Vinyl/CD

[Karl Helmer Torvmark]

On Thu, 27 May 1999, Bill Layer wrote:
> 
> It's primarily due to the limited bandwidth and dynamic range of the CD
> media; 22KHz isn't enough bandwidth to quantify enough ambient information.
> The limited dynamic range of the CD (98dB vs 130dB+ for vinyl) forces
> compression, which obscures the  micro-dynamics and other details that give
> a recording fullness.

Sorry, but this is wildly optimistic. 130 dB is a theoretical figure
only, and is not meaningful in any practical context. The text-books I've
read on the subject quantifies the dynamic range of vinyl to be around
80-90dB in practice. The dynamic range of CD isn't 96dB (that's the 
theoretical limit), but also lies between 80-90dB. A dynamic range of
130dB is not achivable in practice, for either analog og digital media.
(Remember that you have to look at your stereo system as a whole, as long
as you don't  pipe the digital data from the CD directly into your
brain.....)

>22k bandwidth is also a bit of a myth. Listening test have not
shown that the ear has ANY ability to perceive frequencies over 18-19kHz
under the best conditions.
Rather, the probable problem is that steep analog anti-aliasing filters
result in phase-distortion. By increasing the sample rate, you can use
less steep filters, and get a better phase-response. Another solution is
the oversampling used in most equipment today.

Take a look in literature to see what the limitations of the human sensory
apparatus are. 
 
> Another big problem with all contemporary digital media: as the signal gets
> quieter, you begin to lose resolution... At very low levels, you may be
> down to only a couple of bits of resolution. Now, dynamic changes really
> get lost, as the machine tries to decide if that sound is a "level 1" or a
> "level 2" and there is no in between. Even at the extremes of it's dynamic
> range, the analog system still has theoretical infinite resolution of small
> details (ie have you ever noticed that the infinity of numbers between 0
> and 1 is as big as the infinity between 1 and 10? Paul Davies would
> disagree here)

This is a more probable cause of the preceived differences. However, in a
practical situation, you will have noise. This will result in limited
resolution in the analog case, as well. Did you know that the biggest
problem in 24-bit DAC/ADCs is resolution in the analog sections? No
"24-bit" systems are true 24-bit for this reason.

Also, the range of human hearing is about 130dB. However, this is from the
pain threshold down to the lowest sound you can hear. For musical
purposes, you don't want (or need) the entire range, and a dynamic range
of around 100-110 dB is more than sufficient.

IMHO, the biggest problem with digital signal processing, is when you try
to realise non-linear systems. Then you really have problems with
aliasing, and have to adopt a sampling rate of over 1Msamples/sec to get
an "analog" response.

> CD is and always was a huge sonic compromise; it is a convenience medium
> and should never be regarded as anything else. I owned a variety of
> recordings in both formats, and the CD version was always a pathetic joke
> compared to the vinyl. Even if they are listenable (many are not) they are
> _not_ correct sounding.

It is perhaps a compromise, but a good one, especially when you think it
was designed over 20 years ago! It's not perfect, but neither is vinyl.
Ever thought about the unlinearities involved?

WOW, take it easy here, people. This is a religious issue. Perhaps your
points apply to CD-pressings of old vinyl recordings, but I refuse to
believe that you are correct when it comes to modern recordings. Remember
that the signal has already passed through a digital recorder, and
perhaps a digital mixing console....

Instead of dissing the media, take a look at loudspeakers. The SUCK in
comparison with all other parts of the audio chain.

Sadly, the audio debate is down to firmly entrenched fronts, and very
little constructive comments are done. (Same as the tube vs. solid-state
battle). Please don't post more of these one-sided comments. This question
is much more complex than most people realize. Read up on some
psycho-acoustics, it is very interesting reading and should shed new light
on a lot of these issues.

-------------
Karl H.