Hey folks, I was recently doing some testing of Jamin, Ecasound and
Jack, remastering some field recordings. I was looking pretty closely
for errors and weirdness as I was testing several software packages for
the first time. Two issues that came up were caused by Audacity.
FYI, I am using audacity-1.2.0-0.pre3.1.rh90.ccrma.i386.rpm, the latest
RPM packaged for Planet CCRMA (redhat 9.0) on a PIII-550 w/320 MB RAM.
1) Sometimes, ~50% of the time, a single sample is cut off the end of
the wav track after it is exported. This is certainly not audible but
it does raise a (pedantic) technical question and actually causes me a
I am working with live recordings that segue, so they need to be
recorded in DAO mode with no sector boundary errors at the ends of the
tracks. I have been careful to set the 'snap to' to min:secs.frames,
and to always turn it on when making edits in the hopes of eliminating a
step in the mastering process (repairing the boundaries with shntool).
Is this a known issue, or is this something new? What could be causing
Audacity to be dropping a sample when it exports to wav?
2) When using the normalize function with its default settings,
including the don't allow clipping checkbox, the tracks always end up
with clipped peaks in the waveform display. Is the dB waveform the only
one that is relevant here? Is there some other issue that I am
overlooking? Normalize to 0 dB should set the highest peak in the
waveform to be at 0 dB, no?
I have run across some other minor flakiness in an otherwise great
program that I use frequently. Thank you for all of the effort you have
put into making this application available us.
From: Matt Brubeck <mbrubeck@cs...> - 2004-02-26 18:33:56
Barton Bosch wrote:
> 1) Sometimes, ~50% of the time, a single sample is cut off the end of
> the wav track after it is exported.
I haven't managed to reproduce this yet. Are you using "Export As WAV"
or "Export Selection As WAV" (or both)?
Exact steps to reproduce and verify the problem would be useful, even if
they don't work 100% of the time.
> 2) When using the normalize function with its default settings,
> including the don't allow clipping checkbox, the tracks always end up
> with clipped peaks in the waveform display. Is the dB waveform the
> only one that is relevant here? Is there some other issue that I am
> overlooking? Normalize to 0 dB should set the highest peak in the
> waveform to be at 0 dB, no?
That's right. The Amplify command (with "don't allow clipping" checked)
should never introduce clipping. You shouldn't see clipping in either
Are you sure the peaks were not already clipped before the Amplify was
applied? Is the clipping audible? (If not, you may have zoomed in
vertically by accident, by clicking in the vertical ruler.)
Some before/after screenshots (zoomed in to show the peaks before and
after clipping) or audio clips might be handy. You can post them on a
web page, or send them to me at <mbrubeck@...> (this list rejects