Re: Digital vs. Analog

[P3D Eric Goldstein <egoldste@xxxxxx>]

Peter Davis notes in response to John R:

> My understanding is that higher frequencies, although inaudible in the
> "real" world, will show up as aliases unless filtered.  Ideally, this
> would be accomplished with a brick wall filter, which completely
> removes everything above 22.05 kHz (since, by Nyquist principles, the
> sampling rate must be double the highest sampled frequency), but
> leaves intact everything below.  In practice, however, there are no
> such ideal filters.  They all have a sloping curve which attenuates
> some audible signal, and leaves in some of the "inaudible" frequencies
> present as aliases.

We are wayyyy off topic here, so as briefly as possible...

Peter correctly points out just one of the many system distortions which
digital sampling introduces, and there are others. *All* recording
processes introduce anomolies and distortions. The great debate between
those who prefer vinyl versus those who prefer CDs will continue to
rage, and each side will have valid points to raise in support of their
positions.

To say however as John R did that:

> As a general rule, for example, people can not hear the difference
> between an analog sound and its sampled sound (sampled at 44.1 kHz) if > the D/A converter is good (i.e. accurate with smoothing filters,
> though the smoothing filters should not actually be necessary, either, > because the harmonics are all beyond human hearing.)  Anyhow, there
> are no 'gaps' between the samples in a reconstructed sound.  The
> difference in noticed in sound (the "warmth" of vinyl, for example) is > actually due to artifacts introduced along the path of recording,
> manufacturing, and playback, and can be convincingly simulated with
> DSP techniques.

is to misrepresent the facts. We could if you'd like go back to the 60's
and look at AR (and other) research which demonstrates that in some real
world listening environments, the sample (a/k/a "people") generally
could not detect the difference between live and recorded (analogue)
sound if the signal path is "good." Does this mean that the universe of
people cannot discern this difference in most real world circumstances?
Of course not!

The same is true for A/D processing... without getting into a discussion
of which is "better," suffice it to say that "people" can a discern
difference between analogue, sampled and live sounds. Further, by saying
that vinyl anomolies can be replicated through DSP, your implication is
that DSP is without anomolies, and nothing could be further from the
truth.


Eric G.


------------------------------