Have you ever noticed how much more fun it is to turn up old recordings than new ones? I know we all talk about the level wars that the CD medium ushered in, and it seems most people who voice an opinion about the subject are in opposition to excessive levels. Me? I'm on the fence. In truth, I must admit that there are aspects of squashed recordings that I really like. I can get rock tracks to grind with bus compression in a way that makes the mix a lot more powerful. It also affects the bottom end in a very cool way. Conversely, there are times when squashing a mix does not work. It gets hard-sounding and two-dimensional. Both approaches have their place, but it is choosing when to apply each one that's the tough part.
When I know an album needs to be loud, I make it loud in the mix. The idea that you can mix an album with big dynamics and then let the mastering engineer bring the level up is flawed. When you mix for dynamics and then squash all the peaks, the drums disappear, the vocals get thick and the guitars get dark. Extremely high-level recordings need to be mixed differently so that the intended balances translate.
Additionally, loud-level CDs almost always have superaggressive midrange to compensate for the closing up of the mix that results from drastic peak-limiting. When you turn it up loud, it hurts. Not like an old Fleetwood Mac record, which just gets better when you turn it up. So what does that mean? Well for me, it means that if a CD is mixed or mastered too loud, it's less enjoyable to turn it up on a stereo. One could even say that the louder a CD is mixed, the quieter the listener might listen. So if the intention is to provide a robust and loud listening experience, perhaps the best bet is a low-level CD.
I think it's also important to mention that the analog electronics in the players we use don't always like dealing with such high-level signals. They often are prone to clipping, and when they do it's not pleasant.
So when does a loud CD translate better? A restaurant is one example, and perhaps also a shopping mall. If we plan our music to be audio wallpaper, we should make it loud. In addition, if you are mixing a type of music that is supposed to be loud and squashed-sounding, then by all means make it loud. Would a super-dynamic Slipknot album really be better?
But if your target is radio or iTunes, you really needn't worry about smashing it unless you want that sound. Both of those formats will competitively level your song for you. Radio uses broadcast compression to do it, and iTunes has the Sound Check feature. You can make it loud if you want to, but I'm not sure of the benefit.
Neither approach is right or wrong. They are just different. The only dilemma that comes up for me sometimes is that it's not always easy to tell at the beginning of a mix which approach is the correct one for that particular song. Sometimes either method can produce good results.
Is it better to provide a positive experience for a passive audience that puts no effort into hearing your music, or for the people who actually go out and buy your record? Will they still buy it if they don't have a good passive experience first? Should I use an approach purely because its result is the one my client is most used to? Will they reject my mix if it's not louder than dude mixer X's? Is it better to be safe or evolve? Should I just do what I think sounds best?
I know these seem like stupid questions, but as George Massenburg used to say to me, “The only stupid question is the one you don't ask.”
Nathaniel Kunkel (studiowithoutwalls.com) is a Grammy- and Emmy Award-winning producer, engineer and mixer who has worked with Sting, James Taylor, B.B. King, Insane Clown Posse, Lyle Lovett, I-Nine and comedian Robin Williams.