Finding most similar sounds

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Finding most similar sounds

John Vonachen
Given a set of short sounds is it possible to use SoX to compare and find most similar sounds?

--
rDch0FkiKFXdpHLLyDHVsTXVGdBAIHIUbjJG9mlYYCfMWBC0CSilnxRbxASWwc
My final thoughts on Taoism and the philosophy of Wittgenstein:
Beliefs are made of words but the world is not.


------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Ulrich Klauer-2
John Vonachen wrote:

> Given a set of short sounds is it possible to use SoX to compare and  
> find most similar sounds?

I don't think so. The only effect that does such a similarity analysis  
is splice (for finding the optimal splicing point), but I don't see  
how to (ab)use that for your case. Probably best to use Octave, or  
write a special-purpose program that cross correlates sound pairs.

Ulrich


------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Peter Shute
I'd like something similar. SoX can produce spectrograms. I wonder if there are any image processing programs that could compare the spectrograms of both tracks.

> -----Original Message-----
> From: Ulrich Klauer [mailto:[hidden email]]
> Sent: Wednesday, 31 July 2013 2:11 AM
> To: [hidden email]
> Subject: Re: [SoX-users] Finding most similar sounds
>
> John Vonachen wrote:
>
> > Given a set of short sounds is it possible to use SoX to
> compare and
> > find most similar sounds?
>
> I don't think so. The only effect that does such a similarity
> analysis is splice (for finding the optimal splicing point),
> but I don't see how to (ab)use that for your case. Probably
> best to use Octave, or write a special-purpose program that
> cross correlates sound pairs.
------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Rafal Maszkowski
On Wed, Jul 31, 2013 at 07:07:06AM +1000, Peter Shute wrote:

> I'd like something similar. SoX can produce spectrograms. I wonder if there are any image processing programs that could compare the spectrograms of both tracks.
> > From: Ulrich Klauer [mailto:[hidden email]]
> > John Vonachen wrote:
> > > Given a set of short sounds is it possible to use SoX to
> > compare and
> > > find most similar sounds?
> > I don't think so. The only effect that does such a similarity
> > analysis is splice (for finding the optimal splicing point),
> > but I don't see how to (ab)use that for your case. Probably
> > best to use Octave, or write a special-purpose program that
> > cross correlates sound pairs.

I hope to use MPEG-7 (there are other methods too) for this purpose some
day. Extracting MPEG-7 characteristics of sounds isn't difficult and it
could be implemented in SoX but comparing them is another important and
more difficult part of the task.

R.
--
Жить надо так, чтобы тебя помнили и сволочи. (Фаина Георгиевна Раневская)

------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Nick Hughes
On Wed, Jul 31, 2013 at 7:51 AM, Rafal Maszkowski <[hidden email]> wrote:
On Wed, Jul 31, 2013 at 07:07:06AM +1000, Peter Shute wrote:
> I'd like something similar. SoX can produce spectrograms. I wonder if there are any image processing programs that could compare the spectrograms of both tracks.
> > From: Ulrich Klauer [mailto:[hidden email]]
> > John Vonachen wrote:
> > > Given a set of short sounds is it possible to use SoX to
> > compare and
> > > find most similar sounds?
> > I don't think so. The only effect that does such a similarity
> > analysis is splice (for finding the optimal splicing point),
> > but I don't see how to (ab)use that for your case. Probably
> > best to use Octave, or write a special-purpose program that
> > cross correlates sound pairs.
 
I have been playing with the Sound Visualiser (open source) to see if there is any relationship between the last few seconds of track 0 to the first few seconds of track 1 which may suggest that the two tracks are contiguous.  For instance, the second half of The Beatles Abbey Road album has contiguous "gapless" tracks.  I'm hoping to detect such tracks with audio analysis and additional information from the original CD (each audio file represents a track from a CD, unfortunately may not include a TOC with gap info).  This may have some similar goals as a spectrogram is involved.

So far ideas are to use knowledge that each audio file represented a real track on the CD and the following:

 o Use RMS volume between samples, variance should be minimal
 o Spectrum analysis between samples
    + Peak Bins between samples
 o Use Melodic analysis between samples 
    + dB for certain percentage of frequencies line up between samples
    + Phase?
 
Here are some screenshots which displays Track 10 and 11 joined together.  These two tracks play from a CD contiguously without gap from 2:26:26 position (cursor is there in screenshots).  Thoughts on this are appreciated.



------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Peter Shute
In reply to this post by Rafal Maszkowski
> -----Original Message-----
> From: Rafal Maszkowski [mailto:[hidden email]]
> Sent: Wednesday, 31 July 2013 9:52 PM
> To: [hidden email]
> Subject: Re: [SoX-users] Finding most similar sounds
>
> On Wed, Jul 31, 2013 at 07:07:06AM +1000, Peter Shute wrote:
> > I'd like something similar. SoX can produce spectrograms. I
> wonder if there are any image processing programs that could
> compare the spectrograms of both tracks.
> > > From: Ulrich Klauer [mailto:[hidden email]] John Vonachen wrote:
> > > > Given a set of short sounds is it possible to use SoX to
> > > compare and
> > > > find most similar sounds?
> > > I don't think so. The only effect that does such a similarity
> > > analysis is splice (for finding the optimal splicing
> point), but I
> > > don't see how to (ab)use that for your case. Probably best to use
> > > Octave, or write a special-purpose program that cross correlates
> > > sound pairs.
>
> I hope to use MPEG-7 (there are other methods too) for this
> purpose some day. Extracting MPEG-7 characteristics of sounds
> isn't difficult and it could be implemented in SoX but
> comparing them is another important and more difficult part
> of the task.

Rafal, isn't MPEG-7 just a standard for describing track contents with time code? It's not going to help find similar sounds unless someone else has already added the descriptions. I hadn't heard of it before. Is it in use anywhere yet? It sounds like it would be very useful for adding event descriptions in nature recordings.

John, depending on what these recording are, one of the bioacoustics analysis programs might help you. Raven is one, but I don't think the free lite version will do these comparisons. I think the paid version is limited to detecting sounds that have similar frequency, time and energy ranges, and that it's intended to help a researcher by shortlisting similar events for manual checking, not for finding precise matches. I might be wrong about that.

There are others, but as far as I can tell, all cost money. I'd appreciate if someone could prove me wrong about that.

There's a list of bioacoustics analysis software here:
http://zeeman.ehc.edu/envs/Hopp/sound.html

I've been meaning to have a play with this one:
http://www.soundid.net/

Peter Shute
------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users
Reply | Threaded
Open this post in threaded view
|

Re: Finding most similar sounds

Jan Stary
In reply to this post by John Vonachen
On Jul 29 16:25:38, [hidden email] wrote:
> Given a set of short sounds is it possible to use SoX to compare and find
> most similar sounds?

What exactly do you mean by "most similar"?
Is 10 seconds of a 400Hz sine wave
more similar to 11 seconds of a 400Hz sine wave
or to 10 seconds of a 410 sine wave?


------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
Sox-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/sox-users