[linux-audio-dev] Radio receiver.

fons adriaensen fons.adriaensen at skynet.be
Fri Oct 28 00:16:16 UTC 2005


On Thu, Oct 27, 2005 at 10:20:06PM +0000, carmen wrote:

> > There really is no such thing as a 'noise filter' - noise usually
> > occupies the whole band, so a noise filter would remove all of the
> > signal ! 
> 
> there is definitely such thing as a noise filter. i know i had a
> couple dozen VSTs in my 'filter/noise' folder on windows... 

A filter in the usual (audio) sense is a linear process that
has a frequency dependent gain. It will have the same effect
on any signal, be it noise or discrete tones. In that sense,
a 'noise filter', supposedly acting only on a continuous
spectrum (noise) does not exist.

There may of course exists things called 'noise filter',
and having the effect of reducing unwanted noise in some
application.

These could be just linear filters attenuating what their
designer considered to be the unwanted part of the spectrum.
But that is application specific - one (wo)man's noise is
another one's signal. If you are recording an interview
at the seaside, the waves are noise. If I'm trying to
capture the sound of sea, people's voices are 'noise'.
There is no filter that can know what either of us wants.
So we will both use some filter adapted to our needs,
and call that a 'noise filter'.

There are also processors that analyse a signal and remove
those parts of the spectrum that have little energy, or the
parts that correlate with a reference signal. The removed
part would be noise in many cases. But these are not filters
in the usual sense.

>From my experience on the short waves, it's not noise in
the normal sense (i.e. a continuous hiss) that makes reading
Morse difficult - we can quite easily 'remove' that mentally.
It's other forms of interference, and filtering can help to
reduce these.

You could indeed make some processor that would regenerate
a Morse signal (and also decode it on the fly). But you would
have a difficult time trying to outperform human hearing in
that application. For example, our hearing mechanisms are not
limited by the usual 'uncertainty principle' that limits the
product of resolution in time and in frequency, they can do
much better.


-- 
FA


 




More information about the Linux-audio-dev mailing list