naturerecordists
[Top] [All Lists]

Re: hiss reduction

Subject: Re: hiss reduction
From: Walter Knapp <>
Date: Tue, 09 Mar 2004 20:57:44 -0500
From: Dan Dugan <>

> This is the area of "restoration" software. I have quite a bit of
> experience with it, and I hate it. It isn't possible to remove hiss
> without affecting the rest of the recording. With a lot of time spent
> in trial and error, it's often possible to make an acceptable
> improvement.
>
> The problem I have is that when I'm working with the restoration
> software (DINR, Ionizer, Waves Restoration Suite) I'm paying
> attention to removing the noise. In the process of getting
> satisfaction with that, other values tend to slip away. When I listen
> to my work the next day, I often wonder what I could have been
> thinking, it sounds dead.

I certainly agree, it's really tough to remove hiss without messing
things up. Nothing teaches you to try to do better in the field quite
like spending hours trying to clean up a recording.

One thought, when you remove hiss, you remove the hiss that's under the
calls you want. That will, inevitably, change the character of the call.
  As you hear it in the original it's a combo of the call and the hiss.
The 'real call' sounds different. Do not judge hiss removal as failing
just because the call changed, judge the call itself and see if it's
still all there even if it sounds changed. Does it agree with what you
heard in the real world when you recorded it?

Even in a hissless recording, removing the quieter stuff under the calls
with a dynamics filter or such like has to be done with a extremely
light hand. A lot of the structure of a call is light sounds, not just
the heavy dominant ones.

Ideally it would be nice to have filters that we could set somehow to
just filter between the calls. I've sometimes done that manually with
short clips, but it's a lot of work.

Also experiment with several light applications of a filter, not just
one heavy one. Often the light approach will come out better.

And always keep a unfiltered copy to fall back to. Even when you think
you have it right. And have others listen to it. When I was working on
the frog CD, I'd provide them with several choices that I could make out
differences. Very often no one else would hear any difference and would
like them all. You can overmanipulate a recording, get stuck on minute
details.

Walt





________________________________________________________________________
________________________________________________________________________

<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