Mike Keithley | 1 Jan 2005 18:51
Picon
Favicon

Noise-reduction error

When I try the noise reduction effect on a file, I get this error at some 
point:
sox: noisered.c:227: process_window: Assertion `s >= -1 && s <= 1' failed.
Aborted

The file is mono and sampled at 22050.  The sample is about 45 seconds 
of noise and is high-frequency hiss with a slight 2500 hertz peak.  I 
tried a threshold of 0.25 but this resulted in the same error but earlier 
in the file.

I read in an earlier thread about a library called fttw in regard to 
noise-reduction but this library is not on my machine.  I am running the 
2.6.7 kernel and sox version 12.17.7.

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
Kevin Cosgrove | 1 Jan 2005 20:48

Re: Noise-reduction error


On 1 January 2005 at 10:51, Mike Keithley <mkeithley <at> pobox.com> wrote:

> I read in an earlier thread about a library called fttw in regard to 
> noise-reduction but this library is not on my machine.  I am running the 
> 2.6.7 kernel and sox version 12.17.7.

Not sure about your error report or problem.  But on my Mandrake 10.1 
system fftw is in in libfftw2 and libfftw3 packages.  If your 
distribution is pretty new, then you might be similarly lucky.

Cheers...

--
Kevin

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
John Que | 2 Jan 2005 14:04
Picon
Favicon

converting gsm to wav with sox

Hello,

I am taking a gsm compressed file and try to play it with
Windows Media Player.
As I understand, windows media player does not support playing
gsm compressed files.
So I want to convert it to a format that windows media player will
recognize.
When I try:

sox -V in.gsm -s -w out.wav
it does succeed; However, the output file
is more than 10 times bigger ! (the input was 15 K and the output more than 
150 K).
It seems to me that the reason is that the out.wav in this case is 
uncompressed.

Is there a way to do it so that it will have ,more or less, the original 
size ?
(I mean : converting this gsm file to a format which Windows Media Player 
will
recognize)

regards
John

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar - get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/

(Continue reading)

Mike Keithley | 3 Jan 2005 20:20
Picon
Favicon

Re: Noise-reduction error

It appears the problem was bad sampling in the source file.  It was 
originally an mp3 file and I think the encoding got interrupted , or 
something like that.

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
Justin Gombos | 3 Jan 2005 22:15

Re: converting gsm to wav with sox

* John Que <qwejohn <at> hotmail.com> [2005-01-02 10:08]:
> 
> Is there a way to do it so that it will have ,more or less, the
> original size ?  (I mean : converting this gsm file to a format
> which Windows Media Player will recognize)

If you recall my issue of going the other direction (wav to gsm), it
turned out that sox did exactly what it is asked, not necessarily what
made sense.  Since the GSM format has constraints such that the
bitrates and other parameters are degraded to only accommodate voices,
I would guess that you have to use options for maintaining the same
constraints on the wav file.  ie. Don't use a default sample rate of
44100.  You can probably reduce the size just by doing a -r 8000, and
doing a mono encoding.

Additionally you can compress the thing, according to the man pages:

  SoX currently can read PCM, ULAW, ALAW, MS ADPCM, and IMA (or DVI)
  ADPCM.  It can write all of these formats including (NEW!)  the
  ADPCM encoding.

To use these compression methods, the following switches are provided:

  -s/-u/-U/-A/-a/-i/-g/-f

But the silly man pages don't tell us which switch is for what
algorithm.. so it seems you'll have to experiment.

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
(Continue reading)

Duncan McQueen | 7 Jan 2005 20:48
Picon

Converting 5.1 PCM to stereo

Can sox convert 5.1 PCM files (extracted from a DVD-Audio) to stereo?

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
W W M | 12 Jan 2005 21:40
Picon
Favicon

Probably using stretch effect incorrectly

I was trying to squish together a wav file (which had been generated from
"timidity -Ow foo.mid foo.wav" ) using the stretch effect, along the lines:
sox foo.wav foo_faster.wav stretch 0.9 [20]
What I get is of course faster without altering pitches, but I got all these extra 'artifacts' which seemed to be a higher pitch as I lowered the 'window' parameter (which is 20 by default, iirc)
I tried reducing the window value to 0, but it takes *significantly* longer -- in fact it was still going when I had to leave for work this morning.  Is zero not a valid value for 'window'?
 
The artifacts at window value 20 are below and parallel to the original pitch, but as I lowered the value to, say, 0.3, the pitch of the artifact was higher and parallel, but still not completely gone.
How can I get a 'clean' version?  Or should I just reload the midi file into perhaps rosegarden or Muse, and speed up the tempo that way?  Probably more efficient....
TIA,
Mark

Do you Yahoo!?
All your favorites on one personal page
Kevin Cosgrove | 13 Jan 2005 00:32

Re: Probably using stretch effect incorrectly


On 12 January 2005 at 12:40, W W M <scottishwalrus <at> yahoo.com> wrote:

> Or should I just reload the midi file into perhaps rosegarden
> or Muse, and speed up the tempo that way?  Probably more
> efficient....

ABSOLUTELY, if you have MIDI, then that is the best way to change the 
tempo.  It doesn't take as long to do, and the sound is better.  I'm 
not fully versed in TiMidity.  It might have a tempo option.

G'luck....

--
Kevin

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
W W M | 13 Jan 2005 01:24
Picon
Favicon

Re: Probably using stretch effect incorrectly

Kevin Cosgrove <kevinc <at> doink.com> wrote:


On 12 January 2005 at 12:40, W W M wrote:

> Or should I just reload the midi file into perhaps rosegarden
> or Muse, and speed up the tempo that way? Probably more
> efficient....

ABSOLUTELY, if you have MIDI, then that is the best way to change the
tempo. It doesn't take as long to do, and the sound is better. I'm
not fully versed in TiMidity. It might have a tempo option.

G'luck....

--
Kevin

I tried with timidity, using the -T n% option, but it didn't seem to be doing anything.  I'm fairly sure rosegarden4 or Muse will do it, though.  Any timidity experts out there?  I'm using Agnula 1.2.0, FWIW.

Thanks for your reply, Kevin!

Mark


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
Sox-users mailing list
Sox-users <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sox-users

Do you Yahoo!?
The all-new My Yahoo!
Kevin Cosgrove | 13 Jan 2005 01:29

Re: Probably using stretch effect incorrectly


On 12 January 2005 at 16:24, W W M <scottishwalrus <at> yahoo.com> wrote:

> I tried with timidity, using the -T n% option, but it didn't
> seem to be doing anything.  I'm fairly sure rosegarden4 or Muse
> will do it, though.  Any timidity experts out there?  I'm using
> Agnula 1.2.0, FWIW.

Rosegarden, which I use, would do that easily.....

 
--
Kevin

-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Gmane