I read somewhere that one should double the vocal track -- and alternate the eq settings at full on / full off, as you move up the frequencies. The left one starts full off and the right one starts full on. That is what I did, and that may have given it too much noise. I do notice that once the song speeds up in your version -- the voices also vanish and sound odd, just like they do on mine. Maybe I should have put the raw (and unprocessed) tracks up instead of putting processed ones up. Oh well, yours sounds better than mine did. Thanks.
SoundCloud: soundcloud.com/westcrown (be nice, they are kids) -- Running Cubase 7.0.x on an MSI Gaming Laptop / Focusrite Sapphire PRO 40 / Steinberg UR824