Production-212-Logo-1I guess the number one question I get from production newbies is how to get the voice track to pop without going through a lot of arduous ducking and dodging. The number two question is how to “phatten” the VO to make it sound huge, without obliterating the music and effects. Well, the two are more closely linked than you might think. Getting the voice to pop first is key to making it phat. Many regular readers will recognize this as a sort of extension of what I talked about last month: textures. This month, I’m getting more into the nuts and bolts of voiceover processing. Let’s start with getting your vocals to pop. By far, the easiest thing is to use EQ. Assuming that all your levels have been ”normalized,” pumping up the high end of the vocal will make it pop almost immediately, without ducking the music.

The reason might not seem so obvious. Most folks think that the bass in a man’s voice is low frequencies. Some of it is, but not very much. When you talk to someone on the phone, you’re not hearing any bass at all (basically nothing below 400Hz, which is just a few semitones below middle C on a piano), and yet you almost always know whether it’s a man or woman who’s speaking. The key to differentiating a male voice from female is mostly in the “glottal fry.” If you listen closely to a man’s deep voice, you will notice that you can actually hear the vocal chords slapping against each other. That’s glottal fry. The frequencies of glottal fry are up in the 600Hz range. When you hear a man’s voice with a lot of glottal fry, your brain makes the connection immediately -  “That’s a man speaking. (Psychoacoustics is what that’s called.) Going back to the phone analogy, sometimes you’re wrong about the caller’s gender. If it’s a woman who has smoked heavily for years, she’ll have a ton of glottal fry, and you just won’t be sure.

Okay, having said all that, how does it help you? Cut OFF everything below 400Hz on the VO. Pro Tools users, open the 1 band EQ module in your Audio Suite folder and click on the left-most button. That’s the high-pass button. (Everything above your designated frequency passes through.) If you’re not a Pro Tools user, just about every EQ module I’ve seen has a similar setting. Set the frequency to 400Hz and test a couple of voice tracks. They’ll sound a bit thinner at first, but that’s normal. We’ll take care of the thin-ness next. The glottal fry is still there, especially in a man’s voice. (By the way, I do this to ALL of my voiceovers.)

Step two is compression. Narrow the dynamics to 3db or so (the difference between the softest and loudest parts) with any comp-limiter, and make it as loud as you can. Done deal. This voice track will print over just about anything. In fact, you may have to lower the gain on the VO to make it blend better. Too much is always better than too little. You should also play with the input gain settings to make sure you don’t splatter the VO.

Okay then, let’s phatten up the sound. There is a method I sometimes use and one I always use. The one I always use is to run all the VO tracks through a sub-master with a delay module on it. Left channel is delayed 20msec and the right channel is 40msec. This “Stereo Masking,” as I call it, puts the voice track out-of-phase all the time. You need to reduce the MIX on the delay module to something like 20%. In other words, 80% of the signal is the original signal, 20% is the delay. You might find an even smaller MIX setting sounds better for you.

When I need to really make it sound big, I do a mechanical fix, with no plug-ins involved. Start with a mono VO channel. Duplicate the VO to another mono channel and offset it 20msec earlier than the original track. Duplicate the original track again to another mono track. Offset this one 20msec later than the original. Grab BOTH of the new VO tracks and move them to a single stereo track. Reduce the gain on the stereo track by -6db. Hit play. All of a sudden, you have the voice of GOD speaking. It’s truly PHAT!

You MUST reduce the gain on the offset tracks or your station’s engineer will scream, especially if you have a phase-chaser in the chain. It’ll totally go berserk because it won’t have a solid “center” channel to lock onto, and that would sound bad. Besides, if the stereo mask is louder than the original VO, it will be unintelligible. Not good.

If you check out my track on this month’s CD (track 3), a promo for this year’s Z100 Jingle Ball, you’ll notice that I use the delay on most of my VO. My entire VO has no frequencies below 400Hz. I only use the stereo mask when I really need to emphasize a word or phrase. I don’t use any kind of delay when I’m popping a hi-pass filter (1kHz), nor do I use any on the artist’s VO tracks because they have their own “star-power” to make them jump out.

There you have it; Z100’s VO processing in a nutshell. Once you’re set up, it’s unbelievably easy. You can make an even mediocre voice sound really PHAT, without having to spend a lot of time ducking and dodging the other tracks. You’ll still have to do some fine-tuning, but in most cases, you won’t even worry about it. Your VO will pop almost by itself. It’s almost like turbo-charging your production.