[linux-audio-user] Re: Digital Fidelity

Carlo Capocasa capocasa at gmx.net
Mon Feb 27 15:40:08 EST 2006


I Maluvia!

I'm sorry I didn't mean to offend you. It's just that *I* happen to be
heavily into *certain types* of hocus pocus, and I was hoping to
exchange a little bit of subtle humor with you, since you made the
impression you'd be at least halfway open to that type of thinking,
which is rare!

As for the subjectivity, the pages you quoted about the digital myth are
based on three things, which are all scientific: 1. CD jitter errors, 2.
different types of A/D converters, and 3. recalculation loss.

None of these would explain properly why you can hear differences
between different copies of files on HARD DRIVES.

A/D converters, sure. The sound they are picking up are in an area where
quantum physics starts playing a role, so of course there are all kinds
of differences to what exactly happens to the sound there.

Jitter errors are measurable. You create a mint 44.1/16 WAV file that is
your master copy. You slap it onto a CD with your CD writer, that
introduces errors because our consumer optical media is fault-prone. You
rip the CD on your friend's computer, reading gives more errors, but
none are reported, they are simply padded through your CD peripheral's
error 'correction' systems. The sound quality is degraded. So far so good.

Now say you transfer the whole file back from your friend's computer
over a USB hard drive. Now hard drives are fault proof while they last,
and when there are errors they are reported immediately. Scan your hard
drive with some kind of tool (Norton Utilities used to be fashionable at
a time when this was more of an issue than it is now), and you see where
there are 'bad' sectors. When you do have bad sectors on your hard
drive, bad things happen. Programs crash constantly. File opens fail.
You lose data. Your songs start sounding not worse quality, but radio
static. You notice it. It's measurable. (In fact, hard drives are not
completely foolproof, but that is backed by redundantly storing the data
on the disc at the in-disc electronics level, so the computer kernel
doesn't even notice... All it gets is a certain series of numbers in,
and the same series out. The same things happen when you store software
on CD-ROMS, or WAV files on CD-ROMs. There is no error 'correction',
errors get reported and if one bit is wrong the disc is trash, immediately).

So you take your faulty audio file from your friend's computer, and put
it back on your computer using a WAV file, checksums, and generally
reliable storage. Your friend's file is inferior, but that is
NUMERICALLY MEASURABLE. You run a checking algorithm over the files, and
by golly the files are numerically no longer the same, simply because
the old CD-AUDIO format is an unreliable way of transmitting digital
information, and errors are not reported but attempted to be hidden.

Expensive audio players simply happen to be a little better at this
hiding process, but transfer your data as WAV files on a USB stick and
you don't even need that because you're not trying to fix a two dollar
radio with a two thousand dollar solderer.

As for all the recalculations and algorithms that get introduced into
our audio without us knowing it, yes, this is very common. Everyone
wants to throw in a magic formula algorithm to 'improve' what isn't
broken, and consumer software often does it without us being aware of
it, blatantly ignoring psychoacoustic algorithms don't improve the basic
property of digital fidelity, and that is data resolution.

But, also, if you'd run an algorithm on an effect-distorted file, it
would be numerically detectible. Perfect science. You make the numbers
worse, the sound gets worse.

So far so good. All perfectly scientific.

Now here's where it get's interesting (To me, at least :)

You claimed you can hear a difference between two files that are stored
on different hard drives. Now that's a completely different story. Hard
drives usually store software. There is no automatic error correction
with a subtle loss of quality for software. You change a single bit in
an executable program, and the thing crashes. It could break your
computer, physically, and ruin all your data which represents years of
work. In many cases, a faulty program can KILL PEOPLE. Software depends
on 1:1 numerical reproduction of data, at least for the lifetime of a
computer, and when that's not given, there is no longer a computer,
there is only a piece of scrap junk.

And yet you claim you can hear the difference between different copies
of data on different hard drives. To someone who has been working on
computers for more than half of his lifetime, that statement in itself
sounds esoteric. So don't be surprised if the man who believes you uses
esoteric theories to explain it :)

Of course, the phenomenon of subjectivity and energy relations is
currently explained within contemporary science as simply a convenient
way of discribing trances and other subjective psychological states,
which are diverted and deluded and somehow diverted from a generally
accepted way of experiencing things.

To me, this is a very dogmatic, and also quite arrogant way of looking
at things. I don't deny that thinking scientifically is USEFUL, because
it enables us to create very sophisticated models of how the world
works, which in turn lets us create neat things such as digital audio
devices and airplanes and cars. I wouldn't want to make do without all
that. But in all its usefulness, that doesn't make the theory TRUE.

Newton's theory is useful. It helps explain things WITHIN CERTAIN
BOUNDARIES. But then Einstein came along and created a different theory,
which had the same results within Newton's range, but continued to be
effective when Newton's theory failed. Then Einstein's theory failed,
and quantum mechanics was introduced to better explain other and
different phenomena. Great efforts are on the way right now simply to
bring all those observations to common ground, and when they succeed, we
can be certain that YET MORE unexplainable phenomena will occur, and
that the theory will have to be placed in proper context yet again. The
theory is never true, it is only useful within certain contexts. It's
'true' because it's useful. But it can just as well be untrue where it
stops being useful. That's simply the way science works.

However, there is a tendency among the scientific community to cling to
what already has been explained, simply because it makes things so
comfortably predictable. However, that comes at the price of gaining new
knowledge, at the price of the very way the current explanations have
been found. That's why it is so very important to keep an open mind
about things. I was ready to trash my theory that bits are always bits
that has worked so well for me for over twelve years AT THE BLINK OF AN
EYE, simply because of your unconfirmed subjective observation that you
can hear a difference between the same files on different hard drives,
even if it might mean making the way I work in the studio infinetly more
complex, at least until I found a better theory.

That's what I was trying to do there, take my present theory and expand
it with new knowledge that a couple of Hawaiian shamans tought me (and
my dad gave me a glimpse of when he told me what he and his buddies were
up to anno '69)

I have heard of people who could PHYSICALLY REPAIR CARS simply by
thinking about them. If you think that's hocus pocus, remember that the
'round earth theory' was considered hocus pocus by most only five
hundred years ago. Five hundred years! On an earth scale, that's not
even one acoustic sample. That's way below any D/A converter's noise
margin. We need to stay open about things.

Now that doesn't mean I'm getting all hoozy shmoozy on you. If people
come to me and tell me all I have to do is think about the pink flame in
me and all is well, I tend to mentally put them right where they belong
in my opinion, next to cult followers, opiate addicts and battered
housewives. What do these three have in common? Unhealthy addictions.
But, if new knowledge comes along that has the potential to genuinely
improve my life (and that of others), and that knowledge happens to
sound different, well. Let us investigate.

The statement about the cars no longer sounds so far fetched when you
study this guy's work a little:

http://www.whatthebleep.com/crystals/

He has people think about water, freezes the crystal, and documents the
results. The principles are esoteric (in that subtle energy phenomena
are used as explanation) AND scientific (in that the results are
reproducable and observable).

Wanna create miracles? Do things that are unexplained in our school system.

So in my book, no, I do not believe that you are deluding yourself. Bob
Moog is said to have had EMOTIONAL CONNECTIONS with his synthesizers,
and as far as I am concerned, they sure sound genuine. Maybe that's just
the geek me, but I've seen people have really healthy relationships with
technology and other 'inanimates'. And maybe that's just the nerd me,
but I've seen some really unhealthy ways of people interacting with each
other. So I like to joke a lot about technology and people being
interchangable relationship-wise. It's a way of getting over a lot of pain.

I do not believe you are deluding yourself. I believe you are using your
yet unexplained physical properties to influence your environment that
could be observed by other people in the same situation also, but maybe
not by someone using the same brand but different device as you, or a
different person. But that's just a theory. It could be wrong.

The remark about your respected partner in the end, on the other hand,
was deliberately formulated to be a little edgy, if not theoretically
unsound. Please forgive me, I can write impressive essays very well, but
I have a very cruel sense of humour that is constantly tempting me.

Carlo




More information about the Linux-audio-user mailing list