View Single Post
  #33   Report Post  
LASERandDVDfan
 
Posts: n/a
Default Why I don't have a plasma or LCD TV either

16bit audio was the best they could come up with in the late
70's


Irrelevant.

State a more clear and definitive answer.

(Well... they did quickly come up with 24bit but Sony who was
spearheading the digital technology had too much invested in 16bit
encoding, also they'd paid big bucks for the development of the medium
that was to usher in the digital era- CD's.)


First off, the concept of the CD was invented by Philips. Sony came in the
game later to help with its development. If Sony hadn't been in the game, the
CD would have been Philips' way, which would be 14-bit at 44 KHz, if the final
product would even work reliably enough to be practical.

Secondly, Sony wasn't the only company involved with digital audio which
involved these specs. Pioneer, Denon, and even the British Broadcasting
Company had involvement in the development of digital audio, among other
companies. As for PCM, those were devised on theories and mathematical models
established in the early 20th Century!

16bit just doesn't have enough sampling rate to encompass the mass of
data in complex musical scores. i.e. anything more complex than the
average 5 piece rock band.


And what do you define as mass in this regard?

Quantization is nothing more than measuring a wave's amplitude and assigning it
a value.

16-bit quantization allows up to 65,536 possible amplitude values for the sound
wave, as calculated with 2^16.

24-bit quantization allows up to 14,777,216 possible amplitude values for the
sound wave, as calculated with 2^24.

Obviously, the higher the quantization, the more values of amplitude an encoder
can assign to a wave in creating a mathematical model. This is how
quantization has a say in determining the dynamic range and signal-noise ratio
in a digital audio system.

In theory, 24-bit should sound better but, in practical use, 24-bit is hardly a
quantum leap over 16-bit.

First off, regardless to what you believe, the dynamics that 16-bit
quantization allows is still superior to any analogue system that is out there.
Analogue systems that achieve high enough levels to be comparable in regards
to dynamic range and signal-noise do so by using companders, which has their
own set of problems when they do what they do. Meanwhile, 16-bit can achieve
anywhere from 96-98 dB of dynamics with approximately equal SNR WITHOUT the aid
of companders, and their inherent audible artifacts.

However, you still have to factor another important aspect of digital audio
encoding, which is sampling. This determines your maximum frequency response.

Digital audio for CD is sampled at a rate of 44.1 KHz, which gives it a 22,050
Hz maximum frequency response. The extra bandwidth above 20,000 Hz is to
accomodate filtering with brickwalls, although digital oversampling filters do
a much better job at filtering by upsampling the 44.1 KHz feed upwards to 352.8
KHz with the typical 8-times oversampling digital filter.

Both quantization and sampling work in concert for digitization of audio.

As for analogue recorders, they still have finite abilities in recording a
sound wave since they are limited by their own capabilities.

Plus, like digital audio, analogue audio recorders also do a conversion of the
soundwave into another form suitable for storage on the medium they deal with.

Digital audio has a grave misconception of "butchering the sound" in that it
doesn't capture everything. In reality, digital is nothing more than another
way of dealing with storage and reproduction of information.

In the same way, you could say that a lathe butchers the sound by turning a
sound wave into a mechanical interpretation cut on an LP mother using a source
that is rolled off at around 15-16 KHz to keep the cutterhead from overheating.
With the severe roll-off, LPs capture less than CDs, even with the use of dbx!

Analogue tapes may butcher the sound because it's converting a sound wave into
magnetic print patterns. It may be further butchered if you use companders in
an attempt to eliminate the inherent noise in analogue linear magnetic
recordings. And, again, analogue tapes using fixed heads can't capture
frequencies above a certain range unless you want to run the tape at a
ridiculously high linear speed, which can allow greater levels of wow and
flutter.

Hell, you can even say that microphones will butcher the sound because it's
turning a sound wave into electrical impulses of varying amplitude and
frequency of a finite nature. No microphone will ever be able to capture sound
at the infinite frequency ranges with infinite possibilities of amplitude. The
limits you place on sound start right when you pick it up from the mics.

It even starts when you listen to sounds live with your own ears, as the range
of human hearing is also finite in terms of perceiving frequencies and
amplitudes.

Comb filters were used in the circuitry to "filter" out extraneous
artifacts i.e. noise.


These aren't called comb filters, but brickwall analogue filters.

Problem was when faced with more data than could be processed it used
the comb filter to restrict the flow of data to prevent overrun or
distortion.


Bull****.

The rate of process by the digital section, and therefore pitch, is governed by
a quartz oscillator. The information is transferred from the serial device
(the disc drive) to the digital section and is reassembled from its interleaved
state, which are arranged as block sectors on the disc, into a linear form.
The read is buffered, undergoes 8-14 modulation, is run through CIRC to correct
any errors, oversampled (which is done on just about all but the oldest CD
players), and then converted by the D/A.

While analog have alot of failings fidelity wise, analog system don't
filter anything out and you get everything.


Bull****.

In analogue tapes, companders is, guess what, a form of filtering! The very
things that are used in hi-fi VCRs, for which you give praise.

Companders designed to minimize the inherent noise of tape hiss as well as
increasing the audio's signal-noise is a form of filtering. If you play a tape
that was treated to noise reduction without the NR process engaged on playback,
you'll get some crappy sound, all the way from too bright a sound with Dolby B
tapes to high levels of intolerable intrinsic noise with dbx.

As for LPs with dbx, same problem.

I had a pro sound studio where we did alot of post production work for
film in the late 70's and early 80's.
I was anxious to go digital and got the then available digital studio
units available to test them out.


That was when they first came out. Have you ever tried the stuff that is the
current state-of-the-art now? If you haven't, then, I'm afraid, you are no
longer an authority.

But when we listened with our ears we noticed that the bass was
thready and we lost almost all ambients during busy passages.


Was the encoding equipment dithered?

I'll bet it wasn't.

That's what can happen if you don't dither the encoding equipment. - Reinhart