the crest factor is the difference between the highest peak and the average rms. If you can get the crest factor right(as low as possible), using compression and saturation to achieve the right balance, you can then push the master bus limiter/s a little bit harder before distortion. Judicious use of dynamic eq before the final limiter can reduce the amount the limiter has to work, making it more transparent and allowing you to push it a little harder. Cutting below 20-30hz on bass channels or any channel that has information down there and cutting above 16-18 khz can also take away some unwanted information that can cause problems. Also chaining a couple limiters can reduce the amount of work each limiter has to do, this can make the limiting a bit more transparent without crushing the hell out of your final master. This of course takes the dynamics out of a musical piece, the loudness war continues and the first casualty of the loudness war is dynamics. A well balanced mix has a good crest factor, and a reasonably flat frequency response across the spectrum. When mastered the wav looks very healthy and thick, without the distortion. Some genres push the loudness up to peak at -0.4db and averaging around -6 to -7db. I find it pretty easy to get around -8 to -9db, beyond that it can be a bit harder to push without the distortion becoming apparent.that's just a phrase. what do you mean
and rather how do you make
Some engineers swear by soft clipping too, but I dont use much soft clipping. Maybe I should.