If it's a voice you want to level out, then use the tube-modelled compressor. This works well on voices, and is easier to set up than the Dynamics processor. It will automatically reduce the peaks and leave the lower levels alone - as will any compressor. The only control you will have to set up carefully is the Threshold level - this is the one you can adjust with the slider. This control will determine the level at which compression starts. The rest of the controls determine how much compression is applied, with the exception of the make-up gain control. This control alters the level of the whole signal to compensate for the reduction caused when you compress it.
I used it, but nothing really happened. I set the Treshold level at -4.2 which is almost at the top of the track I see. What did I do wrong?
Ok, I got it. I just used Voice leveler instead of custom.
There are a couple of things to mention here. First is that it's easiest to see what is going on if you Normalize the signal first, and put the highest peaks at about -0.1dB. This generally means that you don't have to pull down the threshold control anywhere near as much to see the effect on the red gain reduction meter.
And the red meter is the second thing, of course. You can see how much reduction you are getting on peaks very clearly with this, whether you normalize your signal first or not. This is the one thing that makes this particular compressor much easier to use - the ease with which you can set it up!