Something to tax your verbal recognition

[Guest]

Archived from groups: rec.audio.pro (More info?)

This may not be the right group for this request, but it's a question
about audio recognition and processing, so let's try. Some time ago I
made a recording of a group of men shouting nearby, which I have
uploaded to the following site;

www.MARTIANDAWN.COM/m10-t06.mp3

The 17-minute segment is 16 megabytes which will download in about
three minutes over broadband. Some of you listen to speech audio
professionally, so perhaps you could give opinions on what you can
hear from this track. My "take" is as follows;

00:39 "kurwa"
00:44 "he's watching"
00:50 "kurwa"
01:12 "I don't wanna shoot people... take pictures"
05:09 "something wrong with the guy"
05:49 "you're crazy"
07:12 "crazy"
15:04 "yes! wasn't funny"
15:23 "I am not a whore"

where mm:ss is minutes:seconds from start of recording.

The audio is unclear although some of it may be comprehensible. The
men are obviously shouting abuse, if what I hear is correct.

The other part of my question is whether it's worth running Soundforge
or other noise reduction on this segment, or whether that would only
destroy signal as well as noise. If noise reduction is feasible, what
would be the filter or settings to achieve it for this segment?

 

[Guest]

Archived from groups: rec.audio.pro (More info?)

On Sun, 05 Sep 2004 22:29:17 +0100, wep@MARTIANDAWN-removethis.COM
wrote:

>This may not be the right group for this request, but it's a question
>about audio recognition and processing, so let's try. Some time ago I
>made a recording of a group of men shouting nearby, which I have
>uploaded to the following site;
>
>www.MARTIANDAWN.COM/m10-t06.mp3
>
>The 17-minute segment is 16 megabytes which will download in about
>three minutes over broadband. Some of you listen to speech audio
>professionally, so perhaps you could give opinions on what you can
>hear from this track. My "take" is as follows;
>
>00:39 "kurwa"
>00:44 "he's watching"
>00:50 "kurwa"
>01:12 "I don't wanna shoot people... take pictures"
>05:09 "something wrong with the guy"
>05:49 "you're crazy"
>07:12 "crazy"
>15:04 "yes! wasn't funny"
>15:23 "I am not a whore"
>
>where mm:ss is minutes:seconds from start of recording.
>
>The audio is unclear although some of it may be comprehensible. The
>men are obviously shouting abuse, if what I hear is correct.
>
>The other part of my question is whether it's worth running Soundforge
>or other noise reduction on this segment, or whether that would only
>destroy signal as well as noise. If noise reduction is feasible, what
>would be the filter or settings to achieve it for this segment?


Would anyone please listen to this track and respond? Some parts of
the track eg 01:12 and 15:23 are quite clear. The material is
unpleasant but another opinion would help.