naturerecordists
[Top] [All Lists]

Re: Difficult Project

Subject: Re: Difficult Project
From: "Trond Lossius" tlossius
Date: Wed Apr 30, 2014 6:38 am ((PDT))
Hi,

You=92ll have to determine for yourself what factors or qualities you want =
to correlate with respect to, but there=92s lots of research being done in =
this area. Brobably you don=92t want to compare raw waveforms, but rather w=
ork on shape of envelope, and interpretions of spectral information. You mi=
ght e.g. want to look into research by Diemo Schwatz, Norbert Schnell and o=
thers at Ircam in relation to the CataRT real-time environment for realtime=
 concatenaive synthesis, and the possibilities for analysis offered by FTM,=
 Gabor and in non-real-time by AudiSculpt and SuperVP.

As far as I remember, a lot of research has also been carried out at Queen =
Mary University of London, e.g. extracting audio information from the big d=
ata of LastFM.

Best,
Trond


On 30 Apr 2014, at 11:28, Michael Dalton <> wrote:

>
> Howdy All,
>
> I am looking for an easy to use program to attempt to correlate sound fil=
es. The purpose is for a paper I am writing. Years ago, I had a graphic cor=
relation program, but it is long gone--I was disappointed that I only got f=
igures in the 0.30 range.
>
> Does anyone have experience looking at two different sound files and sear=
ching for confirmation that there is a relation between the files, which co=
uld be of different lengths, onset, etc.? The files are recorded at differe=
nt times.
>
> I am still grappling with showing that my transcriptions of a bird speaki=
ng English (that I clearly hear) are related to a recording of my voice say=
ing the same words. Any other approach that might work would be acceptable =
as well. All suggestions are appreciated.
>
> Regards,
>
> Mike
> Florida
> http://www.parrotspeech.com/
>
>







<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU