The 12-Step Mastering Program

They say mastering is a black art, a discipline that not everyone can learn. But if you are just now trying your hand at it, or even if you are a seasoned vet, there are a couple universal rules that, when abided by, will make all the difference in your end product. So, like Letterman, let’s give a quick countdown.


Perhaps the hallmark of a great mixing engineer, like a Serban Ghenea, Pat Vialla, or Chris Trevett, is the lack of squashed mixes they turn out. As a mastering engineer, it’s important to get mixes that breathe, that have some headroom. We can take care of things better than any single compressor or plug-in any mixing engineer has in their studio, so tell them loudly — scream if you must — “Don’t kill the dynamics!” They’ve worked hard on their mixes, so they shouldn’t go ruining them; they should let someone else ruin them instead, so you can displace the blame later.

Tell them not to maximize; it can lead to nasty distortion. If they are unsure about whether distortion is imminent, they can always try bringing the mix up on two channels, flipping the phase on one channel, and mono their monitor section. They will now hear only the stereo information. If they hear crunch, and chorus vocals start breaking up, particularly on kick drum hits, then you’re printing too hot. Also, tell them to look at their meters! Are they continuously in the red? Then they are printing too hot. Is their volume knob on five, but deep red fluid is pouring out of their ears? They are printing too hot! Stop printing too hot! Tell them to let you make it loud for them.


A 24-bit DAT, when used with high quality converters in front of it, sounds great. The headroom is much greater in comparison to a print back into Pro Tools. I also prefer it to 1/2" most of the time, as there are just too many variables with 1/2" to trust putting final masters on it, unless you are really in a perfect environment. Because many recording studios’ tape machines are not well maintained due to lack of use (many technicians don’t understand bias well and much of the tape produced today is of questionable quality), I highly recommend not printing to 1/2", regardless of what many purists may say.

Digital file formats offer both consistency in sound and flexibility through the use of printed stems. A good mixing engineer working in Pro Tools will create a Pro Tools session for each song that contains both the final stereo passes and the mix stems — as well as organize the final masters, Pro Tools sessions, and mix passes so as to make the mastering process both easier and cost-effective.


I know it sounds crazy, but sometimes the mixing engineer nails it. A mastering engineer’s job is to make a song sound its best, not to recklessly step on a mixing engineer’s toes just to par the course. As an example, “Do What it Do,” on Jamie Foxx’s album (a Serban Ghenea mix) and a couple of songs on Lyfe Jennings album (mixed by Rich Keller), only needed a simple gain increase and a tiny bit of top end shelving at 12kHz. Otherwise, it was best just left alone. It’s a life lesson that applies to mastering as much as anything else: If it’s not broken, don’t break it.


A great console is key. My Dangerous Master analog console allows me to use one insert in sum and difference mode. This allows me to EQ the mono material independently from the stereo information. Therefore if a lead vocal needs to be louder I can boost some 2kHz to 5 kHz on the left channel of my Prism EQ and only affect my center/mono audio where the lead typically sits without affecting chorus vocals, which typically sit in the Left/Right stereo information. Conversely, I can make chorus vocals sparkle without over EQing the lead vocal. Working this way provides significant flexibility, but you must always be conscious not to disrupt the overall balance of the sound field.


You can increase the overall depth, width, and definition of your digital audio path simply by using a high quality external Word Clock generator to sync all of your digital devices. Try it; you’ll see. An Apogee AD/DA converter sounds like two different units when A/B’d with and without external word clock. Jitter is real! While using shorter AES cables also help reduce jitter to a certain degree, the real trick is to sync.


My file playback device is Wavelab with a Lynx 2 soundcard that is externally word clocked. Why Wavelab? It’s virtually an audio toolbox add kung fu grip; it recognizes almost every type of file and accepts Direct X and VST plug-ins. Plug-ins, you say, with a sneer? Yes. The plug-in can be your friend; there are some real excellent ones out there, whether you are looking to spend $300 or $3,000. The beauty of plug-ins is tied to the virtues of working in the box. Computers can read program material before it processes it, allowing compressors to “see ahead” and let kick drum transients to pass and not get squashed — something analog compressors have a hard time achieving. So the trick is to apply subtle compression via a plug-in before you hit your analog chain, with the keyword being “subtle.”


My compression approach is a little different than most engineers — I like to add little bits of compression three or four times in the signal path, and I suggest you try doing the same. I first apply compression via a plug-in for the reasons listed above, second that by going into the analog domain with an API 2500 or old NTP 179-120 to get that sought after character (this is when I recommend applying your EQ as well), and then touch –1dB with a Waves hardware L2 in the digital world, at the very end of the signal path, before going into my DAW. Any one of these used alone, excessively, destroys musical transients, in my opinion. The cumulative effect of a little bit of each is loud and clear — this is how you can achieve the volume many clients want today without losing musical transients, and it also reduces the amount of EQ work you have to do.

This is the best way to preserve headroom when mastering and, regardless of what trends arose in the Loudness Wars, that’s precisely what a good mastering engineer should do. The soft spots are what allow the loud parts to have optimum impact. If you kill that, then you’re killing a crucial part of the listener’s experience.


Before you EQ anything you should raise the overall gain to the point just before it distorts, and then compress. This can eliminate half of your EQ work. Doing the reverse will only change the EQ curve when you raise the gain, and you’ll just be stuck having to re-adjust your EQ settings again and again.


Where to start? The approach changes from artist to artist and song to song, of course, but one technique I like to use when EQing bass is to boost anywhere from 80Hz to 160Hz with a shelving EQ and then apply a 32Hz highpass filter to remove any excess sub build up. This will give you that hard hitting, kick-you-in-the-chest bass, a tight sub by removing unwanted frequencies that will not only make your sub-bass sound loose and sloppy, but will also eat up some of your valuable headroom. And I recommend using digital EQs, as they are tight, precise and great for filtering. The Roots “Game Theory,” for instance, has this method applied throughout, with the addition of some aggressive 10k shelving for top, and air from an old Neumann PV76 EQ. Sometimes it takes more though — Outkast’s song “BOB” required more than +9dB at 90Hz shelving to get some real bounce out of it.

#3 . . . OR, BABY, BE GENTLE

On the flip side, the recent Waltz for Venus and Acquiesce records I mastered for Mark Owen required very little EQ, and almost no compression at all. When I first asked what compressors he used, the reply was astonishing: “None!” It was all slammed to tape during tracking — total old school. In that case, I simply added some soft 10kHz shelving to the center and sides to sparkle up the guitars, and some 1.4kHz in the center to bring the lead vocals out in the mix.


One EQ setting may not be appropriate for the entire song. To illustrate this, the Roots song “Take It There” takes a sharp left hand turn half way through. Jason Goldstein’s mix of the song was dope the whole way through. My mastering settings, however, were not. Come the second half of the song those settings were downright unjust. I took the extra 15 minutes and equalized the second half separately, then spliced the two halves together in Sonic Solutions. Perfect. Doing otherwise would have just been lazy mastering, which is too much of a trend these days, something that’s done by “seasoned” professionals who are . . . well . . . maybe a little too “seasoned.” It’s called “HMO Mastering”: A client comes in to get better, and gets rushed out the back door sounding worse. So care enough to take the extra time to do it right. Your name is on the record too.


Phil Ramone posed this question on a vocal session once. The engineer spent so much time obsessing over tape levels and the compressor’s meters that Phil finally stopped the session and yelled, “I don’t care about the meters! How does it sound? It sounds fine, so let’s start recording!” I was 19 at the time and have never forgotten that moment. Don’t worry about what the meters, waveforms or the oscilloscope looks like. Remember, we must master with our ears, not our eyes. Just because it’s not a clear science doesn’t mean that it’s a black art.

Dave Kutch is a mastering engineer at Song Music Studios. His credits include The Roots, DMX, Alicia Keys, and Outkast. Drop him a line at