Audio Peaking Questions :::

Gonzoman wrote on 4/24/2005, 9:26 PM
So I have this piece that I'm working on for a friend.....behind the scenes type thing. The camera man's voice is so much louder than everyone elses. Everytime he speaks...my audio peaks out - sometimes with a little distortion. On some of the footage the volume isn' t loud enough. I normalized all the audio to bring up the lower volumed stuff but I still need to deal with the "peaking out" stuff. What's the best way to tackle the peaking out stuff?

I was thinking about doing a volume envelope and just going thru and lowering the volume everytime I see clipping - is this the best way? Hopefully there's a plugin or something that can do this for me?

Thanks,

Comments

PeterWright wrote on 4/24/2005, 9:58 PM
Since the camera op's voice may need other treatment in addition to level, I would split either side of those parts and move them to a separate track. Then any settings you change, whether volume, EQ etc, will apply to all the cameraman's voice sections.

In future, avoid using the on-camera mic or tell the camera op to either keep quiet or speak gently!

Incidentally, normalizing in this situation would not be verty effective. It increases the loudest sounds to max and everything else comes up by that amount. If there are already parts near max, nothing will change much if at all.
Gonzoman wrote on 4/24/2005, 11:23 PM
Hey Peter - thanks for your thoughts. Here's something I just did and it appears to work - what do you think?

As a test....I imported some audio into Sound Forge - Once I highlighted a section of the audio that was peaking - I would go under "process" and click on "volume" where I could lower the volume of only that small piece of audio. After I did that to the rest of the peaking parts of that piece of audio - I then normalized it. Now it sounds much fuller and richer and with no peaking. Is there any reason that this process would not work?

If this would work, I may just render out the whole audio track as a wave file - import into SF - clean it up and enhance it - render it back out and throw it back up on the timeline.

Thoughts?

Thanks-
PeterWright wrote on 4/25/2005, 12:10 AM
By highlighting in turn all the peaking sections, in effect you are doing something similar - once they are reduced then Normalization has some room to move ...

As I said, voices which are a little near the mic, like the camera operator's, can be recorded with more lower frequencies than further off voices, and that is why it can be helpful to have them all on their own track, so that Track EQ can be applied to the lot at once.