How do I use a limiter to sound balance a vo?

Rokko · Post by **Rokko** » Thu May 18, 2006 1:49 pm

Hey dudes,
I work as a video editor and often have to spend hours going through speech/voiceovers balancing the audio. I was thinking that I could maybe use a limiter to do this automatically, but I don't really know a hell of a lot about using a compressor\limiter.
All i really need to do is make sure the voice is as loud as it can be without going above a certain level. Maybe there's not a better way to do it than going through it bit by bit adjusting the volume envelope, but it would be good if there was. We use Final Cut Pro and Ableton. Anyone able to share their wisdom on this matter?

Thanks

Rokko

christhomasscog · Post by **christhomasscog** » Thu May 18, 2006 2:07 pm

Rokko,

I'm no expert (disclaimer!!

) but what you're best off doing is using a compressor set to about 4-5:1 ratio and about 20dB with a quick attack (1-10ms) and a medium release (about 250ms I think). This should give you a reasonably level sound. If you're editing the audio separately I would also recomend using a noise gate to cut down any background noise (ie camera motors or people coughing etc)
Hope that helps!
Chris

Rokko · Post by **Rokko** » Thu May 18, 2006 2:50 pm

Thanks mate!

Just tried that and it seems to work pretty well.

What should I adjust if I want to raise the quiet parts even more but keep the overall level the same?

Rokko

eisnein · Post by **eisnein** » Thu May 18, 2006 2:53 pm

hire an audio post guy?

awesome you guys use ableton for audio!

Rokko · Post by **Rokko** » Thu May 18, 2006 3:19 pm

We don't tend to do all that much in way of audio, mainly just re-editing music and balancing sound levels, so there's not much call for a dedicated audio specialist.

I'm in the process of trying to convince the company to buy their own copy of Ableton, i just installed it here using one of my licences but i want it back. So i'm trying to think of as many reasons as i can for using Ableton in a video production environment.
1. It's great for cutting the tune up so it'll fit the video better - Make a 3 min song last 9 mins without just repeating the whole thing 3 times etc.
2. Transitions between 2 bits of music. I can beatmix 2 pieces of library music or match the key of the music. Up untill I started working here the standard practice was just to crossfade between the two songs, at any random point in the tune. There's been time when 2 tunes have sounded awful together, and producers have said "why not just do a really long crossfade?" I strongly disagree!
3. If a client doesn't like a certain part of the music - like the way it ends - I can usually sort it out with ableton.
4. Sound balancing?

I've been asked to do a presentation on it, so I need to wow them with the wonders of Ableton. Should be fairly easy!
Only problem is that it's dangerous to have ableton at work, i'll probably just spend all day writing tunes.

christhomasscog · Post by **christhomasscog** » Thu May 18, 2006 3:31 pm

I don't think you'll have any probs convincing them that you need Live at work, especially since, as you say, it can be used to fit a piece of music to a limited time frame (which I should immagine you have to do a lot!)
As for pulling up the levels of the quiet bits on your audio pice, you really should normalise the original file (something Live doesn't yet do unfortunately). To do this you may find that any decent audio editor will do, eg Steinbergs Wave or even better, as it's free, Audacity which is available from http://www.sourceforge.net

M. Bréqs · Post by **M. Bréqs** » Fri May 19, 2006 1:05 am

christhomasscog wrote:Rokko,

I'm no expert (disclaimer!! ) but what you're best off doing is using a compressor set to about 4-5:1 ratio and about 20dB with a quick attack (1-10ms) and a medium release (about 250ms I think). This should give you a reasonably level sound. If you're editing the audio separately I would also recomend using a noise gate to cut down any background noise (ie camera motors or people coughing etc)
Hope that helps!
Chris

I think that a quarter second release is going to give you some audible "whooshbacks". After dudes stop talking and there's a pause, the ambient noise will suddenly woosh back in. Not cool.

What you'll need to do is as follows:

1. Get ambient noise recordings from the set where the actors did their speaking. That is to say, record "silence" on the same set for each scene filmed. If the sound guy didn't record this, then go to his house and slap him, cause now you're behind the 8 ball. That's just basic, and if he didn't do this, he just fucked you as the editor in the ass big time.

2. Use settings similar to what christhomasscog suggested - though I would use a faster release (say 80-150 ms, but use your ears) for what comes in step 3.

3. Insert a gate (or expander) effect BEFORE your compressor. Set the threshold so that it will kill all the ambient noise you don't need; there won't be wooshbacks, cause there will just be silence.

4. Now that there's absolute silence between dialog, you need to add (guess what....) Ambient noise! So, just loop your recording that the sound guy recorded in step 1 (he DID do that, right?!?!), and mix to taste. Bear in mind that you can't "set and forget" the ambient noise level - this isn't a kick on a techno track. In fact, you'll have to bring it out and in depending on the dialog on the set, and you should have a pretty complex volume envelope drawn for your arrangement.

Hope that helps.

M. Bréqs · Post by **M. Bréqs** » Fri May 19, 2006 1:09 am

M. Bréqs wrote:... In fact, you'll have to bring it out and in depending on the dialog on the set, and you should have a pretty complex volume envelope drawn for your arrangement.

Actually, some will use a subtle side chain compressor (1.2 to 1.5 ratio, with gain reduction triggered by the dialog track(s)), but since Ableton Live doesn't really support sidechaining well, that's a moot point.

FaX-01 · Post by **FaX-01** » Fri May 19, 2006 1:15 am

Rokko wrote:Hey dudes,

All i really need to do is make sure the voice is as loud as it can be without going above a certain level.
Thanks

Rokko

Import to audio editor and normalise to -1 or -2db VU .
Then use subtle compression afterwards would be the easiest way possible IMHO.
Make sure you have an audio editor that allows you to select the maximum db level of the highest amplitude peak before normalising.
In Adobe Audition for example you can normalise from 0db down to like -30db etc if you need too.

forge · Post by **forge** » Fri May 19, 2006 2:38 am

Waves will help alot - there's the RVOx plug which is made especially for that and the L2 is the bomb when it comes to maximising

How do I use a limiter to sound balance a vo?

How do I use a limiter to sound balance a vo?

Re: How do I use a limiter to sound balance a vo?