I'm guessing you're eqing a compressed sample. In that case, the peak volume goes up, but the RMS volume is still going down (if you're only cutting). So the loudest parts of the track will be a little louder, but since the average volume is less, it'll sound quieter to a listener.