2
Date: Wed Jun 11, 2014 10:21 am ((PDT))
David:
Perhaps you mis-read Philipp's text.
High pass filtering is equivalent to a 'low cut' filter and it is used by n=
ature recordists to remove the low-end rumble and anthrophonic garbage that=
plagues many environments. Ho high frequencies are filtered out when using=
such filters and it wouldn't affect bird sounds unless it was something li=
ke the 'drumming' sound of a ruffed grouse.
Best,
Greg
From: " [naturerecordists]" <=
ps.com>
To:
Sent: Wednesday, June 11, 2014 8:05 AM
Subject: [Nature Recordists] Re: text spectrograms?
=C2=A0
> My question might sound somewhat exotic but it is something that
> occupies my mind this morning.
Philipp,
That is what we are here for, to answer questions if we can.
> While editing my recordings I sometimes do a bit of equalization. And I
> am still not satisfied with it. I am blind and not able to view the
> spectrogram of a sound file.
You can't balance a recording using a spectrogram. That is useful for
finding the frequency of an interfering sound. For instance there is a grai=
n
mill a mile away from my house which whines faintly at 780 Hertz which I ca=
n
reduce with a notch filter I made in the Audacity sound editor.
> That's why I have to adjust frequencies by
> trial and error method.
That's how you should do it.
> That's fine for High pass filtering but it is
> not practicable for subtle tasks.
Why do you filter the high frequencies? A lot of wildlife sounds are up
there. I think it is better to leave a little hiss in and not lose the
wildlife high frequencies. In any case, natural sounds also include hiss,
and hopefully this will drown out the hiss from the mic.
If listeners want to adjust your recording to their playback system, they
will have knobs to do that. It is better to record everything "flat" unless=
it is a major problem, in which case filtering won't help. The one exceptio=
n
is a bass cut or "roll off" filter which can stop wind and other low
frequencies from overloading your recording. If you later want more bass,
you can equalise that back in afterwards without losing any quality.
To lose some mic hiss, you need a sound editor noise reduction application,
but this can make the recording sound odd, so it is usually better to leave=
the hiss in if it is not too bad. The difficult part of nature recording is=
minimising unwanted sounds. Planes and cars are the usual worst offenders.=
> That is why I am wondering if there is
> any way of exporting/converting a spectrogram into a concise text file
In short, no. It would be too complicated a table of figures and it really=
would tell you very little. Trust your ears.
> ... so that I have a clear overview of the prominent developments of
> frequencies?
Do this by listening. It really is the only way. this is the art of sound
recording. If you want an example of various frequencies, I can make a soun=
d
file with a range of frequency beeps, so you know what each frequency sound=
s
like. That would also give you an idea of the frequency response of the
whole recording and playback chain from the mic, recording, amplification,=
headphones, and your ears.
Either a spectrogram or your ears won't show an even level, it never does.=
The plot is usually all over the place, and constantly changing.
It is useful to memorise frequencies like the human voice (low hundreds)
various birdsong (higher hundreds to lower thousands) and small birds,
higher harmonics, wind rustle and sometimes small mammal squeaks in the
upper thousands.
There is no way that you can create a sound balance by using a spectrogram,=
especially with birdsong. A spectrogram can show you the frequencies of
calls, but it simply cannot be a guide to a balanced "sound picture". A
spectrogram will not help you select the best stereo image.
The other very good reason for not using a spectrogram for recording is tha=
t
you don't have one outside. When it comes to pointing a stereo microphone,=
your ears are the only judge.
David Brinicombe
|