OK, you'll be sorry!

First, I think you should use 44.1k, since we're going to end up at
that speed anyway, if we record to CD.  If 48K is all ya got, then
that's how it is.  But in any case, it shouldn't be the cause of your
trouble.

I'm convinced that there are clock rate problems.  I have a clock rate
ON MY OWN SYSTEM, because audio recording uses my laptop's built-in
audio chip's clock source, and my MIDI uses the "system" timer.  More
on that later.

There are a number of problems a clock can have, but we'll focus on
two: frequency offset and frequency variation.  We'll ignore drift.
If yo want to learn lots about these things, go to
\link{http://www.udel.edu,www.udel.edu} and search for NTP.
I did some work in NTP and had a few conversations with this
genius, who invented the protocol computers use to synchronize
their clocks.  (He even acknowledged me for a trivial contribution
I made to one of his RFCs, which are specs for internet protocols!
 Whee! I'm famous!)

Now, a decent crystal clock, like you find on any synth, and like
you should find on any half-decent sound card, should have a
frequency variation we'd never notice.  And the frequency offset
should be small enough that we should only notice it on a very very
long song.

The system clock on my PC, on the other hand, seems to be an RC
oscillator.  I assume this because it loses about a minute a month
on the time-of-day clock.  I'm not sure that the system clock that
n-Track has to use for my MIDI is the same as the TOD clock, but
anyway the TOD clock is off by 0.0032%.  This means that in a 5
minute song at 100 bpm, by the end I'd be off by a 64nd note,
which is musically meaningful and not exactly "in the pocket",
but not too bad, close enough for me most of the time.  By the end
of a 20 minute opus, though, it would be off by a 16th note, which
would be garbage.

Now, Freddy is very right about a lot here.  If your system uses the
same clock for everything, then everything on your system should line
up without a hitch.  If Newt records ten second of audio on a system
with a steady but slow clock at 100 BPS, his system will still assign
44,100 frames to that "second", although it will really be longer than
a second.  When Bob plays it back (you didn't know that Newt & Bob
collaborate?), the track will be EXACTLY 10 seconds long, according
to Bob's clock.  Why?  Because it contained 441,000 samples!  If
Newt had clicked once every second (according to his system's clock),
Bob will hear a click once every second, exactly, according to
his system's clock.  Regardless of whether it's fast or slow.  But
let's assume Bob's clock is fast.  Where does the missing time go?
Let's say Newt actually recorded 10.1 seconds (his clock is 1% slow)
and Bob's playback actually took 9.9 second (his clock is 1% fast).
Of course, nobody notices the error.  But if Bob has perfect pitch,
he might notice that the pitch is 2% higher than it should be.

MOST CLOCK ERRORS CAUSE PITCH ERRORS, not time errors.  This is
why Freddy is very right about clocks rarely being a problem.
there would have to be a REALLY bad clock for it to cause enough
of a pitch problem for even the sharpest of us to notice it.
A 0.1% error, worse than I expect from a clock (it would lose or
gain a half hour in a month!) would only cause about a 1 cent
pitch change.

So what's the problem?  Well, here's a problem I have.  If I have a
5 minute song with an audio track and a MIDI track, the timing of the
two drift apart as the song proceeds.  The reason is that I'm using
two clocks (the sound card's clock for audio, and the system clock
for MIDI), and the difference between them is significant.  n-Track
has an option to use the sound card's clock on the MIDI, but my
sound card doesn't support the necessary feature.

So, I suspect that sync problems are within each person's DAW,
and not a matter of posts.  Again, Freddy is right here.  Only,
it's not necessarily that human error is involved, it's that the
problem is occuring at the source.  If I post something and say
it's at 80 BPM, well, nobody else will agree -- and neither will
my own DAW!  The problem is within my system.

We could do a test, and have everyone post 100 BPM tracks -- but
all we would learn is whose systems are using different clocks
for audio and MIDI (or whatever the source of the click is).  And
then only if the two clocks on the person's system are different
enough to matter.