Hi John--
Sorry about any confusion created. I use volume automation to reduce
the peaks that would be inordinately loud _at the same time_ I EQ and
boost volume.
To use normalization in the context of mixing assumes that loudest
transient peak-- as your recorder created it-- is _ideal_ in terms of
relative volume. My experience is that that birds and amphibians
don't happen to perch at perfect distances and the insects don't
respect my "no fly mic zone" etc. Using the loudest peak to establish
overall track volume in a mix often translates into middle-ground and
background subtleties that folks can't hear unless they're using
headphones.
The reverse is true, of course. If one only wants folks to concern
themselves with the foreground sounds and there's adequate separation
in the original field recording, normalization is a quick way to
greatly increase volume without creating over-modulation.
Why not just do a quick normalization and follow that with selective
gain reduction etc and other measures? You can, and making further
changes to a file that has been "bumped-up" to 24 bits and normalized
at the same time seems to help preserve quality. I try to plan for a
minimum number of digital generations that involve processing.
There's a slight but pesky difference between real-time playback with
EQ plugs and playback after the EQ from the plug has been digitally
written. So, I usually equalize and do critical volume automation
adjustments at the same time that I create the 24 bit "Submaster"
(from a 16 bit or 24 bit field recording). I boost gain with my EQ
plugs and mixer settings making sure that the mix peak does not
exceed -.3dB. Its common for quite a few events to hover around the
peak. Does this address your question? If not, let me know. Rob D.
At 8:43 PM +0000 8/2/06, John Hartog wrote:
>Hi rob,
>
>I've been pondering your statement but still having difficulty
>understanding the difference between "normalizing" and "increasing
>the gain" as you explained. I was thinking normalizing meant
>increasing the gain to a determined level below 0. Could you (or
>someone else) explain the difference a little further?
>
>Much appreciated,
>John Hartog
>
>> I rarely use the "normalize" function. I tame the transient peaks
>> with volume automation and wave form editing and then increase the
>> gain at the same time I apply EQ to obtain a fully saturated 24 bit
>> submaster. This seems to help preserve more "body" or overall tonal
>> balance and more spatial clues where normalizing (especially with
>low
>> saturated recordings) tends to make the transients harsher while
>> suppressing the lower mid range. How else could we be sensing that
>> one of the animals is coming from "behind" were it not for cues
>from
>> reflections? To get a twig click or sniff recorded at 2am a remote
>> woodland to be heard and play "naturally" in a living room with
>> 35-40dB ambient background sound is, er,.. well,.. fun! Rob D.
>>
>
>
>
>
>
>
>
>"Microphones are not ears,
>Loudspeakers are not birds,
>A listening room is not nature."
>Klas Strandberg
>Yahoo! Groups Links
>
>
>
>
--
Rob Danielson
Peck School of the Arts
University of Wisconsin-Milwaukee
http://www.uwm.edu/~type/audio-art-tech-gallery/
|