I have a single clip of a mother and child talking. The mother's voice overpowers her daughter's. I've surounded the child's dialog with keyframes and lifted its gain.

But I still can't get the word "creepy" loud enough (it's only a blip on the waveform). And of course, "creepy" is important to the scene. I've raised the keyframes to +12 gain. I don't have anywhere higher to move them to.

Is there some way to get this one word even louder?

Thanks much!

sounds like you need to apply compression to the track. before you do, however, what are the RMS levels? It should be around -20 dBFS

Bill,

Thanks for the reply. It's peaking around 6dB and averaging probably around 20.

To be honest, I don't understand is why there is a limit of 12dB gain increase. Is it common knowledge that adding more than 12dB adds too much noise?

I'm still pretty new to this world, so sorry if the ?'s are basic.

Thanks again.


P.S. The clip will be converted to a .wmv and most mostly viewed on computer screens.

I think the answer to my question can be found in the concept of "gain staging." I'm using automation (adujsting with keyframes) to increase the gain for the girl's dialogue. But automation seems to be more for fine to medium tuning. All of the girl's dialog is too soft. And some of the mom's dialogue is too loud.

What I should do, IMHO, is create separate audio clips for the girl's dialogue and use clip gain to boost the level. Then use automation to tweak it. The combination of clip gain and automation seems to give more than enough boost, IMHO.

Any thoughts on this approach?

Peter,

Adding compression is the way to go. Open up the Audio Suite tool. Select the audio tracks and make sure the blue line is over the clips.

Click on the fast hamburger menu of Audio Suite and choose "Compresser: 8 to1 voiceover". You can mess around with the options in there to get a good result. Since you say +12 isn't enough for you, try +16 gain with a threshold of -28db. If that doesn't do it, try increasing the gain and keeping the threshold low enough to prevent clipping (the louder voice will tend to clip as you increase gain so the threshold will control compression of the higher values).

This is way easier than trying to mess around creating separate clips for the two voices.