[Table of Contents]


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ARSCLIST] Sampling Theory (was Fred Layn's post on the Studer list re: Quantegy)



According to master audio designer Rupert Neve though we do not "hear"
the frequencies above
20KHz we can tell when they are missing! This is one of the reasons why
analog purists shunned
the earlier versions of the ADC process, as it simply filtered out
entirely what is essentially the harmonic
overtone series above the stated upper threshold of the original
Nyquist curve (44.1KHz). Now that the sampling frequency
in the new Pro-Tools HD system has an upper threshold of 192KHz,
meaning an effective reproduction
of 96KHz, these objections of reproducing the so-called "air"
frequencies has been effectively eliminated.

As an old die hard analog guy, I am sheepishly willing to admit that
the new HD system sounds downright
superb, no ifs ands or buts...It is simply a very good reproduction
system at this point marry that to the death
of magnetic oxide and mylar and you have the future of audio recording.
Now as far as the digital mixing console
or mics go, well that's another story for another day...

aaron
On Jan 16, 2005, at 7:44 AM, Don Cox wrote:

On 16/01/05, Steven C. Barr wrote:

So...the d/a converter actually creates a smooth waveform rather than
a set of steps. Is this done by averaging the values of adjacent
"steps" somehow?

Yes, by applying a steep filter above 22.05KHz. (Nowadays the steps are doubled or quadrupled in frequency first, so the filter can be more gentle.)

I had no idea this was done...would the unfiltered signal sound essentially the same, or could we hear a distinct difference?

You would hear a difference if your hearing were good above 22.05KHz. On early players, the very steep filter could have effects in the audible range.

There would probably also be audible subharmonics of the sample rate
without the filter.

See "Art of Digital Audio" (Watkinson) or "Mastering Audio" (Katz) for
full explanations.

Regards
--
Don Cox
doncox@xxxxxxxxxxxxxx



[Subject index] [Index for current month] [Table of Contents]