The Elements of Home

Studio Audio Mastering

How to use software processors to finalize your mix

Though the picture shows Logic setup for mastering, you can achieve this with Cubase, Sonar, Performer or others just as easily.

E veryone wants the best sound quality. That is perhaps one of the only statements we can make about "everyone" involved in music production that always holds true. Think about all the issues we have covered: 16 vs. 24 bit, digital audio converters, analog vs. digital, which monitors? mics? which sequencer? Its all about the sound. For many of us, it is akin to a Search for the Holy Grail. How do we get that million dollar sound? You should know by now that your material must be good. You can polish a piece of audio garbage. You can add sugar to it, put it in a shiny cellophane wrapper, but hey, don't ask me to listen to it! :) You know that to capture good sound you have to have decent mics and preamps and a clean recording chain. You know you have to mix your tracks in a way to make the project sound listenable and enjoyable. Now lets go to the final stage, and it is here where you need to pay attention to mastering processors like those in the UAD-1 from universal audio, Waves bundles, and Izotope's Ozone.



Let me say up front that I am not a mastering engineer. I am a composer and sound developer, and as the latter, I try every tool I have and tweak it to find new sounds. That's basically what a tweakhead does :) But as I have played with hardware and software processors, finalizing everything form tiny samples that last a tenth of a second to my own epic pieces, I have learned a few things.



Mastering, or finalizing, is the last stage of the process of making audio. It is the final high resolution version of the production, the one from which you will spin off red book copies for cd and mp3 files for the internet. When you have a great, not just a good, song you hear the advice to get it mastered professionally. Dude, do it. They have the gear, a specially treated room, and excellent monitors that most of us could not afford. Plus it is always a good idea to have another set of ears that can listen to the piece more objectively.

But for many of us who have yet to discover our magnum opus, we may want to try our own hand at it, to make the cd for friends sound better, or to make the demos we hand out sound great. Or you may be a working towards becoming a sound designer, or building material for radio shows, or as an indie film composer who has to produces volumes of material so fast that mastering is out of the question. Thanks to developments in plugin technology over the past few years, we can now turn our computers into home mastering labs. While the result will not match that of an experienced mastering engineer with tens of thousands of dollars worth of hardware, with practice and a good ear you will be able to dramatically improve your production.



Getting Started

You may already be doing your own form of home mastering. Are you adding compressors, EQs, and limiters after your mix? Then you are.

Develop 2 listening environments. The first is obviously in your studio room where your computer is. The second should be in a good sized comfortable room where the speakers are farther away, a living room, with its average home audio components, can work. This gives you a reference outside your studio. Speakers that are 8-10 feet away and interacting with the room will sound immensely different than those in your control room that are 4-5 feet away. One goal is to get the audio sounding good on both.



Software file formats

Your mix should have ended with an uncompressed stereo file, ideally at high resolution, without any dither. Use 24 bit depth, and if you can during the rendering of the mix, go to a 88.2 or 96 kHz sample rate (or higher if your gear and software supports it). This lets you start with a high quality format which we will retain for the final master. You might consider using the same software in some circumstances. If you did the mix in Cubase, you could master there too. Same for Logic and Sonar. If you are going to use a different application, like Sound forge on the PC or Sound track pro or Peak on the Mac, just make sure the mix's file format is compatible. No modern application should have trouble with a 24/96 .wav or .aif file.

Most software will let you work in similar ways. Essentially you have a mixer strip for the file and an output strip for the master out. On the output strip is where your basic processors go, chained in a series in a definite order.

Note the meters from Logic illustrating the basic mastering setup. I made this simple for the sake of illustrating the basic concept. The file plays through its mixer channel without any processors here. You could add them, but you risk overloading that channel, which you don't want to do. The output channel strip is where the processors go, and they always end with a limiter of the "brick wall" type. This kind of limiter will not let audio pass into the red no matter how hard it is pushed. Indeed, this allows you to get your tracks up to commercial volume levels.

Loudness

Note in the example, the mix is down significantly, nowhere near the top of the scale. It peaked at .09, but its average level is well below. Note how after going through the 3 plugins the audio is slamming against the ceiling at 0.1, and if I wanted I could set the limiter to -0.05. There is your loudness. If I stick that in my CD player I am as loud as any commercial cd.

Not so easy. The example in reality sounds like garbage, because I just flattened all the dynamics out of the mix First we see the audio file before limiting.

And now we see it after limiting. Look at how squashed it is. But none of that material ever overshoots 0db, so it technically has no errors.



That's an example of what not to do. Of course we know that is the first thing you will try with your new software limiter, so go ahead, do it till you get sick of that sound. With a limiter you can have your audio end up anywhere between the two extremes, its really a matter of how much you dial in. You actually make a decision here, based on everything you know about what sounds good.

Below you see the UAD Precision Limiter, in my opinion, the best one out there in the software realm that a home studio owner may aspire to. This is a very conservative setting from one of my actual masters at a random point in time. Notice I increased the gain by 3db, which makes the whole mix 3db louder. I have the output set to -0.10 so it will never go over that. You see the gain reduction meter pegged at 0.50 which means I would have gone over by that much without the limiter. At that setting, the audio is indistinguishable from the mix, its just louder and only a tiny bit of the dynamics was lost. Yet that is another extreme. For most pop type songs you want to end up between the two and you choose by ear (not by waveform display). For a classical piece you want to be more conservative. For death metal you might like the sound of supreme squash. The more limiting you add, the louder the quietest parts of the mix will be--at the extreme all parts will be loud. Newbies often ask for the perfect setting for a limiter. There it is, somewhere in between the 2 extremes I presented.

Tonal Balance

But loudness is only one aspect to mastering audio. Its the easy one. Things get more complicated when we get into the tonal balance of the piece, which is effected by equalization or EQ. Here you "shape" the mix into final form and can correct problems with it that might make it unlistenable under some conditions. It here where you need to put in hours of experimentation to learn to use this tool. EQ at the mastering stage can drastically change the song's aural imprint. They allow you to select the bands you wish to modify and raise and lower the volume of those bands. The bands can be as narrow or as wide as you need them to be, from very wide gentle boosts or cuts to very narrow slices that are boosted or removed.

Buzzwords. By boosting or cutting the bands on your equalizers, you can make your sound more or less "airy" (16khz), "bright" (3-10kHz) "harsh" (which is excessive brightness) "edgy" and "brittle" (2-6k) "sweet" (a slight but wide cut at 2-8k), "warm" (slight upper bass boost and slight 4k cut). You can make your mix sound "thin" by reducing an wide band of frequencies from around 200-400 hz and make it "thick" by increasing those. If you increase it too much you'll have a "muddy" mix. Your bass can go from "missing" to "buried" to "solid", "fat", "boomy" depending on how you set the low frequency controls.

Can any EQ work? To some extent yes, but overall, for best results, you need excellent plugin eqs. Some software eqs are best for the tracking stage. At the mastering stage they will make the sound worse. Mastering EQs are usually phase compensated. Some may upsample the audio to high resolution, alter the sound, then downsample back to help prevent distortion and digital artifacts from creeping in.



The first example is the Cambridge EQ from the UAD-1 collection. Above you see a gentle curve applied to one of my mixes. Notice the bass roll-off starting at 50Hz removing subsonic frequencies. The bass then is slightly boosted for 60-80Hz then there is a slight dip @400hz. The high end starts rolling downward gently to give a smoother more ear friendly result. The signal is gently "nudged" and "shaved" into place. The Cambridge is great for those getting started, due not only to its sound, but to the great graphic display. To get started finalizing your work it's essential that you have an audio graph deeply burned in your brain and know what each band sounds like when boosted or cut. You also need to know what sounds bad at different frequencies so you can back out quick when its artifacts start to appear. Every bandwidth, boosted to extreme, sounds wretched.



Here is another example using the UAD Precision Equalizer with a focused setting. Note, in this example, I want to brighten the whole mix. Notice a fairly narrow boost at 68hz for bass then a dramatic 4db cut at 315Hz. The treble bands at 4k is barely touched and at 17k the amplitude is actually cut 1 db. Yet the mix is much brighter thanks to eliminating a whole lot (4db!) of muddy frequencies at a relatively narrow band around 315Hz. Just with that control alone I can make the mix lighter or heavier sounding. So which do you dial in? Perhaps that is the hardest part--deciding which you like the best.

Perhaps my favorite EQ is the Pultec EQP1A for the UAD. It has a pleasing sound that you notice within 10 seconds of twiddling its dials the first time. It can be used tracking too, and is a great enhancer for just about any instrument, and lovely on vocals. It has the amazing ability to both boost and cut at the same frequency, which at first may sound crazy, but has many applications. On the left you have bass controls and on the right you have treble. The attenuator dials in your roll offs for the high and low end. Here you see bass boosted at 60Hz and everything under that is rolled off. The treble is boosted at 8kHZ, something that results in harshness on other eqs--but not on the Pultec! That makes it a must-have.

Compression

Compression is another tool the mastering engineer uses to bring out the flavor of audio. Used effectively, compression can smooth out the piece. It raises the volume of the softer sounds and reduces the level of the louder ones, to make them all more uniform to the ear. Setting the attack and release of the compressor can yield a pleasing sense of dynamics that can set the whole mix in motion where all the instruments sound like they are on the beat and surging forward in the groove (even when they may not be). The loudest element of the mix that the attack segment "captures" will trigger the subsequent gain reduction. The decay will determine how long that reduction will last and the audio will rise again in volume till the next loud trigger comes through and starts the cycle all over gain. Mastering engineers tend to love compressors as each has a different sonic imprint on material.

Another type of compression used at the mastering phase is a multi-band compressor. This is a processor that works to both tonally balance the piece by breaking up the audio bandwidth into 3 or more bands and having a separate compressor for each.



The Waves Linear Phase Multiband has 5 bands that are totally user definable. This is a dangerously powerful tool. It's quite easy to ruin a mix you worked months on with it. The MB compressor can do major surgery on your mix when used with care. You can dial in different compression (or expansion) settings for each band till it does what you want it to do. These are settings I discovered on a fast trance like song to bring out more punch in the lower region and create a sharply defined edgy high end that was deeply buried in the original. Now that those frequencies are alive I can add other processors to smooth it out. The Multiband compressor usually goes early in the chain and may be followed by a smoothing eq and a rounder compressor. I could have used the low bass band to roll-off the bass, but here I wanted to accentuate a clean subwoofer "thump", so i will have to follow this with something to roll off the subsonic bass.

The Order of Processors in the Plugin Chain

There is really only one rule. The brickwall limiter has to be at the end to prevent any "overs" from occurring. Otherwise the order is determined by your goals for the piece you are mastering. Mastering engineers earn their pay not just because they know how to set up the machines, but because of their experience in assessing the overall mix's strengths and weaknesses and then being able to develop a plan of attack to remove or reduce the weaknesses while maintaining or enhancing the mix's strengths. You add a processor when there is a reason and need to do so. You might not need to add eq if the mixdown engineer had the troublesome frequencies under control, or you might have to surgically remove an irritating frequency. Rather than ask a question which has no absolute answer, why not experiment by changing the order in your plugin chain a few times. How did it work when eq was put before compression? Vice versa? Now you are gaining experience.



The End of the Session

By using your ears and your knowledge of what each processor may provide, you begin to develop strategies for making the master shine. The session ends when you have decided you have achieved this. Often that is a hard decision! When is done, "done"! Mastering audio is about listening and making decisions. Its your ability to listen carefully, knowing what tool to apply and when, and a strong internal sense of what good sound is that leads to a successful session. An interaction of ear and mind. We all want good sound. As you master your work, you have a shot a defining what that means, not only for others, but more importantly, for yourself.



May your audio shine,

Tweak

Go to the Next Class

Go to the Previous Class

This book is highly recommended

Threads

Digital vs. Analog compression and limiting - ILLUSTRATED...

Forums

The Art and Science of Mastering Audio with Massive Mastering

How does one set up their studio for mastering? What are some of the practices the mastering engineer follows? What hardware and software is needed to accomplish the task?



Resources on Mastering on the Web

Paul White's "20 Tips on Home Mastering"

Recommended Services

Links to our Mastering Engineer

Massive Mastering

And our Recording/Mix Engineer

Blue Bear Sound

More Articles by Tweak on Mixing

Mixing and Mastering

Choosing a Mixer for your Studio

Understanding your Mixer

Digital Mixers

Classic Analog Mixers

How to Hookup a Mixer

Guide to Control Surfaces

How to use EQ

How to Use a Compressor

Using Pan Controls

The Perfect Mix

Review of the UAD 4k Processors

Mixing on a Virtual Console

Tascam DM3200 Resource

Mackie 8 bus Console Resources

Elements of Mastering

Mixer catalog List