Wave vertical width & clipping

Discussion forum for Amadeus users

Moderator: Martin Hairer

Post Reply
Cherrill P. Heaton

Wave vertical width & clipping

Post by Cherrill P. Heaton »

In a sound wave, I frequently have adjoining tracks (same kind of music) that are quite different in vertical width (volume). I normalize each to 97%. Then I go hunting for Maximums in the narrower wave and trim them visually so they look more like the spikes that surround them, then re-normalize. But even after a bunch of such trimming, the newly normalized narrow wave is still quite a bit narrower than the adjoining wave. If I amplify the narrow one to, say, 110%, it reflects clipping even though the wave is narrower than the other wave with no clipping. What is my conceptual or executional error here, pls? Thx.
CP=

User avatar
Jim Edgar
Posts: 118
Joined: Mon Mar 01, 2010 7:20 am
Contact:

Wave vertical width & clipping

Post by Jim Edgar »

It depends what you are trying to do.  I can't quite tell if you are trying to get a multi-instrument mix into shape or make two different recordings that you are splicing end to end (though on different tracks) play at similar volumes. 

First - are these tracks you are recording of the same instruments, or are they different?  If they are different types of instruments, then it's probably not the best approach - in other words recording a trumpet, electric guitar and a violin will create very different timbres - the wave forms will be quite different so you can't really just go by the visuals.  They may look quite different, but that at doesn't necessarily mean they are inappropriately recorded such that you'd need to modify the amplitude of the wave (your percentage normalization action).


Also - if what you mean when you "trim them visually" is that you are highlighting the taller waves and then negatively amplifying them (reducing the height) you are basically doing a manual version of "Compression".  You might find the AU Dynamics Processor a more effective tool. 


Normalization is just an Amplification process with a specific target.  (Amplification by itself is a relative increase or decrease - a plus or minus from where you are).  If you are using the Normalize to "Maximal Sample Value" then it will find the highest wave and make that point your "97%" and keep everything else in the wave form relative as it increases or decreases the highest wave.  That means if your loudest sound was a loud pop at -0 dB and everything else was relatively quiet, using Normalize would actually reduce the volume of that track - taking the -0 dB (max loudness) and reducing it to 97% and taking everything else down with it.  


If you are trying to mix multiple tracks, having every track at high levels can peak/clip/distort the stereo mix. I'd tend to keep my individual tracks down around -6 dB or less at the initial stages.


You might find using dB is a bit more accurate.  Also, the Analyze->Waveform Statistics tool can give you a sense of the relative "power" of each track - the RMS value is such a measurement. 


- Jim
Jim Edgar

How you say it matters
jimedgarvoices.com -  @jimedgarvoices

justaskjimvo.studio - @justaskjimvo







On Sun, Aug 7, 2016 at 1:25 PM, Cherrill P. Heaton <heaton5@cox.net (heaton5@cox.net)> wrote:
In a sound wave, I frequently have adjoining tracks (same kind of music) that are quite different in vertical width (volume). I normalize each to 97%. Then I go hunting for Maximums in the narrower wave and trim them visually so they look more like the spikes that surround them, then re-normalize. But even after a bunch of such trimming, the newly normalized narrow wave is still quite a bit narrower than the adjoining wave. If I amplify the narrow one to, say, 110%, it reflects clipping even though the wave is narrower than the other wave with no clipping. What is my conceptual or executional error here, pls? Thx.
CP

Post Reply