Mixed Up About Mastering
First published in L’Amour Report
It is no secret that these days a lot of music is recorded, mixed, and mastered using computers. In fact, many DAW (Digital Audio Workstation) systems have at least some ability to do all three things. Back in the day, though, these processes were separated from each other because of the tools we used. Mastering, in particular, involved hardware that was quite different from what was used for recording and mixing. A mastering console was a lot different from a mixing console, with fewer input channels but much more elaborate EQ and processing. From the time that sessions were recorded to tape instead of directly to a disc lacquer, mastering gradually became a more specialized process, and eventually there were engineers who dedicated themselves primarily to mastering.
Today mastering continues to be a craft unto itself, and years of training and experience are necessary to qualify as a mastering engineer in a top-flight facility. The record industry relies on mastering engineers to provide the final polish to recordings before they go to the replication plant. Now that computers can be used for mastering, though, a lot of project studio owners and engineers are getting confused about the boundaries between mixing and mastering. Sometimes this confusion makes the job of a mastering engineer much harder than it should be, and a lot of mastering engineers have their own horror stories to tell. I'd like to do what I can to make their lives easier, so here goes...
Which is Which?
Mixing and mastering should not be confused with each other, and they DEFINITELY should not be done at the same time. Each process has its own purpose, and they work together better when they are done separately. It is easier to keep from confusing the two processes if we start by properly defining them.
The purpose of mixing is to process (if necessary) and combine the various tracks from the recording session. This is where the balance among different parts is decided, and it is the last step in the creative process for the individual song. The question that the mixing decisions must answer is “What does the song want?”
The purpose of mastering is to fit the various mixes together as a complete record (CD). This is where the order of songs, the pauses between songs, and the processing needed to help fit the songs together is decided. This is also where the overall sound of the finished product is determined. The question that the mastering decisions must answer is “What does the record want?”
We generally want the mastering engineer to make our songs sound better and (especially) louder. What many don't realize is that there is only just so far that you can “pump up the volume”. It is possible to process a mix so heavily that there is no place left for the mastering engineer to go. If you are going to invest in mastering, you want to get the most for your money, and the way to get the best value is to NOT try to do the mastering before the mastering gets done.
Knowing what mastering is for helps, but you should also know why you are paying someone else to do it. It can be tempting to Do It Yourself, but going to a professional (even if YOU are a professional) is generally a good idea. There are two things that you go to a mastering studio for: the equipment and the engineer.
A good mastering studio is usually better set up for mastering than a recording studio is. As much care as recording studios take with their monitoring systems, mastering rooms usually take this much further. The more detailed and powerful your processing capabilities are, the more accurate and reliable your monitoring has to be. The mastering process is generally the most exacting stage of all in creating a record, and the better mastering studios are very serious about the accuracy of what they hear. Where a recording studio has to handle many channels at once, a mastering studio is dealing with no more than 8 channels, and usually only 2 channels. Although they do not have to invest in as many channels, they may (and often do) put a lot more money into each channel. In the hands of a skilled mastering engineer, this difference definitely shows.
While top grade equipment is important to mastering, the skills of the mastering engineer are MUCH more important. He (or she) is trained to focus on sonic details that can easily be overlooked in recording and mixing, even by the best of recording engineers. The mastering engineer has, in a way, fewer things to keep track of, which leaves him better able to focus on subtle details. There is also the matter of being in touch with general industry trends and expectations, such as what is expected in a recording for radio airplay. Finally, it is always a good idea to bring a fresh set of ears and perceptions to a project. There is no one person so skilled and talented that they won't ever miss something that someone else may catch, and no seasoned professional feels threatened by the ability of one of his peers to find mistakes or problems that he didn't notice. After all, the next time the process may be going in the other direction.
You go to a mastering studio/engineer to get the benefit of their capabilities, and you will generally benefit most when you do not try to do part of their job for them. By knowing what a mastering engineer wants or needs from you, you can help him give you the best value for your money. Here, then, are a few helpful “Do's and Don'ts”:
The first rule is that any process that the mastering engineer may apply to your mix should not be done within the mix. The mastering engineer cannot apply any effect or process just to individual tracks, only to the entire mix, so anything that isn't done to the entire mix at once is your job. Once everything gets to the final mix buss, the only “effect” you should apply is the setting of the master fader (and in some cases not even that). This means that you should NOT compress the entire mix, you should not apply any kind of distortion or enhancement plugin to the overall mix, and you should not apply anything else that might be considered a “mastering” plugin or process to the overall mix. Do you have one of those nifty outboard mastering processors? If so, do NOT use it on your overall mix! If you are taking your mixes to a professional mastering studio, the engineer there probably can do a better job applying any effect you might be tempted to try than you can. Any “mastering” process you apply will leave that much less that the mastering engineer can do for you, which means you can actually prevent him from giving you the best value for your money. (Did I just repeat myself in there? Probably a good thing...)
Another reason to avoid “global” processing is that it can distort your judgement while you are mixing. More than once I have been asked to consult on a mix, and I had to start by removing the processors that the client had slapped across the mix buss. Sometimes the reaction is something like “Is that why I was having trouble with my mix?”, and sometimes the answer is “Yes.”
Naturally, if you have heard somewhere that a certain process helped the sound of a mix, it is tempting to just throw it on right from the start. The reality here is that there is NO “magic bullet”, no process, plugin, or method that can substitute for judgement and skill in a mix. When a process helps a mix, it is not just because the process was applied, but because it was applied correctly. Even the best plugin or process, applied badly, will hurt rather than help a mix, so don't just grab something that looks cool and “throw it on there”.
If there is a specific part or instrument that does not “sit right” in the mix, fix it by adjusting or processing that specific part or track, not by applying some overall process to the entire mix. Yes, a mastering engineer will sometimes process a mix to “bring out” a specific part or sound, but he does that only because he cannot go back inside the mix and fix only that sound. He does not have the option of altering the settings within the mix itself, so he works with the tools he has available.
The second rule is that what the song needs and what the marketers (and the radio stations) demand are not always the same thing. That is why the “single”, which was designed for radio airplay, was once often a shorter version then the “album cut” of a song. Radio stations were more likely to play short songs, even if it meant cutting out some of the better parts. As a result, instrumental solos were often severely shortened, for example. For some types of music, shortness is still desirable for airplay, with about 3 minutes being considered the optimum length. This does not mean that the songs in the CD you sell should all be that short, but it may mean supplying special edits for airplay.
In recent years the big issue for airplay (in some formats) is that the recording must be LOUD. This is often the most important demand made of the mastering engineer. Much of the music now sold is processed so heavily that the overall quality of the sound is actually damaged. Music that is too loud all of the time can literally make our ears tired. Have you ever been out driving with the sun glaring right into your eyes? You have to squint to keep from being blinded, and the resulting muscle tension is uncomfortable, sometimes even causing a headache. Consistently loud sound makes us sort of squint our ears, tensing muscles to fight off the noise, and sometimes even ending up with a headache. The scientific name for this effect is “listening fatigue”, and this problem is part of why some of us seldom listen to music on the radio anymore.
Of course we want to get our music on the radio so that people will want to buy our recordings, so we give the radio guys what they expect so that they will give us a chance. Once someone buys your CD, though, you want them to like it so much that they play it over and over again. If they like it that much, they are more likely to recommend it to their friends, which means more sales. Too often lately I will pick up a CD that has good songs on it, but the mastering of the recording is so agressive that after a playing or two I can't stand to listen to it anymore. I have heard similar complaint from others, so I know it's not just a case of me being the “lone nut job”!
The logical answer, of course, is to do with processing what you might do with edits. If your budget permits, have the mastering engineer run two versions of your project: one that is “jammed to the gills” for radio play, and another that is less processed that will then be manufactured for sale. That way you can keep the radio goons happy while still selling a product with quality that you can be proud of.
The third rule is, don't send the mastering house a finished audio CD to work with: instead, send them the best quality source materials from which they can assemble the CD for you.
These days most original mixes are done with at least 24 bit resolution, while the resolution of a “playable” audio CD is limited to 16 bits. A certain amount of resolution is generally lost in the mastering process, but a mastering engineer working from 24 bit source files can give you the best possible quality 16 bit finished product. If you send your material in on CD, make it a data CD with 24 bit audio (such as .wav or .aiff, etc.) files. If your original recording was done at a sample rate other than 44.1k, leave the sample rate conversion to the mastering house, as they will usually have sample rate conversion tools that are most better than most of us can afford.
When possible, it is best to also leave editing and fades to the mastering guys. If a mix is a bit noisy, and the mastering will include some noise reduction, they generally need some of the “extra” stuff at the beginning or end of the file so that they can grab a noise sample, which the cleanup process then uses as a sort of template to work from. After that, they can edit out the heads and tails, leaving only the music.
For fades, it is best to try out your own fades, then carefully note the beginning and ending times for each fade and send out your mixes without the fades, but with directions about the fades. The big reason for this has to do with the processing the is done in mastering. Recordings are usually made “louder” by using some form of compressor. The harder you “hit” a compressor, the more “loudness” you can gain. If the music going into a compressor is faded out, the sound of the compression will change as the music fades. Leaving the fades to the mastering engineer allows him to compress “pre-fade”, which keeps the quality of the processing consistent even through the fadeout.
Some people will send more than one mix of a song to the mastering house to allow for changes that can happen in processing. It is not uncommon, for example, to send “vocal up” and “vocal down” versions along with the “regular” mix. Sometimes versions with the intstrumental solos louder or softer are sent. If necessary, the mastering engineer can even edit together different parts of different versions.
Always get a “test CD” copy of the finished master before having your CD's replicated, and make sure that you listen carefully all the way through to make sure that there are no problems. Leave yourself plenty of time before your delivery deadline so that you can make sure that you are comfortable with the sound of the finished product.
Finally, never be afraid to ask questions, especially before sending your material for mastering. Most mastering engineers would rather be asked several “stupid” questions than be sent one mix with stupid mistakes in it. Some mastering engineers will have their own personal preferences for how you prepare and document your materials, and they will gladly let you know about these if you ask.
By the way, even if you want to do your own mastering, it is best to keep the mastering process separate from mixing. Do not master work that you have mixed on the same day if you can help it. Approaching the material with fresh ears will help you focus better on the details that you need to listen for. If, in the process of mastering, you discover that you need to go back to the mix and make changes, that's OK too, as long as you don't try to do both things at once or jump back and forth between processes without resting your ears a bit.
There are probably other potentially important things that I didn't think to cover here, and I did not try to cover all possible situations, but I have tried to hit some of the high points. I hope you will find these ideas helpful. Just remember not to confuse the processes of mixing and mastering. Give each process its own space so that you can get the best results out of both.