Text 11848, 94 rader
Skriven 2008-09-30 00:10:06 av James Bradley (1:342/77.0)
     Kommentar till en text av Maurice Kinal
Ärende: smoke gets in your eyes
===============================
09-28-08 16:01: Maurice Kinal to James Bradley, "smoke gets in your eyes"

 > The "modern" studio, as well as any home-based ones worth their
 > muster lose the analog designation *before* the signal is store-able.

 MK> Maybe.  Depends on whether or not an AD filter is being
 MK> applied during the capture stage.  A passthru should
 MK> reproduce whatever the input is with the only filtering
 MK> being what the hardware is capable of reproducing. Using
 MK> mathematical definitions then the output for analog is
 MK> contiuous while for digital it is disrete.

 MK> digital:  . . . . . . . . .
 MK> analog:   _________________

 MK> Does that help?

Not really. Digital words need "yups" and "nyets" along with the stuff that
indicates that a new word is comin' right up, but them's just yups and nyets
anyway. So your nine-bit representation would suppose you are including a
redundancy check? Something I don't think is done with audio A-D coding. IIR,
they were too busy trying to stuff as much data into the distribution media,
that corrupt words would be tossed out before the D-A could be told if the
signal should increase its slope-path, hold the same, or decrease the
slope-path with the miss-read word. "Too late. Throw it away." 

Which brings me to the analog slope. Your representation would be a pretty
boring tune. <G+D>

 > Magnetic tape; maybe to helically stripe data onto it.

 MK> See above.  An anolog signal is easily reproducable on magnetic
 MK> tape.

The signal on the tape, best represents the original if it is recording a
digital signal. I think that was pretty close to being true at 44.1k 16bit.
(Isn't that the CD spec?) Regardless, at 48k - easily doable with '90s HD
recording HW - you would *have* to be pretty *stellar* to release a record that
could shake a stick at that digital quality. Then, we can toss into the "mix",
non-destructive editing, and lossless copying regardless how many generations
you make. There is a reason, even the best made analog decks were reduced to
complicated effects boxes.

 > CD... What's analog about that?

 MK> I might be wrong about that.  Offhand I thought CD music is
 MK> in a fancy wave file which seems to ne to be analog (ie a
 MK> continuous stream). The fact that a CD spins in a
 MK> continuous manner makes it an analog piece of hardware no
 MK> matter what type of file is etched onto the CD.

Likewise, a trail of bread crumbs might lead you out of the woods, but the same
trail leads to Grannies house. Analog signals need to remain sinusoidal to
retain the analog designation. Even your radio signal is a modified analog
signal, but there is also a good reason that the signal presented to the radio
stations modulator is most often a D-A conversion. Analog signals molded into
vinyl or shellac just isn't able to weather playback like CDs. Nothing touches
the data on a spinning CD, but the light from a laser. Run the best needle, on
the best arm, on top of the best turntable, and you *still* reduce the fidelity
of the plater after every pass. The same degradation occurs on tape, hence the
ninth-CRC bit on streaming data tape cartridges. 

 > What came first, ASCII or machine code?

 MK> For me?  Both started at the same time, except that ASCII was

You missed the "Chicken and the egg" inference. WWII math machines used code to
decrypt alphanumeric messages. Turn it back a few thousand years, and we have
an abacus doing pure math. In the middle, a slide rule could accomplish
logarithms and such. I've heard arguments that these early math machines were
the first computers, able to store and retrieve data. Now, I use a compression
alogarithm on my ASCII to economize bandwidth. I don't understand your aversion
that.

 MK> to an attached modem to act upon ought to have the desired
 MK> result, such as in the case of AT commands, which are
 MK> ASCII.  No?

Because I use a POTS for giving you a hard time, it matters nothing to me, if
the switching station is digital, or a city block of rotary relays. As long as
the call gets through, t'hell do I care how the tel-co gets my modem to connect
to my provider. 


 MK> No doubt.  I am analog.  :-)

Analogous, to be *sure!* }B-]


... James
___ MultiMail/Linux v0.49

--- Maximus 3.01
 * Origin: -=-= Calgary Organization CDN (403) 242-3221 (1:342/77)