NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/rec/audio/17207

Internet Message Format | 1992-12-22 | 3.6 KB

Path: sparky!uunet!wupost!waikato.ac.nz!canterbury.ac.nz!equinox.gen.nz!equinox!tragula!vanz
Newsgroups: rec.audio
Subject: Re:  CD SOund QUality
Message-ID: <[e-mail]>
From: [e-mail] (Martin Nieuwelaar)
Date: 23 Dec 92 09:01:05 +1200
References: <[e-mail]> <[e-mail]> <[e-mail]>
Organization: Not an organisation
Lines: 67

In article <[e-mail]> [e-mail] (carl.w.muhlhausen) writes:
>In article <[e-mail]> [e-mail] (jj, the really irritated tonight) writes:
>>In article <[e-mail]> [e-mail] (Nicholas E. Damato) writes:
>>
>>>I say lets leave it at ONE disc format for simplicity,
>>>and work on a 32 bit 96KHz sampling system.
[stuff deleted...]
>
>jj,
>
>I (and I think other rec.audio crazies) would be interested in what
>audible improvements we'd get out of more dynamic range and a
>higher sampling rate. I was under the impression that 16 bits
>gives more dynamic range than we need and a higher sampling rate
>would benefit dogs and bats but few humans.

Just joining the conversation...

Say you sample at 44 KHz.  The maximum theoretical limit of frequency
you can capture is half of this, 22 KHz.  However, at this rate there
are only two samples per cycle.  With two samples per cycle, a sine
wave will sound the same as a square wave.  Surely this is not hi-fi?
Suppose you say that you cannot hear this high a frequency.
Well, at 11 KHz, there will be 4 samples per waveform.  How close
to a pure sine wave can you get with 4 samples?  By looking at it,
not very.  I'm not sure how noticable the difference in the sound is.
I guess it wouldn't be difficult for someone with a computer with
reasonable sound capabilities to try this out.

I haven't introduced oversampling, as I'm still not too sure what it
is.  If someone could tell me if what I think is correct, is
correct I'd be pleased.  What I think happens is that you take
two of the samples, and do a straight line interpolation between
the two.  eg, with 4 times oversampling, there will be three extra
values created between the originals.  If this is so, then I
would expect the maximum sampling frequency would remain the same,
which makes sense.

One result of this (assuming I'm correct) would then be that
an analog 22 KHz sine wave, could be sampled at 44 Khz.  The result
would be a square wave.  Oversampling this with linear interpolation
would result in a triangular wave.  Interesting.

As for dynamic range, all I can say is that I'm pleased when I can
turn up the volume so the loud parts are loud, and there is no
background hiss in the quiet parts.  As I can only do this with
certain CDs, it leads me to believe that the companies that put
the music on the disks have some work to do!  I guess this could
be expected as I listen to house and dance music mainly, and
that's pretty notorious for 'Remix #9645324 made at that cool
party wot Eric had'.  I sometimes wonder about the equipment they
use...

One clear advantage in using more bits would be that making the
first digital recording would be much easier.  I use an 8-bit
sampler on my computer, and it stresses how important it is
to use as many bits as you can without clipping.  If I could
sample with 16 bits, the volume adjustment would be much easier.
The sound could then be normalised down to 8 bits.  Perhaps
more bits are only necessary in the recording industry?  I don't
know.

>Carl Muhlhausen        "Kum wit me iff you vant to liff"
>att!taz!ledzep

Anything I have written is just my opinion, or at least was
intended to be.