Planewalker .:.:. Waveform Wizardry

I’ve often heard it said as a rule that you should never master your own tracks. I followed that rule for awhile, but recently I’ve come to a new place with it after years of study, and I don’t agree anymore. In my opinion, if an artist can learn to mix his own tracks well, he can also learn to master them. The truth is that mixing well is a more involved process than mastering well, and if you can’t do a good job mixing your own tracks, good mastering can’t fix them. On the other hand, if a track is well-mixed, mastering it is much simpler. I’ve sent tracks to people who have mastered Pearl Jam, Alice in Chains and other really big artists and they come back with the track saying that all they had to do was “pump up the level”, which I could have done just as easily. My new philosophy is – do a really good job mixing your tracks, and mastering becomes easy as pie. The biggest names in mastering seem to agree – all you have to do to a well-mixed track is run it through a peak limiter and dither down. If you are mastering a whole album, you have to do other things like get even volume between the tracks, determine the track order, and maybe put ISRC codes into the CD, but that’s not so tough to do.

That said, it can take years to get a good feel for mastering. It is its own skill and has its own lingo. But it is learn-able. If you are just starting out, I think it’s worth it to pay $50 or so send a track from your album to a professional mastering engineer. Make sure to ask them for their mastering notes. Take a look at the difference between the mastered and un-mastered versions of your track, and then read the notes. Pay particular attention to the spectral balance and the waveform. For this you need a really hi-res spectral analyzer and oscilloscope plugin. I recommend the Spectrum device in Ableton Live version 7 and above, and the s(M)exoscope from smartelectronix. The Spectrum device is great because you can click the arrow in the title bar of the device and make it take up the entire screen, where you can zoom in on any part of the spectrum. This is critical, because a 3db difference in mastering is a BIG difference. Most spectrum analyzers cannot show you that much detail.

You also need to understand the frequency response of your monitoring environment, particularly in the sub-300hz range. For this, play a sine wave and sweep the frequency across the spectrum, noting where it is louder and where it is quieter. Most setups will have resonant frequencies at around at 80hz and 160 hz where the bass suddenly gets way louder or quieter. In a properly designed and acoustically-treated room with good monitors, the frequency sweep will be even, and the sine wave will remain at the same volume as you sweep with less than a 3db variation between frequencies. This is not easy to achieve, and you will probably notice much bigger volume variations. However, if you understand where the problem frequencies are, you can make note of them and know that the room is influencing the volume at those frequencies. Ideally you would want to smooth out those problem areas, but if you can’t, you can at least know where they are and avoid making mixing and mastering decisions by ear in those frequencies, relying instead on the spectrum analyzer.

A well-mastered track will have a slowly tapering downward spectral slope in the spectrum analyzer, starting louder on the bass side and slowly rolling off as you get farther into the high frequencies. Just run a well-mastered track though the spectrum analyzer to see what I mean. It won’t have any big spikes anywhere in the spectrum. You can use either an EQ or preferably a multiband dynamics processor to smooth out the spikes and get a more even spectrum, but if you did a good job mixing down the track and compressing/EQing the channels, you won’t need to. On the oscilloscope, the peaks of the waveform should hit zero on both the positive and negative side at the highest energy points of the track when the song is at its loudest. The peaks of quieter parts should fall somewhere between -6 and -12db. The final level itself is a matter of taste, but this is where having a reference track by a pro engineer really helps you. Try to get the track at the same level as the reference track while remaining just as clear. Use a brick-wall limiter to adjust the level. Any other effects like stereo widening or harmonic excitation are not necessary if the track was mixed well. If it wasn’t, it’s better to go back and fix it in the mix than apply such effects in the mastering.

Once you get an idea of what a well-mastered track looks like and what the appropriate level is, you can attempt to duplicate it in other mastering efforts. If you keep doing this and refining your sound, you’ll be able to master your own and other people’s tracks just fine. If you get good enough at reading your oscilloscope and spectral analyzer, you can even do a good mastering job on headphones or in environments where the acoustics are bad.

If you’re new to the whole process, but you have good compositional skills, it’s still true that you are probably better off having someone else master your tracks until you really get the hang of it. Until you do, however, it’s best to get a professional to mix your tracks before you hand them to the mastering engineer, because a badly mixed track with great mastering will sound a lot worse than a well-mixed track with great mastering. The better your mixing skills, the easier it is to master your own tracks.

Filed under: Business Tips - Music Tips
by planewalker