I started commenting on this post about the loudness wars, and decided fuck commenting I’M GOING TO USE MY MR. MICROPHONE HERE TO SAY MY PIECE!
I do mastering for CDs and digital delivery, so I pay a lot of attention to real and apparent loudness, and that post throwsaround dB values without explaining what you’re comparing them to. Is he talking about RMS level relative to 0dB full digital scale? Is he using a single RMS calculation for the entire track? I don’t know, but I do know that it’s absurd to say that Venetian Snares averages -1.25dB RMS. That would mean that his music is nothing but a pulse-width-modulated square wave just below 0dBFS! I’m not a big Venetian Snares fan, but I know his music well enough to say that whatever methodology the guy on MusicMachinery.com is using isn’t producing meaningful numbers.
When I started mastering my own and other people’s music, I picked out 5 or 6 commercial CDS from the modern era, and did some analysis on their average RMS value, and the RMS level for their loudest passages, to come up with a general rule of thumb to use.
What I decided is that there’s a sweet spot in whole-track RMS average between -13 and -11 dB RMS below digital full scale. That’s for rock and dance music; for classical music, which demands a larger dynamic range, you have to allow for larger dynamic range or it sounds like shit.
At any rate, that -13 to -11 range has served me well over the years. It leaves some dynamics intact, but it pumps the music up to where listeners don’t have to tweak volumes to match levels with other commercially mastered CDs. You occasionally run into CDs that are well above that range, but their shitty mastering usually complements their shitty music, so I don’t listen to them. Problem solved!
After a while I got more sophisticated in how I did my mastering, or rather, I started using RME’s nifty Digicheck Utility to monitor the real time RMS level of the tracks I’m mastering. Instead of judging the RMS level of a whole track, I’d set the gain such that on the loudest sections of the track, the RMS level was bouncing between about -12 and -6 dB RMS. This means the loudest bits are loud enough, but the quiet bits are still quiet, maybe -18dB or even -24dB RMS.
I’m no Bob Clearmountain but I do get repeat business, so I can’t be all bad. Furthermore I hear music all the time mastered by guys who charge 10 times what I do and I think they sound like shit, so I feel good about the results I’m getting. Part of the reason I’m so cheap is that I change the music i receive as little as I can and still make the music sound finished. That’s the dirty little secret of mastering — if a mastering engineer is doing his job right, 95% of the time, he’s just using a brickwall limiter and monitoring the RMS level of the output. $5K for an full length CD? I’d love to get that kind of money, but I would probably still only take a few hours to do the job.
But to get back to the original topic — i.e. people talking out their asses about ‘Loudness Wars’ — I’m just saying read with skepticism. Sure, Metallica Nickleback and the Red Hot Chili Peppers released some albums with no fucking dynamic range. But really, who the fuck cares? Any music actually made for people who actually enjoy music does not get the ‘brick waveform’ treatment. You can rip CDs and look at them in an audio editor all day long and only the shittiest and most transparently commercial CDs are over-compressed.