22 Oct
2007
22 Oct
'07
3:17 p.m.
Fons,
Thanks for your reply.
Yes, the format is U8, it is a simple recording program to make '.wav'
files. The filtering does work when adjusting the 'silence byte'. It
appears that the device drivers are not consistent with this value?
So I will be adding an option to change the value of the 'silence byte'.
How do I use the RMS value to remove noise? A URL that I could read?
Thanks again for your time.
--
William Estrada
MrUmunhum(a)popdial.com
Mt-Umunhum-Wireless.net ( http://Mt-Umunhum-Wireless.net )
Ymessenger: MrUmunhum
Message: 6 Date: Mon, 22 Oct 2007 08:53:34 +0200 From:
Fons Adriaensen
<fons(a)kokkinizita.net> Subject: Re: [LAU] USB device gives wrong data? To:
linux-audio-user(a)lists.linuxaudio.org Message-ID:
<20071022065334.GB5840(a)linux-2.site> Content-Type: text/plain;
charset=us-ascii On Fri, Oct 19, 2007 at 02:44:08PM -0700, william estrada
wrote:
> When I use my laptop's sound device and
record without a mic, I get a
> file full of x'7f's. If I do the same with my cheap-o USB device I get a
> file full of x'81's. The sound quality from the laptop is unacceptable,
> so that is way I want to use the USB device.
Those values are somewhat confusing - your device doesn't produce 8-bit
samples, does it ?
A 'perfect' interface would probably produce zeros on silence. You may be
seeing a small DC offset.
Anyway, silence detection based on one particular sample value can't be
expected to work well, there will always be _some_ noise. Better calculate
the RMS amplitude and set a threshold on that.
-- FA Laboratorio di Acustica ed Elettroacustica Parma, Italia Lascia la
spina, cogli la rosa. 
23 Oct
23 Oct
8:54 a.m.
New subject: USB device gives wrong data?
On Mon, Oct 22, 2007 at 10:17:50AM -0700, william estrada wrote:
Yes, the format is U8, it is a simple recording
program to make '.wav'
files. The filtering does work when adjusting the 'silence byte'. It
appears that the device drivers are not consistent with this value?
So I will be adding an option to change the value of the 'silence byte'.
How do I use the RMS value to remove noise? A URL that I could read?
It's probably not worth the effort for 8 bits. What you see is a simple
DC offset. Probably the easy approach is to consider a range of values
e.g. 0x7E...0x82 as silence. Any value outside the range starts the
recording, until you find a long series of samples within the silence
range.
So the algo would be:
If you find a value in the silence range, increment a counter.
If the counter reaches a certain value, stop.
If you find a value outside the range, start and reset the
counter.
You could also use a counter to avoid too many false starts.
Ciao
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
Lascia la spina, cogli la rosa.
3:58 p.m.
New subject: USB device gives wrong data?
Fons,
Thanks for the response.
What you suggested is what I did but I only use one 'silence byte'. That
is definable to match the 'silence byte' for different devices. It seems
to work OK.
Do you think that a range of 'silence bytes' would work better?
William Estrada
MrUmunhum(a)popdial.com
Mt-Umunhum-Wireless.net ( http://Mt-Umunhum-Wireless.net )
Ymessenger: MrUmunhum
Fons Adriaensen wrote:
On Mon, Oct 22, 2007 at 10:17:50AM -0700, william
estrada wrote:
Yes, the format is U8, it is a simple recording
program to make '.wav'
files. The filtering does work when adjusting the 'silence byte'. It
appears that the device drivers are not consistent with this value?
So I will be adding an option to change the value of the 'silence byte'.
How do I use the RMS value to remove noise? A URL that I could read?
It's probably not worth the effort for 8 bits. What you see is a simple
DC offset. Probably the easy approach is to consider a range of values
e.g. 0x7E...0x82 as silence. Any value outside the range starts the
recording, until you find a long series of samples within the silence
range.
So the algo would be:
If you find a value in the silence range, increment a counter.
If the counter reaches a certain value, stop.
If you find a value outside the range, start and reset the
counter.
You could also use a counter to avoid too many false starts.
Ciao
25 Oct
25 Oct
9:55 a.m.
New subject: USB device gives wrong data?
On Tue, Oct 23, 2007 at 10:57:58AM -0700, william estrada wrote:
What you suggested is what I did but I only use one
'silence byte'. That
is definable to match the 'silence byte' for different devices. It seems
to work OK.
Do you think that a range of 'silence bytes' would work better?
It would depend less on
- the silence being 'absolute'
- the actual device
- drift of the DC offset with ageing, temperature etc.
Ciao,
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
Lascia la spina, cogli la rosa.
6053
days inactive
6056
days old
linux-audio-user@lists.linuxaudio.org
3 comments
2 participants
participants (2)
-
Fons Adriaensen -
william estrada