OK. I'm going to take one last stab at this. You are right. But at the
same time, you are wrong also. You are applying an out-of-date
definition to a grey subject, and trying to make it black and white
by using that definition.

The definition of what is analog data and what is digital data is fixed, and
will not change. The reason you're reluctant to accept it is that, like all
human beings -- myself included -- you're reluctant to alter the way you
think.


To use the linguistic analogy again. Thirty years ago, the word "gay"
meant happy and carefree. Then it was hijacked by homosexuals
to describe themselves in what they felt was a less contentious way.

This is not correct. The term, as a synonym for "homosexual", appears to
have been in use since at least the middle 17th century. It actually means
"loose", "depraved", "low-life", etc, and was originally derogatory (or at
least disapproving). The Oxford English Dictionary gives the following
definitions (among others) and source examples:

2. a. Addicted to social pleasures and dissipations. Often euphemistically:
Of loose or immoral life. Esp. in gay dog, a man given to revelling or
self-indulgence; gay Lothario: see Lothario.

1637 Shirley Lady of Pleasure v. K1b, Lord. You'le not be angry, Madam. Cel.
Nor rude, though gay men have a priviledge. 1700 T. Brown tr. Fresny's
Amusem. Ser. & Com. 130 Every Dunce of a Quack, is call'd a Physician+Every
Gay thing, a Chevalier. 1703 Rowe Fair Penit. v. i, Is this that Haughty,
Gallant, Gay Lothario? 1754 Adventurer No. 124 37 The old gentleman, whose
character I cannot better express than in the fashionable phrase which has
been contrived to palliate false principles and dissolute manners, had been
a gay man, and was well acquainted with the town. 1791 Burke Let. to Member
Nat. Assembly Wks. VI. 36 The brilliant part of men of wit and pleasure, or
gay, young, military sparks. 1798 Ferriar Illustr. Sterne ii. 40 The
dissolute conduct of the gay circles in France is not of modern date. 1847
H. Rogers Ess. I. v. 214 For some years he lived a cheerful, and even gay,
though never a dissipated life, in Paris. 1849 Macaulay Hist. Eng. vi. II.
103 The place was merely a gay suburb of the capital. 1851 Mayhew Lond.
Labour I. 382 The principal of the firm was what is termed ‘gay’. He was
particularly fond of attending public entertainments. He sported a little as
well, and delighted in horse-racing. 1891 E. Peacock N. Brendon I. 302 This
elder Narcissa had led a gay and wild life while beauty lasted. 1897 J.
Hutchinson Archives Surg. VIII. 224 My patient was a married man, who
admitted having been very gay in early life. 1900 G. Swift Somerley 54 Oh!
that first kiss! how proud of it we are, what gay dogs we feel! 1910 S.
Kaye-Smith Spell Land xix. 221 He felt rather a gay dog. 1849 Macaulay
Hist. Eng. ii. I. 196 On the vices of the young and gay he looked
with+aversion.

2. b. Hence, in slang use, of a woman: Leading an immoral life, living by
prostitution.

1825 C. M. Westmacott Eng. Spy II. 22 Two sisters—both gay. 1857 J. E.
Ritchie Night Side Lond. 40 The gay women, as they are termed, are worse off
than American slaves. 1868 Sund. Times 19 July 5/1 As soon as ever a woman
has ostensibly lost her reputation, we, with a grim inappositeness, call her
‘gay’. 1885 Hull & Linc. Times 26 Dec. 8/4 She was leading a gay life.


The fact that any particular word has different meanings in different
contexts is meaningless in this discussion -- in any discussion, for that
matter. The /only/ example in science or engineering I'm aware of is
"torque", which is supposed to mean "rate of change of angular momentum",
but also has the meaning of "force x distance" (where distance is the
distance from the point of application of the force to the center of
rotation).

English is a wonderfully rich and complex language. Unfortunately, there is
no official body (as the French have) attempting to keep the language fixed.
This can be a bad thing when words are redefined by common usage, * which
weakens the language's vigor. Perhaps the worst example (to a logophile like
myself) is the recent use of "impact" as verb (which it didn't used to be),
to replace many stronger, more-appropriate, more-vigorous, and more-subtle
words: alter, modify, influence, affect, confront, vary, harm, ruin, change,
etc, etc, etc...

I also dislike changing the pronunciation of words, which has the same
effect: "Clique" is pronounced "cleek", not "click". Pronouncing it the
latter way creates an unnecessary homonym.

* The OED is based on common usage, not the opinions of "experts".


Now if I want to be linguistically accurate, the word, thirty years later,
still means nothing other than happy and carefree.

What do you mean by linguistic? For better or worse, the English language
has changed, and will continue to change.


But a gazillion homosexuals around the world would disagree with me
very strongly.

I, for one, don't like the word. I describe myself as "homosexual" or
"queer".


Are they wrong, and me with my lone voice, right? No, neither of us is
right or wrong. I am theoretically right, and they are practically right,
because the word has now been accepted into modern language to mean
something other than its 'real' original definition.

I, too, would prefer that words kept their "original" meanings. Other than
censoring printed documents, there doesn't seem to be any way to do it,
except complaining vigorously and hoping English teachers will set an
example.

Perhaps it's not pertinent, but what about Shakespeare? He /added/ many
words to English (though whether all of them were his coinage, or merely
their first appearance in print is debatable). And we still add words, or
broaden their meanings, which can be a good thing.

By the way, note the misuse of common expressions, the most-common of recent
examples is the conversion of "the proof of the pudding is in the eating" to
"the proof is in the pudding". (ARRRGGGHHH!) Shakespeare's "one fell swoop"
is almost always misued -- in most cases it should be "one swoop". (Note the
meaning of "fell", and the phrase's usage in "The Scottish Play".)


The same is true of the word "digital". Thirty years ago, your narrow
definition of a digital signal being one that has been quantized into
representative numbers, was correct. You could not have applied an
analogue
audio signal to a Z80 data line, and have expected it to have been able to
do anything with it. However, now, you could apply a two level PWM audio
signal, which you contend is really still analogue, to a port pin on a
completely digital uP IC, and it would have no problem being able to
manipulate that signal, given the appropriate code to do so. The world of
electronics has moved on since the original definition of digital, and the
lines between analogue and digital signal processing, have become much
more fuzzy in the process, to the point where the original 'narrow'
theoretical
definition of digital, no longer hacks it in the real world.

I believe in calling a knife a knife. (Here's an example of (a sort-of)
re-definition: In Italian, "spada" means "sword", not "spade".) Where is the
logic in allowing a clearly defined scientific or technical term to be
incorrectly redefined by common usage?


However, now, you could apply a two-level PWM audio
signal, which you contend is really still analogue.

A "two-level" signal (or a signal with any finite number of defined levels)
is, by definition, digital. But a PWM signal, regardless of its amplitude,
can be analog or digital. It is how the width of the pulse varies
(discretely or continuously) that determines digital or analog, not the fact
that it's a pulse. The width (not the amplitude) of the PWM signal is its
(signal) "level".


Whether or not you like it, or think it right, the word "digital" now
tends
to encompass any means of data transfer between devices or equipments,
or any signal processing technology, which employs just two levels.

But (as I pointed out above), the levels DO NOT convey the data! The change
in amplitude (from 0 to B+, and back again) has nothing to do with the
information being transmitted. To refer to all data-transmission systems
that use pulses "digital" is to ignore the clear meanings of these terms.

There are lots of people who think Einsteinian relativity is a pile of crap.
Does that make it so? If the majority of people did, would it be so?

I don't mean this sarcastically, but does it bother you when I get upset
when people say things that show they don't fully understand what they're
talking about? I think this should bother everyone.


Any engineer working in the real world of electronics will tell you this.
The many millions of people who believe this, and publish on the web,
are *not* wrong, just because you believe that they are.

This is Humpty-Dumptyism -- a word has a particular meaning because I say it
does.


Yamaha for instance, define a PWM audio signal as being digital,
and with no apology, see
http://www.global.yamaha.com/news/2003/20031002.html
Do you honestly consider that the designers at a well-respected
company such as they, are wrong ?

I have no idea what the designers think, as this product sheet was (likely)
written by Americans in Yamaha's marketing department. But engineers have
been known to be wrong.

Analog signal input circuits, pulse width-modulation circuits (Note 2),
bridge-tied load (BTL) output circuits (Note 3), self-oscillating circuits
(Note 4), overcurrent protection circuits, pop-noise suppression circuits
(Note 5), headphone amps, and other components necessary for digital
amplifiers are combined in an extremely small 28-pin TSSOP.

There's nothing about an S/PDIF input, so I have to assume the PWM is
analog, not digital. Note 2 reads "Pulse width-modulation circuits: The
circuits lengthen and shorten the widths of digital pulses, amplifying the
voltage of sound signals."

A pulse is neither analog nor digital. It's how it's used to convey data
that determines "digital" or "analog".


Likewise, Sanken and Sanyo describe their PWM class D
amplifier ICs as being "digital". Wrong also?

Yup.


And Tripath with their famous TA2020 IC used in many
home-cinema systems?

The block diagram suggests that the pulse width is modulated directly (ie,
there is no quantization -- what would be the point, as it would needlessly
complicate the circuit?), so, yes, this is an analog amp.


Kenwood? Sony? JVC maybe? All wrong to call their PWM-based
class D amplifiers, "digital"?

Wrong, wrong, and wrong. They are switching amplfiers, and probably analog.


If you really believe this, then might I suggest that you try e-mailing a
few of their technical or design departments, and put it to them that they
are wrong, and outline your reasoning, based on your definition of the
word,
and then report back what they have to say to you?

Again, I say that you are not wrong, in theory, but neither are you right
in
practice, when referring to today's much-changed world of electronics.

There is an issue of personality here. Being queer, I'm not willing to agree
with the majority. Other people are more-affable, and willing to bend to the
general view. In social matters, they're usually right. In
scientific/technical matters, I don't understand why misunderstanding the
facts doesn't greatly bother people.

"Experts" can be wrong. It's almost certain that Einstein was wrong in
rejecting the implication of quantum physics that, on the sub-microscopic
level, there is no cause and effect -- events occur statistically,
determined by the "maths", not because there is a causative force. He didn't
like it -- "God does not play dice!" -- and rejected it, despite the math
and laboratory observations being against him.


I'm going to do something I hate doing. I have to call Naitonal
Semiconductor next week about a job. I will ask Martin Giles (one of your
countrymen, and an extremely intelligent fellow *) how he feels about this
issue, from both a technical and practical matter. I will also send e-mail
to Bob Pease, National's chief designer, guru, mascot, and resident
curmudgeon, though I doubt he'll reply.

I'll report back what they have to say (if anything).

* How do I know he's intelligent? Because a few years ago, I asked him what
the one basic principle you needed to explain to someone who was learning to
design op-amp circuits was -- and he instantly came back with the correct
answer.