get the polyphony at the 2x sample rate. 
They way they justified it was that they were sampling individual pipes, 
not individual notes or stops (combinations of pipes). There was some 
 >20 kHz energy coming from the pipes, and when playing multiple pipes, 
those ultrasonic frequencies combined to produce frequencies in the 
audible range. So it's a way of getting what you'd hear if you played 
two pipes at once.

The company didn't last very long, at least not in that business. But I 
accepted it as a fair justification for the 2x sample rate, given their 
goal.

-- 
For a good time call http://mikeriversaudio.wordpress.com