Vent Normalization makes everyone's voice the exact same volume. In essence, you won't blow your ears out or have difficulty hearing people.

Here's how:
Go to Setup
Enable Direct Sound
Select the SFX Button
Select Compressor and click Add.
Under Compressor Properties use the following settings
Gain = Adjust for how loud you want people to be. (Try 15)
Attack = 0.01
Release = Around 500
Threshold = Around -30
Ratio = 100
Pre delay = 4.0

What the settings do:

Gain: How much volume you feed into the compressor. Not really relevant unless everyone is too quiet, then you can turn this up a little; turning it up too much will sound EXTREMELY bad.

Attack: How fast volume change will happen. Example: setting this to 500 means that anything coming in will only be dropped in volume after a half a second. For vent this should be set as low as possible.

Release: How fast the compressor stops changing things. Not really that relevant with vent, setting it around 500 is good for voice material.

Threshold: This sets the point where we actually start changing what’s coming in. Whereas 0 is the absolute loudest you can have for an input, -60 is super quiet. Average users will probably come in somewhere around -25 to -15, with the occasional few being really quiet, which is why I have this set so low.

Ratio: How much any sound below the set threshold gets compressed and/or modified in the volume department. If you still want to hear SOME volume variation then set this lower (2-4). The higher you set it the closer in volume everyone will be.

Pre delay: Mostly just deals with processing; with computer and digital processors they can look ahead a few milliseconds to see what needs to be dropped in volume or changed before it actually comes through the speakers. Highest is ideal.