<tt>Hi all,</tt><br>
<br>
<tt>I thought two notes could be added to the suggestions made about hard- and
software audio recording machines for bioacoustics, now that digital audio
is beating records at attractive prices.</tt><br>
<br>
<tt>Watch out.</tt><br>
<br>
<tt>1) On the hardware side:<br>
I've come through an excellent audio acquisition board, famous brand, with
96kHz SR & 24 bit resolution, and was astonished that the device was
actually cutting at 20kHz. High end audio systems like the previously
mentioned Pro Tools Digi 002 have specs only until 20kHz (see
www.protools.com) , above which nothing is guaranteed and I would not bet
that they don't filter above 20kHz too to secure dynamic range and only use
the high SR to reduce the AD/DA converter distortion. High SR (96-192 kHz)
audio boards do not mean "ultrasound enabled" or "can record up to half the
sampling frequency".</tt><br>
<br>
<tt>2) On the software side<br>
Audio software generally provides a plethora of attractive filters, down-up
sampling, interpolation, file format conversion tools, etc... for which
very little is guaranteed or documented at "microscopic" level (phase
distortion, for example) and although to your ear and on your spectrogram
the job sounds and looks good, your data may be irreversibly unusable for
fine analysis, like vocalization pattern recognition etc... Lots of tricks
on audio softwares are pure and hardcore signal processing but meant for
audio, human hearing, that the manufacturer won't even want to disclose
information about when you need it.</tt><br>
<br>
<tt>This is where, for example, on both hardware and software sides,
well-known
instrumentation brands find a niche on the acoustic market - and are not
always that expensive, why softwares like Ishmael were created (free, if i
remember), and why lots of bioacousticians work with their well-known,
regularly calibrated DAT machines (even if they sound old fashioned) or
rough ten year old A/D acquisition boards, and process everything under
Matlab - expensive or Octave -free!. The learning (and the initial expense)
is tougher but always worth it eventually.</tt><br>
<br>
<tt>Still, an excellent audio recording station like a Pro Tools system may be
turned into a scientific instrument for some applications that find its
specs sufficient and adapted to the task. But, although i wish it were -
and I am thinking of getting one for my home-studio, it is not aimed for
acoustics but ... audio recordings, our human ears. Eventually, audio
systems are not designed to provide more precision or signal integrity than
what human hearing needs, and not the precision and integrity that your
analysis may find necessary.</tt><br>
<br>
<tt>Best,</tt><br>
<br>
<pre style="margin:
0em;">______________________________________________________
Eric Delory
Laboratori d'Aplicacions Bioacustiques
D164, Escola Politécnica Superior d'Enginyeria de Vilanova i la Geltrú
Universitat Politécnica de Catalunya
Avda. Víctor Balaguer s/n
08800 Vilanova i la Geltrú
Barcelona, España.
Tel. (34) 93 896 27 72
Fax.(34) 93 896 77 00
H/P.(34) 655 33 0196
[EMAIL PROTECTED]</pre><br>
<br>
|