<<Thank you Scott: I have determined that the popping problem, that
John Hartog fixed by
editing, and you diagnosed, was indeed caused by some sort of digital
clock foul up.>>
BTW, for a preferable first option to cut & crossfade editing, if you
end up with digital glitches in a file & no access to a non-glitchy
version, quite often digital glitches can be drawn out of the
waveform with a pencil tool in the editor. Unlike acoustic glitches,
a digital glitch is usually a perfectly vertical short duration spike
imposed upon a somewhat sine-ish waveform, when viewed at a high zoom
level. These are frequently very easy to smooth out with a steady
hand on the mouse. This doesn't disturb the timeline, as a cut edit
will, & thus often is more transparent.
<<I originally transfered my DAT tape 16 bit files to my Logic Pro 7
Editor in 24 bit, 96K
sampling. They had originally been recorded at 44.1K. Popping. Pop.
Pop.>>
There's the problem. A digital to digital transfer needs to be
synchronized at exactly the same sample rate. With a SPDIF transfer,
wordclock is imbedded in the data stream & forces the destination
recorder to run at precisely the same sample rate as the source
machine. Basically every single sample from the source, all 44,100 of
them per second, is recorded as an exact duplicate in the destination
machine, resulting in a perfect clone. If the sample rates differ you
don't have a one to one sample correspondence, & thus it is not a
perfect clone.
<<I redid things this morning, this time transfering the same 16 bit/
Message: 44.
Subject: 1K tape in 24 bit,
Message: 44.
Subject: 1K, with the synchronization function set to "auto enable external
sync."
This time I got no popping at all, after transferring an hour of
recordings.
The occasional popping was not on my recordings on the DAT tape! They
are fine. It was
on my hard drive. It was caused by some sort of syncing problem it
seems. >>
Right. The two machines were not synchronized, thereby drifting out
of alignment, causing the digital snats.
<<Perhaps only "dufusses" like me go between 44.1 K and 96 K. I
thought if you could take
a 16 bit recording and edit it in 24 bit, you could take a 44.1K
recording and edit it at
Message: 96K.
Subject: Perhaps that makes no sense. Do others do that?>>
You can upsample to 96k with a sample rate conversion in your
software. Or you can record the line out of your DAT deck into an
analog input of your M-Audio interface at 96k, although this will
result in some signal degradation & will not be a perfect clone of
your original recording. Either way, converting an original 44.1k
recording to 96k will not provide any additional audio information,
although some point can be made that some editing or DSP processes
will be smoother when rendered at higher sample rates. As with the 16
vs 24bit thread, there is debate as to whether this possible benefit
is actually audible or worthwhile.
Scott Fraser
|