Proper levels for Dialog in vegas?

musman wrote on 2/22/2004, 3:51 AM
Sorry another question from the guy doing the make a movie in 48 hours thing. I've never had to record live dialog before and am editing it now. The levels rarely peak over midway on Vegas's master volume. I'm use to music that sounds good with my video when I just make sure it doesn't exceed 0db (go into the red).
Am I right that dialog is supposed to run at around -20 to -12db? If so what's going to happen when I lower the music to something that doesn't blow the dialog away? An audio track that's going to sound significantly lower than most others (which would be awkward when it's played with the other 48 hour movies)? I'm affraid to boost the dialog too much as I already have some noise from recording with my pd150 and ME66. Really wish sound forge and noise reduction had arrived on time, they might have helped.

Comments

farss wrote on 2/22/2004, 4:48 AM
Oh dear,
been there, done that!
To get your dialogue to sound OK you need a fair amount of compression, I've found Wave Hammer excellent BUT, whoa up comes the noise floor to smack you in the gob. Last VO I recorded I finally got a good mic and sound gear, all that was lacking was a VO booth so I used my loungeroom.

After I applied the compression I found the previously inaudible resonances were very pronounced at the end of each word, quick snip with a volume envelope fixed it but ut would have been nice not to have it there in the first place.

Sorry none of the is helping you. As you don't have NR2 try to Eq out as much noise as possible, maybe if the noise floor is low enough try the noise gate, then the compression. And heres the wierd part, in realtime the audio FX chain can seem to be working poorly yet if you render it out hey it sounds just fine.

Just to go over it again: Eq>Noise gate>Compression, which isn't the way Vegas does it by default.


Oh, and Good Luck.

To answer your other question, you will need to get the voice peaking to -3dB or close to it. But what really matters with voice is more like the RMS value, that'll give you a better idea of how much punch it has and if you've got music you also want fairly prominant and don't want to drop its level down so you can hear the voice then the voice needs a lot of punch to come through.
musman wrote on 2/22/2004, 2:25 PM
I appreciate it. I'm really puching up the audio levels to get them to approach -3 db, but it's coming at the cost of a lot of hiss. I'll try your methods now that I've done the first render and print to tape for a background.
But, are you sure it wouldn't be better to keep the dialog closer to where it is (hitting around midway on the master volume) and lower the music to match it?
Thank to anyone for the help. I have 2.5 hours to turn this in.
Chienworks wrote on 2/22/2004, 2:30 PM
You have to balance the voices to mix properly with the music. Whether you raise the voices, lower the music, or some of both, it doesn't matter. The end result is the same. If you mix it at a low volume then the people watching it will turn up the volume knob and hear the hiss anyway. If you mix it at a high volume then the people watching it will turn down the volume knob, reducing the hiss to the same level anyway.

The end result is that the hiss will be the same in relation to the finished mix no matter what you do. Personally i'd go for a higher overall level because then any steps that come after the mixing will have a hotter signal to work with and won't suffer as much additional noise due to their own S/N ratios. Keep in mind that if you produce a finished video with a low volume level then the audience will turn up the volume and this will amplify not only the audio you give them, but also the hiss and noise of their player, amplifier, TV, etc.
GlennChan wrote on 2/22/2004, 4:59 PM
Maybe your deadline is over already. Anyways.. you can try:

EQing non-dialogue frequencies out. You can work backwards (raise the volume using track EQ) to find dialogue frequencies, then low shelf and high pass off the non-dialogue frequencies. Other noise reduction techniques would also help.

WIth some music, you can EQ the important voice frequencies down slightly so dialogue punches through more. This won't always work. see http://www.dv.com/dv_login.jhtml?_requestid=131402 (registration required)
"Watch out for music interfering with dialog between 1.5 kHz and 3.5 kHz. This is where the consonants live. Dipping the music a few decibels in this range can make the mix sound smoother."