Re: AUDITORY Digest - 13 Mar 2009 to 14 Mar 2009 (#2009-60)

[Matthew Reynolds <matthew.reynolds@xxxxxxxxxx>]

    Hi Nick,

There are a few common ways of presenting audio data in a printed form each
with their advantages and disadvantages depending on what you want the data
to convey. I hope I'm not being patronising in describing the three most
common by far. 

The most simplest is merely providing a time-history, i.e. A picture of the
waveform as you would see in a wave editing program such as audacity or
Adobe. This is a graph with time as the x-axis and wave amplitude as the
y-axis. This is useful is you want to see unusual physical features of the
sound, such as clicks and pops, or to visualise tonal elements or periodic
modulations. While this is the simplest it is often not used, as waveforms
of recorded sounds often look like random garbage to the naked eye.

Another common tool is to take a fourier transform or calculate the power
spectrum of a signal. These both provide a way of looking at sound in the
frequency domain. So now the X-axis becomes frequency and the y-axis is
amplitude or power. So if a sound has a loud 1kHz tone in it, the graph will
have a large peak at 1000Hz. This is an extremely powerful way of analysing
signals, but does not provide any indication of how the signal changes over
time, but of a total of the whole signal (or a snapshot of the part of the
signal where the fourier transform or power spectrum is applied). You must
also be careful of drawing false conclusions from results as processing the
data in this way can lead to information being lost or clouded due to the
windowing that is required, or spurious peaks can sometimes be seen that
don't necessarily represent a useful part of your signal.

The third tool is called a spectrogram, which is essentially a combination
of the two tools described above. It is a three dimensional graph (x-axis,
y-axis, colour) where the x-axis is normally time, the y-axis frequency and
colour represents the amplitude/power of the signal. Think of it as many
fourier transforms taken over a period of time, and then stacked up next to
each other. So you can see the content of the signal as a function of
frequency but can also watch how this changes over time. The disadvantage of
this is that with so much information represented on a single graph, any
small changes, or fine information will be hard to detect.

I'm sure wikipedia or google will be full of information about these
techniques (along with some complicated looking maths for the uninitiated).
I hope this has helped, and good luck.

Matt Reynolds




On 15/03/2009 04:18, "AUDITORY automatic digest system"
<LISTSERV@xxxxxxxxxxxxxxx> wrote: 

> There is 1 message totalling 39 lines in this issue.
> 
> Topics of the day:
> 
>   1. Representing Sound in non-audio formats, specifically  Journal Articles
> 
> ----------------------------------------------------------------------
> 
> Date:    Sat, 14 Mar 2009 14:16:00 +0000
> From:    Nick Bearman <nick.bearman@xxxxxxxxx> 
> Subject: Representing Sound in non-audio formats, specifically  Journal
> Articles
> 
> I'm in the process of writing a journal article about the research I
> have done on using sound to represent spatial data. As I'm sure you are
> all aware, there are many issues when trying to demonstrate research
> involving sound in a written journal paper, particularly when the
> journal is not used to receiving these types of submissions.
> 
> Does anyone know of a good text exploring the issues of writing an
> article trying to represent sounds in a very non-sonic way?
> 
> Obviously there are ways to make the sound available via the Internet
> such as sound clips or examples of the program in question. However I'm
> specifically interested in ways sound can be represented in a written or
> visual medium for the article itself.
> 
> Thanks,
> Nick.
> 
> P.S. I've specifically been ambiguous about the type of sound to be
> represented, as even though a particular method may not work for my
> data, it might be useful for someone else or prompt ideas for my data in
> a very different way!
> 
> --
> Nick Bearman
> PhD Student
> School of Environmental Science
> University of East Anglia
> Norwich, NR4 7TJ
> E-mail: n.bearman@xxxxxxxxx 
> 
> ------------------------------
> 
> End of AUDITORY Digest - 13 Mar 2009 to 14 Mar 2009 (#2009-60)
> **************************************************************
>

---

The information in this email is confidential. It is intended
solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any
disclosure, copying, or distribution is prohibited and may be
unlawful. If you have received this email in error please delete
it immediately and contact commercial@xxxxxxxxxxx

Sepura plc. Registered Office: Radio House, St Andrew’s Road, Cambridge, CB4 1GR, England. Registered in England and Wales. Registration Number 4353801

---

This email message has been scanned for viruses by Mimecast.
Mimecast delivers a complete managed email solution from a single web based platform.