[Bradford] Convert wave amplitude into a float

[John R. Hudson]

Not at all. This is not about analysing the noise characteristics of
equipment for the design purposes that people would use the B&K
equipment. This is about taking an existing profile and setting a
baseline; each baseline will be different for each piece of equipment.
The problem with 'learning' is that the equipment might 'learn' the
wrong things.

Also they are aiming to produce something significantly cheaper than B&K
equipment.

John
--
On Fri, 2014-03-21 at 12:50 +0000, david at nucleon.co.uk wrote:
> Does seem a bit like they are trying to reinvent the wheel.
> 
> There is already off-the-shelf hand-held kit available that does FFT
> spectrum analysis. We use it all the time, where I work.
> 
> There is also condition monitoring software that "learns" the
> characteristic noise spectrum of equipment, and flags up anomalies (20
> years ago I even had a next-door-neighbour who's PhD was in the subject).
> 
> Some relevant kit is sold here:
> http://www.bksv.com/Products/analysis-software/acoustics/noise-source-identification
> 
> Dave B
> 
> > I think Dave (S) is right,
> >
> > I guess they are going to be doing frequency-amplitude analysis, which is
> > commonly done by Fast Furiour Transforming samples into groups of
> > frequency-amplitude (think graphic equaliser). FFT is much easier to
> > program with floating point samples.
> >
> >
> > On 20 March 2014 20:49, John R. Hudson <j.r.hudson at virginmedia.com> wrote:
> >
> >> Thanks for all the suggestions which I will pass on if they haven't
> >> joined the mailing list yet. Because they want to do it continuously,
> >> reading the Alsa PCM device seems the most likely route. But maybe one
> >> of the other options will be more useful for them once they investigate
> >> them.
> >>
> >> However, thanks for the suggestion about how to export from Audacity as
> >> that will enable them to use the data they have already collected to
> >> develop the software before they need to apply to an actual stream.
> >>
> >> John
> >> --
> >> On Thu, 2014-03-20 at 20:25 +0000, David Spencer wrote:
> >> > What they *really* want to do is probably some kind of realtime signal
> >> > processing... which is not what Audacity is for (it's an editor).
> >> > They should probably be looking at reading the Alsa pcm device
> >> > directly (e.g. from Python); I bet that's what is pencilled in to
> >> > consume these hypothetical floats... either that, or the hypothetical
> >> > float-consumer needs to read an arbitrary well-understood audio file
> >> > container (e.g. WAV) using a library routine. Once the ints are in a
> >> > program, It's not rocket science to convert ints to floats prior to
> >> > feeding them to whatever other library routines are processing the
> >> > data (fft or whatever).
> >> >
> >> > The audacity question as posed just sounds like they caught the first
> >> > bus that came along :-)
> >> >
> >> > -D.
> >> >
> >> >
> >> > On 20 March 2014 17:46, John R. Hudson <j.r.hudson at virginmedia.com>
> >> wrote:
> >> > > Hi
> >> > >
> >> > > I popped in to see Paul Cannon this afternoon and they had just hit
> >> a
> >> > > problem - how to convert a wave amplitude as displayed in Audacity
> >> into
> >> > > a float which can be fed into another program.
> >> > >
> >> > > There is no obvious way in which to do this from Audacity. So it may
> >> > > involve using different software.
> >> > >
> >> > > Any thoughts very welcome.
> >> > >
> >> > > John
> >> > > --
> >> > >
> >> > >
> >> > >
> >> > > _______________________________________________
> >> > > Bradford mailing list
> >> > > Bradford at mailman.lug.org.uk
> >> > > https://mailman.lug.org.uk/mailman/listinfo/bradford
> >> >
> >> > _______________________________________________
> >> > Bradford mailing list
> >> > Bradford at mailman.lug.org.uk
> >> > https://mailman.lug.org.uk/mailman/listinfo/bradford
> >>
> >>
> >>
> >> _______________________________________________
> >> Bradford mailing list
> >> Bradford at mailman.lug.org.uk
> >> https://mailman.lug.org.uk/mailman/listinfo/bradford
> >>
> > _______________________________________________
> > Bradford mailing list
> > Bradford at mailman.lug.org.uk
> > https://mailman.lug.org.uk/mailman/listinfo/bradford
> >
> 
> 
> 
> _______________________________________________
> Bradford mailing list
> Bradford at mailman.lug.org.uk
> https://mailman.lug.org.uk/mailman/listinfo/bradford