sox-users Mailing List for SoX - Sound eXchange

On 31 July 2010 at 3:18, Doug Lee <dg...@dl...> wrote:

> Ah, but I mean transients *within* the recordings, usually caused by
> the recorder being turned on/off at record time, not
> playback/digitization time.  Other causes include cheap tape recorders
> being bumped, cheap microphones being handled or bumping a table, etc.

I have to do that all the time related to live sound recordings I 
make.  The best slightly automated way I've found to do that is to 
use the "fast look-ahead limiter" in Audacity with the edit region 
set to cover the transient in question.  I've improved the recording 
situation using in-line limiters and compressors and don't have to do 
as much by hand now.

Maybe that's not the SoX related answer you wanted.  But, at least it 
works.

G'luck....
--
Kevin

On Sat, Jul 31, 2010 at 07:36:30AM +0200, Jan Stary wrote:
> On Jul 30 17:39:51, Doug Lee wrote:
> > What's the best SoX way to remove the transient spikes from a tape
> > recording caused by turning the tape recorder on/off, so that a subsequent
> > normalization will raise the volume to a reasonable level without causing
> > clipping at the spikes?  I'm hoping there's a way to do this that does not
> > also compress the actual recordings much if at all.  It's almost like I
> > need an analysis that finds and limits short peaks that exceed the rest
> > of the sound, thus making an amplitude ceiling that is appropriate for
> > the actual content.  The compand effect can of course limit sounds over a
> > certain amplitude, but I think it's hard to tune that effect sufficiently
> > to make it cap off very short spikes without causing audible side effects,
> > and also hard to find that ideal amplitude at which to cap off.  But then
> > I come from a longer history of analog recording than digital. :)
> 
> When I was recording my tapes, I had the same problem.
> Simply start recording (with sox) only after you have
> started the tape rolling.

Ah, but I mean transients *within* the recordings, usually caused by
the recorder being turned on/off at record time, not
playback/digitization time.  Other causes include cheap tape recorders
being bumped, cheap microphones being handled or bumping a table, etc.

-- 
Doug Lee                 dg...@dl...                http://www.dlee.org
SSB BART Group           dou...@ss...   http://www.ssbbartgroup.com
"Believe, when you are most unhappy, that there is something for you
to do in the world. So long as you can sweeten another's pain, life is
not in vain." --Helen Keller

On Jul 30 17:39:51, Doug Lee wrote:
> What's the best SoX way to remove the transient spikes from a tape
> recording caused by turning the tape recorder on/off, so that a subsequent
> normalization will raise the volume to a reasonable level without causing
> clipping at the spikes?  I'm hoping there's a way to do this that does not
> also compress the actual recordings much if at all.  It's almost like I
> need an analysis that finds and limits short peaks that exceed the rest
> of the sound, thus making an amplitude ceiling that is appropriate for
> the actual content.  The compand effect can of course limit sounds over a
> certain amplitude, but I think it's hard to tune that effect sufficiently
> to make it cap off very short spikes without causing audible side effects,
> and also hard to find that ideal amplitude at which to cap off.  But then
> I come from a longer history of analog recording than digital. :)

When I was recording my tapes, I had the same problem.
Simply start recording (with sox) only after you have
started the tape rolling.

	Jan

What's the best SoX way to remove the transient spikes from a tape
recording caused by turning the tape recorder on/off, so that a subsequent
normalization will raise the volume to a reasonable level without causing
clipping at the spikes?  I'm hoping there's a way to do this that does not
also compress the actual recordings much if at all.  It's almost like I
need an analysis that finds and limits short peaks that exceed the rest
of the sound, thus making an amplitude ceiling that is appropriate for
the actual content.  The compand effect can of course limit sounds over a
certain amplitude, but I think it's hard to tune that effect sufficiently
to make it cap off very short spikes without causing audible side effects,
and also hard to find that ideal amplitude at which to cap off.  But then
I come from a longer history of analog recording than digital. :)


-- 
Doug Lee                 dg...@dl...                http://www.dlee.org
SSB BART Group           dou...@ss...   http://www.ssbbartgroup.com
"No person is your friend who demands your silence or denies your
right to grow." --unknown source

I have a program that needs to continuously compare a signal sinc-filtered
by sox to the original, non-filtered one. The signal is captured from a
device in real time, and nothing is saved on disk.

Obviously, the program cannot get its comparison (non-filtered) signal from
the same device sox is recording, so it seems I need to convert the signal
to stereo in sox, filter only one channel and then output both channels for
comparison. Is this possible in sox? Or is there an obvious piping hack I'm
missing?

Oona

This is very important to me if there is anyone out there who can help it would be appreciated!

I have also noticed that I can move the last audio piece forward in time to say at 5 seconds and it will fire at 5 seconds. However if I try to move it back in time and set it to 8 seconds it stays at the original spot where it is not suppose to be (around 6.125).  It is like 6.125 is its max although there is other audio playing under it at that point and further down.

It is as if it is putting the audio in a default spot and ignoring the time I am sending in after a certain point in time.  Thanks any help is greatly appreciated!

-Will





On Jul 22, 2010, at 2:54 AM, William Kerber wrote:

> Jan,
> 
> Thanks again for the quick replies.  I read FAQ3 but i mean this is from FAQ 7 also looked through the change log and didn't see that this was depreciated or anything like that (am I missing something you are saying?).
> 
> I know about the whole radio bit....I am just a developer who was asked to do this...sorry that its not choice audio.
> 
> I know it doesn't start at 7.203 because I loaded the wav up in protools.
> 
> If you align the wavs in there you can place file 1 and 0.0, file two at 3.306, and file 3 at 7.203 and KSFM at the end will come in after the hit (right on time).
> 
> Loading the sox mixed wav you can see in the wav form for and hear that around 6.125 or there about (because of the oversaturated radio i can't get exactly).
> 
> -Will
> 
> 
> On Jul 22, 2010, at 2:41 AM, Jan Stary wrote:
> 
>> On Jul 22 02:33:01, William Kerber wrote:
>>> Thanks Jan for the quick reply.
>>> 
>>> I know i dont have 1.wav etc, I just copied and pasted the command on the site for all to see.
>>> 
>>> sox -m /Users/william/file1.wav  "|sox /Users/william/file2.wav -pad 3.306" "|sox /Users/william/file3.wav -p pad 7.204" /Users/william/123.wav
>>> 
>>> is my actual command.  and this is the output I get:
>>> 
>>> sox FAIL formats: can't open input  `-': can't find sox file format identifier
>>> sox FAIL formats: can't open input pipe `|sox /Users/williamkerber/2.wav -pad 3.306': can't find sox file format identifier
>> 
>> Read FAQ3, too.
>> 
>>> sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358
>> 
>> This command is perfectly OK. What exactly makes you think
>> that file3 does NOT start at 07.203?  (In fact, each of the
>> individual files, and the mix, is so digustingly oversaturated
>> radio that it's hard to hear _anything_ in the mix).
>> 
>> 
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Sprint
>> What will you do first with EVO, the first 4G phone?
>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>> _______________________________________________
>> Sox-users mailing list
>> Sox...@li...
>> https://lists.sourceforge.net/lists/listinfo/sox-users
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

Jan,

Thanks again for the quick replies.  I read FAQ3 but i mean this is from FAQ 7 also looked through the change log and didn't see that this was depreciated or anything like that (am I missing something you are saying?).

I know about the whole radio bit....I am just a developer who was asked to do this...sorry that its not choice audio.

I know it doesn't start at 7.203 because I loaded the wav up in protools.

If you align the wavs in there you can place file 1 and 0.0, file two at 3.306, and file 3 at 7.203 and KSFM at the end will come in after the hit (right on time).

Loading the sox mixed wav you can see in the wav form for and hear that around 6.125 or there about (because of the oversaturated radio i can't get exactly).

-Will


On Jul 22, 2010, at 2:41 AM, Jan Stary wrote:

> On Jul 22 02:33:01, William Kerber wrote:
>> Thanks Jan for the quick reply.
>> 
>> I know i dont have 1.wav etc, I just copied and pasted the command on the site for all to see.
>> 
>> sox -m /Users/william/file1.wav  "|sox /Users/william/file2.wav -pad 3.306" "|sox /Users/william/file3.wav -p pad 7.204" /Users/william/123.wav
>> 
>> is my actual command.  and this is the output I get:
>> 
>> sox FAIL formats: can't open input  `-': can't find sox file format identifier
>> sox FAIL formats: can't open input pipe `|sox /Users/williamkerber/2.wav -pad 3.306': can't find sox file format identifier
> 
> Read FAQ3, too.
> 
>> sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358
> 
> This command is perfectly OK. What exactly makes you think
> that file3 does NOT start at 07.203?  (In fact, each of the
> individual files, and the mix, is so digustingly oversaturated
> radio that it's hard to hear _anything_ in the mix).
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

On Jul 22 02:33:01, William Kerber wrote:
> Thanks Jan for the quick reply.
> 
> I know i dont have 1.wav etc, I just copied and pasted the command on the site for all to see.
> 
> sox -m /Users/william/file1.wav  "|sox /Users/william/file2.wav -pad 3.306" "|sox /Users/william/file3.wav -p pad 7.204" /Users/william/123.wav
> 
> is my actual command.  and this is the output I get:
> 
> sox FAIL formats: can't open input  `-': can't find sox file format identifier
> sox FAIL formats: can't open input pipe `|sox /Users/williamkerber/2.wav -pad 3.306': can't find sox file format identifier

Read FAQ3, too.

> sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358

This command is perfectly OK. What exactly makes you think
that file3 does NOT start at 07.203?  (In fact, each of the
individual files, and the mix, is so digustingly oversaturated
radio that it's hard to hear _anything_ in the mix).

Thanks Jan for the quick reply.

I know i dont have 1.wav etc, I just copied and pasted the command on the site for all to see.

sox -m /Users/william/file1.wav  "|sox /Users/william/file2.wav -pad 3.306" "|sox /Users/william/file3.wav -p pad 7.204" /Users/william/123.wav

is my actual command.  and this is the output I get:

sox FAIL formats: can't open input  `-': can't find sox file format identifier
sox FAIL formats: can't open input pipe `|sox /Users/williamkerber/2.wav -pad 3.306': can't find sox file format identifier

Any idea about the "/usr/local/bin/sox FAIL delay: drained asymmetrically!" error message?

when I run:

sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358

Thanks your help is greatly appreciated, is there anything I can do to give you more information or more help in helping me?

-Will



On Jul 22, 2010, at 2:25 AM, Jan Stary wrote:

> On Jul 22 02:18:31, William Kerber wrote:
>> I also tried the commands from #7 on the FAQ page but the second one does not work at all on my machine.  It doesn't even run.
> 
> "Does work" and "doesn't run" are completely useless messages.
> Always provide the complete output, preferably with -V.
> 
>> sox -m f1.wav "|sox f2.wav -p pad 4" "|sox f3.wav -p pad 8" out.wav
> 
> You don't have f1.wav, you have file1.wav
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

On Jul 22 02:18:31, William Kerber wrote:
> I also tried the commands from #7 on the FAQ page but the second one does not work at all on my machine.  It doesn't even run.

"Does work" and "doesn't run" are completely useless messages.
Always provide the complete output, preferably with -V.

>  sox -m f1.wav "|sox f2.wav -p pad 4" "|sox f3.wav -p pad 8" out.wav

You don't have f1.wav, you have file1.wav

I also tried the commands from #7 on the FAQ page but the second one does not work at all on my machine.  It doesn't even run.

 sox -m f1.wav "|sox f2.wav -p pad 4" "|sox f3.wav -p pad 8" out.wav

This seems to be an issue with the lengths of the files and the segue times but I really haven't been able to figure it out and am really in need of some help.

Thanks!

-Will


On Jul 22, 2010, at 12:12 AM, William Kerber wrote:

> When mixing the files without the trim I get this error:
> 
> /usr/local/bin/sox FAIL delay: drained asymmetrically!
> /usr/local/bin/sox FAIL delay: drained asymmetrically!
> 
> I googled and didn't find much of anything of what that means.
> 
> Thanks again.
> 
> -Will
> On Jul 21, 2010, at 11:00 PM, William Kerber wrote:
> 
>> I am in a huge mess!  Any help is appreciated!
>> 
>> I am trying to combine 3 audio files.  
>> 
>> file1
>> file2
>> file3
>> 
>> My command to sox is as follows:
>> 
>> sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358
>> 
>> This should have file 1 playing, then at 3.306 start file 2, then at 7.203 start file 3.  For some reason this command works on other audio I have, but doesn't work right on these files.  The pieces cometogether but instead of piece 3 starting at 7.203 it seems to start around ~6.125 (I can't tell for sure where it starts because it is mixed, but it isn't at 7.203).
>> 
>> Please help!
>> 
>> Links to files are on my server which you can download and are not on this email for everyone's sake!
>> http://www.funinpa.com/file1.wav
>> http://www.funinpa.com/file2.wav
>> http://www.funinpa.com/file3.wav
>> 
>> Thanks again!
>> 
>> -William
>> ------------------------------------------------------------------------------
>> This SF.net email is sponsored by Sprint
>> What will you do first with EVO, the first 4G phone?
>> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>> _______________________________________________
>> Sox-users mailing list
>> Sox...@li...
>> https://lists.sourceforge.net/lists/listinfo/sox-users
> 
> 
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

When mixing the files without the trim I get this error:

/usr/local/bin/sox FAIL delay: drained asymmetrically!
/usr/local/bin/sox FAIL delay: drained asymmetrically!

I googled and didn't find much of anything of what that means.

Thanks again.

-Will
On Jul 21, 2010, at 11:00 PM, William Kerber wrote:

> I am in a huge mess!  Any help is appreciated!
> 
> I am trying to combine 3 audio files.  
> 
> file1
> file2
> file3
> 
> My command to sox is as follows:
> 
> sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358
> 
> This should have file 1 playing, then at 3.306 start file 2, then at 7.203 start file 3.  For some reason this command works on other audio I have, but doesn't work right on these files.  The pieces cometogether but instead of piece 3 starting at 7.203 it seems to start around ~6.125 (I can't tell for sure where it starts because it is mixed, but it isn't at 7.203).
> 
> Please help!
> 
> Links to files are on my server which you can download and are not on this email for everyone's sake!
> http://www.funinpa.com/file1.wav
> http://www.funinpa.com/file2.wav
> http://www.funinpa.com/file3.wav
> 
> Thanks again!
> 
> -William
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

I am in a huge mess!  Any help is appreciated!

I am trying to combine 3 audio files.  

file1
file2
file3

My command to sox is as follows:

sox -M file2.wav file3.wav file1.wav file1-file2-file3.wav delay 3.306 3.306 7.203 7.203 remix 1,3,5p0 2,4,6p0 trim 0 11.358

This should have file 1 playing, then at 3.306 start file 2, then at 7.203 start file 3.  For some reason this command works on other audio I have, but doesn't work right on these files.  The pieces cometogether but instead of piece 3 starting at 7.203 it seems to start around ~6.125 (I can't tell for sure where it starts because it is mixed, but it isn't at 7.203).

Please help!

Links to files are on my server which you can download and are not on this email for everyone's sake!
http://www.funinpa.com/file1.wav
http://www.funinpa.com/file2.wav
http://www.funinpa.com/file3.wav

Thanks again!

-William

>
> i am in need a means to 'correct audio' for  video that has been inverse
> telecine.(ivtc)
> basically i capure video at 29.97 - i can sucessfuly remove the 3 2 pull
> down making it the or (film) 24 fps ( 23.98 fps)
> but now the audio is out of sync.
>
> the video ( after some time like 10 min ) is ahead of the audio. I think
> this is because the ivtc basically throus away frames
>

How did you perform the ivtc operation? What kind of file are you working
with?

Adjusting the audio data is not the correct solution. ivtc should not affect
the length or playback speed of the media, so the audio data should not need
to change.

The correct solution is to adjust the video data to properly account for the
altered frame rate. If you remove frames but keep the frame rate at 29.97
fps, the video will play back too quickly and the audio will be out-of-sync.
This doesn't mean that the audio track needs to be corrected. Instead, it
means that the video track needs to be corrected so that the video player
uses a frame rate of 23.98 fps.

That said, there may be some other problem. Audio/video sync is easy to
screw up, especially with AVI files. You may want to look at a tool like
VirtualDub that can adjust the media file's frame rate to make the length of
the video track match the length of the audio track.

> Jim Shupert wrote:

> Friends,
> 
> i am in need a means to 'correct audio' for  video that has
> been inverse telecine.(ivtc)
> basically i capure video at 29.97 - i can sucessfuly remove
> the 3 2 pull down making it the or (film) 24 fps ( 23.98 fps)
> but now the audio is out of sync.

The only time I messed with that sort of thing the tool (mplayer
or ffmpeg - I don't remember which I used) was able to correct
the audio to match the video.  This would be the preferred
solution.

Did you try that and it didn't work?

> My solution might be use sox and speed up a wav and mux that
> [ speed up ]  sound with the video [ that has had IVTC
> applied }
> 
> something like
> sox -V file.wav speedup.wav speed 1.1
> 
> has anyone ever done this - am i on the right track?

It should get you close.  I have used the linux tool
"soundstretch" for similar tasks.

--         Philip

Friends,

i am in need a means to 'correct audio' for  video that has been inverse 
telecine.(ivtc)
basically i capure video at 29.97 - i can sucessfuly remove the 3 2 pull 
down making it the or (film) 24 fps ( 23.98 fps)
but now the audio is out of sync.

the video ( after some time like 10 min ) is ahead of the audio. I think 
this is because the ivtc basically throus away frames

My solution might be use sox and speed up a wav and mux that [ speed up 
]  sound with the video [ that has had IVTC applied }

something like
sox -V file.wav speedup.wav speed 1.1

has anyone ever done this - am i on the right track?

If you're building with Visual Studio 2008, I suggest using the MSVC9
project files provided in the SoX source tarball. Look in the msvc9 folder
and follow the steps in the readme.txt file. I believe they produce a build
of SoX that is better than the one you get via CMAKE+MSVC9 or Cygwin.

On Thu, Jul 8, 2010 at 5:34 AM, msaun76 <jp...@jp...> wrote:

>
> For others with a similar problem, I've compiled SoX with cygwin and added
> the cygwin DLLs to the Windows path. Although this is far from elegant, it
> works.
>
>
> msaun76 wrote:
> >
> > Documentation states that to enable the waveaudio device driver in SoX,
> > ./configure must be run with the --"with-waveaudio" option. How is this
> > done on Windows? I'm currently using cmake . and then building from
> Visual
> > Studio.
> >
> --
> View this message in context:
> http://old.nabble.com/with_waveaudio-in-Windows-tp29084243p29106547.html
>  Sent from the SoX mailing list archive at Nabble.com.
>
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users
>

What version of SoX are you running? How did you get the version of SoX that
you're currently trying to use? What is the output from running "sox -V6
input.mp3 output.wav"?

The official version of SoX 14.3.1 does not have MP3 decoding/encoding
built-in, but it can dynamically load libmad.dll and libmp3lame.dll if they
are present. However, the support for dynamically loading these is recent
(not present in 14.3.0 and earlier versions) and may not be enabled if you
built SoX yourself (requires several build configuration options to be set
up correctly).

SoX does not support wma. I started writing a plugin to allow it to support
wma (using Windows' built-in WMA codec), but I haven't finished it yet, and
it is possible that I might never finish it because I'm quite lazy. I
suggest using ffmpeg.exe to convert WMA files to a format that SoX can read.

Note: I don't recommend using the SoX ffmpeg plugin. It has several nasty
bugs, several of which are most obvious when you run it under Windows.
Instead, just have ffmpeg.exe write to a format that SoX can accept. You'll
get the same result, it will probably work a lot faster, and you'll avoid a
lot of problems.

Additional note: ffmpeg.exe is great for decoding WMA files, but don't use
it to encode them. Its WMA encoder is not very good. The Microsoft WMA
encoder is much much better.
On Thu, Jul 15, 2010 at 9:52 PM, Mohit Mittal <moh...@gm...>wrote:

> Thanks a ton for your response Philip. I will look into compiling with win
> binaries.
>
> -Mohit
>
>
> On Thu, Jul 15, 2010 at 12:31 AM, Fmiser <fm...@gm...> wrote:
>
>> > Mohit Mittal wrote:
>>
>> > I am running Windows 7 and want to convert mp3, ogg, wma etc
>> > formats to wav format. Can you please suggest how can I do it
>> > using sox command line utility. When I run "sox sample.mp3
>> > sample.wav",
>>
>> That is the correct format for the command.
>>
>> > I get that it needs libmad but even if i keep the
>> > dll for same in the same folder, it does not work.
>>
>> I'm pretty sure that for SoX to read or write an MP3 it needs to
>> be _compiled_ in.  So how it responds to an MP3 is what I would
>> expect.  I can't help with compiling MS Win binaries.
>>
>> > Other formats are also not supported, it says unknown format.
>> > How can I do it, any help in this regard will be appreciated.
>>
>> I have used SoX about 4 times in MS Windows. That was WinXP, so
>> I can't help much.  But I think ogg support is built in even in
>> the MS Win binary.  The version I have  (14.3) doesn't seem to
>> support wma.
>>
>> --    Philip
>>
>

Thanks a ton for your response Philip. I will look into compiling with win
binaries.

-Mohit

On Thu, Jul 15, 2010 at 12:31 AM, Fmiser <fm...@gm...> wrote:

> > Mohit Mittal wrote:
>
> > I am running Windows 7 and want to convert mp3, ogg, wma etc
> > formats to wav format. Can you please suggest how can I do it
> > using sox command line utility. When I run "sox sample.mp3
> > sample.wav",
>
> That is the correct format for the command.
>
> > I get that it needs libmad but even if i keep the
> > dll for same in the same folder, it does not work.
>
> I'm pretty sure that for SoX to read or write an MP3 it needs to
> be _compiled_ in.  So how it responds to an MP3 is what I would
> expect.  I can't help with compiling MS Win binaries.
>
> > Other formats are also not supported, it says unknown format.
> > How can I do it, any help in this regard will be appreciated.
>
> I have used SoX about 4 times in MS Windows. That was WinXP, so
> I can't help much.  But I think ogg support is built in even in
> the MS Win binary.  The version I have  (14.3) doesn't seem to
> support wma.
>
> --    Philip
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users
>

> Mohit Mittal wrote:

> I am running Windows 7 and want to convert mp3, ogg, wma etc
> formats to wav format. Can you please suggest how can I do it
> using sox command line utility. When I run "sox sample.mp3
> sample.wav",

That is the correct format for the command.

> I get that it needs libmad but even if i keep the
> dll for same in the same folder, it does not work.

I'm pretty sure that for SoX to read or write an MP3 it needs to
be _compiled_ in.  So how it responds to an MP3 is what I would
expect.  I can't help with compiling MS Win binaries.

> Other formats are also not supported, it says unknown format.
> How can I do it, any help in this regard will be appreciated.

I have used SoX about 4 times in MS Windows. That was WinXP, so
I can't help much.  But I think ogg support is built in even in
the MS Win binary.  The version I have  (14.3) doesn't seem to
support wma.

--    Philip

Sox is a straight input/output processor, almost like Grep.  It's single threaded so there aren't a lot of easy ways to make a single sox process faster.  BUT, like you've shown running multiple instances is incredibly easy if you have a lot of CPU power and memory.

Try running 10 instances again and check on your CPU load with the top or prstat command.  If your CPU load average is still low, you may be able to run 15 or 20 concurrent processes,


--
Chris J


On Jul 14, 2010, at 6:52 AM, Nilesh Deshpande <nld...@gm...> wrote:

> Hi,
> 
> I have a pair of wav files that need to be mixed. It takes around 35 seconds for this mixing to happen.
> 
> However, when I invoke multiple instances of SOX to mix 10 such pairs simultaneously, it takes almost 350 seconds. Th is is actually the same time taken if I were to mix the pairs one after another.
> 
> Is there something I am missing?
> 
> Any help is much appreciated.
> 
> Thanks in advance
> 
> Nilesh
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

On Jul 14 17:22:45, Nilesh Deshpande wrote:
> Hi,
> 
> I have a pair of wav files that need to be mixed. It takes around 35 seconds
> for this mixing to happen.
> 
> However, when I invoke multiple instances of SOX to mix 10 such pairs
> simultaneously, it takes almost 350 seconds. Th is is actually the same time
> taken if I were to mix the pairs one after another.

So what?

> Is there something I am missing?
> 
> Any help is much appreciated.
> 
> Thanks in advance
> 
> Nilesh

> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users

On Wed, Jul 14, 2010 at 05:22:45PM +0530, Nilesh Deshpande wrote:
> I have a pair of wav files that need to be mixed. It takes around 35 seconds
> for this mixing to happen.
> However, when I invoke multiple instances of SOX to mix 10 such pairs
> simultaneously, it takes almost 350 seconds. Th is is actually the same time
> taken if I were to mix the pairs one after another.
> Is there something I am missing?

Mixing two files you use resourses and you hit some limit, typically CPU
efficiency (could be also disk speed). When you share the same limited
resource between 10 processes the time is 10 times longer.

R.
-- 
Tą bronią zabójczą są media, które robią ludziom wodę z mózgu. I tą bronią
   macie walczyć i zwyciężać. (kpt. Z. Sulatycki do uczniów WSKSiM)

On Wed, Jul 14, 2010 at 7:52 AM, Nilesh Deshpande <nld...@gm...> wrote:
> Hi,
> I have a pair of wav files that need to be mixed. It takes around 35 seconds
> for this mixing to happen.
> However, when I invoke multiple instances of SOX to mix 10 such pairs
> simultaneously, it takes almost 350 seconds. Th is is actually the same time
> taken if I were to mix the pairs one after another.
> Is there something I am missing?
> Any help is much appreciated.
> Thanks in advance
> Nilesh

Are you on a single core non-hyperthreaded CPU?

-Pascal
-- 
Homepage (http://organact.mine.nu)
Debian GNU/Linux (http://www.debian.org)
COMunité/LACIME: École de technologie supérieure (http://www.comunite.ca)

It sounds like the process is purely I/O bound, which is typical for
mixing without re-coding.

You can verify that guess pretty quickly by running half of the
processes on your internal hard drive, and half on a working directory
on a USB drive. If I'm right that should speed it up, how much depending
on how fast the USB drive is.

On 07/14/2010 06:52 AM, Nilesh Deshpande wrote:
> Hi,
>
> I have a pair of wav files that need to be mixed. It takes around 35
> seconds for this mixing to happen.
>
> However, when I invoke multiple instances of SOX to mix 10 such pairs
> simultaneously, it takes almost 350 seconds. Th is is actually the
> same time taken if I were to mix the pairs one after another.
>
> Is there something I am missing?
>
> Any help is much appreciated.
>
> Thanks in advance
>
> Nilesh
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
>
>
> _______________________________________________
> Sox-users mailing list
> Sox...@li...
> https://lists.sourceforge.net/lists/listinfo/sox-users
>   


-- 
Daniel Taylor             VP Operations       Vocal Laboratories, Inc
dt...@vo...                                 952-941-6580x203