how can I best process audio track portions?

BusyBee wrote on 11/4/2006, 6:17 AM

Hello,
I have one DV file featuring 20 interviews that I have captured into Vegas7b.
Some people speak loudly and others speak softly.
I also have SoundForge.
How should I handle this most efficiently ?
Is
a) opening a copy of the sound track in Sound Forge,
b) selecting each interview one by one,
c) leveling up each interview one after the other,

the best way to proceed? Are there better alternatives?
Thanks,
Henry

Comments

DavidMcKnight wrote on 11/4/2006, 6:50 AM

That is one way to do it, although you can also get great results without leaving Vegas. What I do in this case is split the track at each interview, select the audio event you want to work with, and normalize (if that's what you want to do), and then adjust the overall volume level of the track as needed. I believe Forge normalizes on rms instead of peak, however.

John_Cline wrote on 11/4/2006, 7:39 AM

Refer to my post in this thread:

http://www.sonymediasoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=406093

MH_Stevens wrote on 11/5/2006, 6:48 AM

In Forge you can choose rms or peak.

The point here is (and it's something I find annoying) that in Vegas you can only apply audio Fx to a whole track and not to a event. So, to me it seems, that to use the Vegas audio Fx you must put every audio event on its own track!

The use of "normalization" in Vegas can cause problems, as by putting everything to a maximum before you compress etc. it causes a lot of noise.

Chienworks wrote on 11/5/2006, 1:35 PM

You can apply normalize and non-real-time audio effects to individual events. You can even apply them independantly to split sections of the same clip. And of course you can also use the gain reduction on each event separately or apply a volume envelope.

MH_Stevens wrote on 11/5/2006, 3:07 PM

I can not see how to do this. I only see the track FX button.

TorS wrote on 11/5/2006, 3:57 PM

Forget normalize. Why? Because even if you get all the events brought ut to their peak, the weaker parts of them may still be next to inaudible. Even RMS normalization is not good enough for this, although better. You need to raise the volume of the soft parts and maintain (or only slightly raise) the loud parts.
Use light compression (Graphic Dynamics) with gain compensation enabled. This you can do to a whole track (in Vegas) with many events at different levels. The loud ones will be left alone, the soft ones turned louder.
If you have Sound Forge you may also have Wave Hammer? If you do, use the volume maximizer.
All these are highly adjustable, so try them out, tweak, and try again. Use your ears and your patience.
If the recording quality is variable you may want to manipulate some events before you even out the levels. Things like filter out unwanted noise (Sony Noise Reduction is good but you can go a long way with the included EQ). When you have fine audio tools like Vegas and Sound Forge, normalize is hardly ever needed.
Tor

Chienworks wrote on 11/5/2006, 6:24 PM

MH_Stevens, right-mouse-button click on an audio event and look at the menu that pops up.

MH_Stevens wrote on 11/6/2006, 8:13 AM

Kelly - thanks for reply.

However I was meaning the real-time audio FX, the same three included in the track button. I still don't think it is possible to apply those to a single audio event.

For me, I can not use an audio FX if I can not monitor while tweaking. Am I missing something?

Michael

Chienworks wrote on 11/6/2006, 7:49 PM

You can preview the non-real-time effects while tweaking. It's just that after you're done tweaking, a new file with the effects "hard coded" is created and that replaces the orignal audio as a new take. So, it's not as easy to keep tweaking afterwards, but it can be done.