10 Questions About Mastering Your Recordings

Mastering is a crucial process, but it's not always all that well understood by the average musician…so let's deal with some of the basic issues.

By Craig Anderton

Your tunes are done, and you've decided it's time to create a CD — which brings you to the subject of mastering, where all the tunes are assembled and optimized for the best possible sound. You really don't want to make any mistakes at this crucial stage.

Indeed. Mastering can make or break a record, so there's a lot of interest in doing it right. Here are the ten most common questions I hear from people who are about to get their work mastered.

Q. What's the best piece of gear for giving me a professional, "radio-ready" sound?

A. The best piece of gear is a professional mastering engineer who has done this process before for hundreds, if not thousands, of recordings.

Q. So do I just send an audio CD with all the cuts, and the engineer masters them?

A. That's one option, but certainly not the most desirable. Although you should always check with the engineer for specific requirements, if you recorded your music in high-resolution audio, then it's best to provide those high-resolution mixes, as WAV or AIFF files. The mastering engineer will likely do some processing, and 24-bit files give more "calculational headroom."

Q. Wouldn't it be better to send a dithered version of the 24-bit files, as the files are going to end up as 16 bits on a CD anyway?

A. No. Dither is always applied as the very last stage of mastering, when the higher resolution signal gets downsized to the 16 bits required by Red Book audio.

Steinberg's Wavelab includes excellent dithering options, but don't apply these if you plan to hand off your file to a mastering engineer.

Q. I want a couple of cuts to crossfade into each other. Should I do the crossfading myself and send the combined cut?

A. Probably not. Fades can be dicey, and again, the mastering engineer will likely have tools that provide the best possible audio characteristics when creating fades. Also, that will insure dithering happens to the combined file — you don't want to dither two files, then crossfade them. Just make sure that you include full documentation on where you want the fade to begin and end for the two cuts.

Q. So does the same logic applies to fades in general?

A. Yes. Ideally you should send files that don't contain fades, and let them be added during the mastering process for the same reasons described above. Again, include documentation as to where you want the fades to occur, or include a version with the fades so the engineer can hear what you want.

Q. I really like being able to see the names of tunes on my CD player. How does that work?

A. CD Text is not in the official Red Book Audio CD standard, but it has become a sort of de facto standard and many programs include the ability to add it. If you want to include CD Text, provide the name for each song, and check it over carefully for typos, inconsistent capitalization, etc. Better yet, have two people look it over before sending it, because the engineer will enter exactly what you send. A group once sent me an album for mastering and one of the tunes was listed as "Who's Their?" So that's what I entered; I figured they were trying to be cute. But they meant to say "Who's There?" Proof carefully!

When you burn a CD in Magix Samplitude, you can optionally enter CD Text information for the entire album as well as individual tracks.

Q. I've heard it's a good idea not to trim off the beginning of tunes before sending them to a mastering engineer. But that doesn't make any sense. If I want a tune to start at a specific place, why shouldn't I trim it to the desired start point?

A. As mastering engineer Bob Katz says, "Editing is like whittling soap, so it's best to leave a lot more soap on and let a pro produce the very best beginning and ending for the tune. Avoid the problems caused by an overanxious mixing engineer cutting off valuable sound, which you may regret later. Sometimes it's even better to leave room tone or low level noise between tunes; or keep the singer's intake of breath, or the delicate sound of anticipation before the downbeat. Especially for acoustic music and often for electric-based music, it is disturbing to cut off the air' prior to the downbeat. A well-equipped mastering workstation and experienced mastering engineer working in a controlled acoustic environment will know just the right speed and shape of fadeup/fadedown to use on a piece for the smoothest, most natural transition."

Another consideration involves the possible need for noise reduction. Sometimes there may be a slight hiss, hum, or other constant noise at a very low level. If the engineer can obtain a clean "sample" of this sound, it can be loaded into a noise reduction program that mathematically subtracts the noise from the track. Even if this noise is way down in level, removing it can improve the sound in a subtle way by opening up the soundstage and improving stereo separation. Don't try removing the noise yourself prior to sending in the files — as with the other gear, the odds are a pro mastering engineer will have better tools.

Q. I'm on a really tight budget. What if I do some "pre-mastering" myself, like adding a little compression or EQ? That way the mastering engineer won't have to spend as much time on it, and I can save some money.

A. Don't do anything to the raw mixes. The most difficult mastering jobs I do (and the ones which take the most time) are "salvage jobs" where someone tried to master the files, and I have to figure a way to "undo" some of the damage. A good example is if someone used standard compression instead of multiband compression, and there's pumping or breathing. There's no way to "undo" that (well, not that I know of), and if you add more compression, it will accentuate the problem even more.

Q.Well I should at least normalize the tracks, right?

A. No, don't do that either. It doesn't matter if there are level fluctuations among the various tunes, as that will be sorted out during the mastering and assembly process. The reason for hiring professional mastering engineers is you want them to do their magic, so give them the space to do so.

Furthermore, normalization adds one more stage of potentially degrading DSP. Given that you may also need to adjust levels after the mastering process, it makes no sense to adjust levels before the mastering process.

Q. I do some mastering at home, and I'm actually getting pretty good at it. What if I include an example I've done at home to give an idea of the sound I'm looking for?

A. There's no harm in doing that at all. You should definitely discuss with the engineer what you expect. For example, I worked with one artist who wanted his CD to be really, really loud, as is the fashion these days. But his recordings were very open, with good dynamics, and I hated to throw that away. Besides, many listeners want to hear dynamics, because having contrast among the various sections makes for a more satisfying listening experience.

It's also important to remember that overcompressed recordings sound much worse when passed through the processing done by radio stations, because this compresses the sound even further. So we compromised, and I found a "sweet spot" between making something loud enough to satisfy him, but dynamic enough to sound good over the radio and provide a good listening experience.

Acknowledgement: Thanks to mastering engineer Bob Katz of Digital Domain (www.digido.com), and author of the book "Mastering Audio: The Art And the Science" (Focal Press, ISBN 0-240-80545-3), for giving this article a reality check.