Posted March 4, 2006, 10:20 a.m.
By Ernie Rideout
Mixing a song that you’ve lovingly composed, arranged, performed, and recorded can be a tremendously satisfying experience. But if you’re unfamiliar with the tools and techniques available for creating a mix, you might feel a bit of anxiety as you sit at your computer or in front of your mixer, staring at all those faders, knobs, and processors, wondering what to do.
Be anxious no more! You’ve come to the right place. Record U will provide you will clear explanations and practical advice for making your music sound amazing. The good news is that whether you’ve recorded your tracks in computer software or in a hardware multitrack recorder, you have all the tools you need to create everything from a rough mix to a final mix. Let’s survey these tools briefly:
- Faders: Use channel faders to set relative track levels.
- Panning: Separate tracks by placing them in the stereo field.
- EQ: Give each track its own sonic space by shaping sounds with equalization.
- Reverb: Give each track its own apparent distance from the listener by adjusting reverb levels.
- Dynamics: Smooth out peaks, eliminate background noises, and bring up the level of less-audible phrases with compression, limiting, gating, and other processes.
There are a few important things to keep in mind about all these tools:
- They can’t fix poorly recorded material
- Any change you make with one tool can affect the results of the other tools
- The tools were originally designed to do a simple job — but they’re actually full of fun creative potential, too!
In this article we’re going to focus on the tools that many musicians consider to be boring: dynamics processors. But when used intelligently and in tandem with careful listening — that’s the most important tool you have, those ears of yours— dynamics processors can turn an ordinary mix into an extraordinary one. And they’re full of tricks that you can use to make amazing sounds you never imagined.
A word of warning: In the course of this article we’ll be taking dynamics processing to extreme degrees to demonstrate the effects that are possible. For most music, the best application of dynamics processing is the most subtle and gentle. A little can go a long way. Always check your dynamics processing work by listening to the effect in the context of the entire mix. Use the “bypass” switch on the device to compare the processed effect with the non-processed sound. This will help make sure that you keep the intent of your music in mind as you mix!
Let’s take a tour of what kinds of processors fall under the “dynamics” heading. Then we’ll fiddle with the knobs!
What are dynamics?
It’s the same as in your music: Dynamics are changes in loudness for expressive purposes. This category of processing is called “dynamic” because that’s the aspect of sound it’s designed to address — relative loudness and the ways we perceive it. The operative part of that definition is “perception.” Our brains have very funny ways of processing differences in loudness. Louder is better, for example — have you ever felt that way? Most musicians have, especially when they’re in front of a mixing board. But it’s not necessarily so; any increase in loudness in one track can mask more serious problems in other tracks that can continue to cause problems in the mix, even more so once they’re harder to pinpoint. Among other potential problems.
The point of this cautionary tale is this: Dynamics processors use numerical values for their parameter settings, but they require your careful listening to do their jobs properly. They’re not magic.
Well, some are. We’ll get to those in a bit.
So what exactly do these things do with the dynamics of your music? All kinds of things, such as:
- Making bits that are too loud softer
- Making bits that are too soft louder
- Making unwanted noise between notes and phrases disappear
- Making just one part of the frequency spectrum louder or quieter
- Allowing note attacks to pass through unaffected while making the sustained parts louder
- Increasing the sustain of certain sounds
- Allowing certain sounds to increase or decrease in loudness only when other sounds occur
- Emphasize certain parts of the audio spectrum so that you say, “Wow, that just sounds . . . better!”
- Change their behavior depending on the performance dynamics of the music
Dynamics processors come in several flavors. Interestingly, depending on how you set the parameters on some types, they automatically change to a similar corresponding type. Here are the main types:
What gets “compressed” by this process is the range of volume between the softest and loudest levels in a track. At its most basic, a compressor makes the loud parts softer. This process was invented to make some types of material that naturally have a wide range of dynamics — such as singing — more consistently audible compared with its accompaniment, especially when its target was for radio broadcast. It’s often used for vocals, drums, and entire mixes. Most compressors analyze a sound only by its dynamics. Some compressors — called multiband compressors — divide up the sound by frequency, and then apply the compression separately for each frequency band.
Fig. 1. These two waveforms show the effect of compression. The top waverform is without compression; notice the distance between the transients (the beginning of the notes) and the sustained portions. On the bottom is the same phrase compressed. You can see that the attacks and peaks are much softer compared to the original waveform. We’ll listen to examples of this in a bit.
Compressors and limiters are often one and the same device; the difference is in how you set up the controls. Limiting affects only the loudest parts of a performance and reduces their volume. Sometimes this is used on the beginnings of notes with very loud transients. Sometimes it’s used to reduce the volume of entire sections of music. Generally the goal is to prevent clipping or distortion in the signal chain.
Another aptly named process, gating prevents unwanted sounds that occur during pauses between notes or phrases from being audible. Often this is used to eliminate bleedthrough, which is the sound of other instruments that were being played in rooms or isolation booths nearby, and were therefore picked up at a low level during pauses.
Fig. 2. In the un-gated top waveform, you can see the low-level background noise in between the instrumental performance. In the bottom waveform, the background noise has been eliminated by a gate processor.
Gates and expanders go together. Where a gate will close entirely to prevent soft sounds or noise from being audible, expanders simply reduce the volume of a track during pauses. An expander does the opposite of a compressor: The compressor finds loud sounds and decreases their volume, making the peaks closer to the softer parts in dynamic range. An expander finds the soft sounds and reduces their volume, making the peaks farther away from the softer parts in dynamic range. The goal of expansion is to get the best “signal to noise ratio” in a track: You’ll see this written as “S/N” or “SNR” in specification lists, which means “sounds you want/background noise you don’t want.” This ratio is usually expressed as a measurement in dB (decibels) such as 100 dB, which means the signal is 100 times stronger than the noise. The bigger that number, the better the S/N, and the better the sound quality. The good news is that you don’t need to know any of this tech stuff to use an expander!
A maximizer is one of those processor types that at least seems like magic. Maybe that’s overstating the case a bit, but the fact is, processors such as maximizers and enhancers use compression and limiting processes that are optimized to achieve particular results, without forcing the user to fiddle with the entire range of compressor or limiter settings. You put your track or full mix through them, and they just sound . . . better. Specifically, maximizers limit the peaks and increase the perceived loudness of a track or song.
How do dynamics processors work?
Dynamics processors do a variety of tasks, but they all work on the same basic principles:
- They sort sounds out according to whether a sound exceeds a threshold or not
- They apply their effect to sound that’s above or below the threshold
- The timing of the effect depends on settings for attack and release
- The amount of the effect is determined by settings for a ratio (for compression/limiting) or a range (or range of attenuation; gating/expansion)
Let’s look at how a threshold setting affects the way dynamics processors work.
Fig. 3. Here are four sounds at a variety of dynamic levels. We’ve set the threshold for around -24 dB; our volume scale indicates the relative volume that corresponds with that of the channel faders in Reason. Sounds A and B are above or are mostly above the threshold, whereas sounds C and D are below it.
Fig. 4. If we apply compression, it affects the sound above -24 dB. You can see that the volume of sound A has been reduced, and that the beginning of sound B ha, as well. Sounds C and D are not affected. The result is that the range between the loudest sounds and the softest has been compressed.
Fig. 5. When we apply gating to the same sounds with the same threshold setting, we get some dramatic results. Sound A passed through entirely unaffected. Sound B got cut off when its volume fell below the threshold. Sounds C and D have been gated completely, since they were entirely below the threshold. There are ways to use gating so that it doesn’t chop off notes, and there are ways to use it so that the chop becomes a cool effect. More on these in a bit.
Attack and release settings let you adjust dynamics processors to work best with your music. A fast attack time means the effect is applied fully the moment that the sound exceeds the threshold. A slow attack time tells the processor to apply the effect gradually (relatively speaking), ramping up to the full effect. Let’s look at how these settings can affect your music.
Fig. 6. This sound has a loud, fast attack transient and a long decay.
Fig. 7. With the attack and release set for fast response, the attack transient gets compressed right off the bat, as does a portion of the sound that follows the transient. When the signal dips below the threshold, the decay actually seems to jump up slightly in volume.
Fig. 8. With a slow attack, the compressor effect doesn’t take hold until just after the attack transient. Then the sustained part of the sound is compressed, until it falls below the threshold. At that point, the compressor doesn’t just stop working suddenly, it eases off gradually. The result is much smoother, and the attack transient sounds natural.
The ratio setting on a compressor is where you dial in how much compression occurs once the volume exceeds the threshold. Usually this is demonstrated with a diagram that illustrates how loudness changes as compression is applied above a threshold. But we’d rather show you how it looks on your music. Here is a series of identical snare hits, repeated over and over at precisely the same volume. The only thing that changes in Figure 10 is the compression ratio, which we’ll dial as the example plays from 1:1 (no compression) all the way up to infinite compression (maximum; also known as brick wall compression).
Fig. 9. Here are the snare hits, with a compression ratio of 1:1, that is, no compression.
Fig. 10. The only thing that changed for this waveform was that we gradually increased the compression ratio from 1:1 to infinite. As you can see, higher ratio settings can have a dramatic effect.
Using dynamics effectively in your music is a matter of trying out combinations of these settings. You may find that for your music subtle settings work great, making your mixes sound polished and professional. Or you may enjoy using dynamic processors as an extreme sound design tool, making such effects as compression pumping and dramatic gating part of your sound. In the next section, we’ll have some listening examples that demonstrate a variety of dynamics processing applications, from basic to wacky. We’ll use the excellent dynamics processing available in Reason, both within the mixer channel strip and in the included devices.
How do I use compressors in my music?
One of the most common applications of compression is for vocal tracks. The human voice is incredibly expressive, and part of that expressivity is the wide range in volume we can employ within a phrase or even within a single word. Consider the following phrase.
This solo female singer uses a full range of expression, including dynamics:
Why would you want to change that performance at all? It’s beautiful the way it is, in the context of a solo. The decision of whether to apply dynamic effects should always be based on your musical sensibility, and on what you hear. Once you know all there is to know about dynamics processing, let your music tell you what it needs, rather than using these fabulous tools just because they’re there.
That said, consider the following context for the same vocal
Here’s the same solo female, this time backed up by a smokin’ rhythm section:
This track isn’t just telling us the vocal needs compression — it’s yelling for it! The softer words and the ends of the phrases get lost in the sound of the band. Let’s apply some gentle compression from the channel strip in Reason. Here are the settings:
Fig. 11. In Reason, the channel strip features a very smooth, fine-sounding compressor/limiter and gate/expander. Let’s turn on the compressor, set the threshold for around -24 dB, leave the attack slow, set the release on the long side, say around 700ms, and dial in a fairly substantial ratio of 4:1, since we’ve got a lot of sonic competition in the track.
Ahh. Much better. The subtle compression we added in Reason's channel strip makes the vocal sound nice and tight, and perfectly audible — without any adjustment to faders or EQ.
Another common application of compression is to take background parts and squish ‘em to make them sound more like a single instrument or voice, even though there may be very thick chords being sung or played. Here’s an excerpt from a blues track.
In this blues track, notice how the horn section gets a little lost in the mix, and how it sounds like three individual sounds:
If your artistic intent is to communicate the individuality of the three horn players in the section, then let them sound independent. But if you really wish they sounded more cohesive, try some excessive compression.
This time, we’re going to set up Reason's MClass Compressor device as an insert effect, and we’ll send the three horn tracks through this compressor. The attack on each note is important, especially on the trumpet, so we’ll use a slow attack on the compressor to preserve them. The notes are warm and breathy, so we don’t want really obvious compression artifacts; the soft knee setting lets the compression build gradually, rather than suddenly, so we’ll use this as well. To really get a squashed sound, we’ll dial up the ratio to nearly the highest setting. Then we’ll have a gentle release. Finally, we’ll add some output gain to make up for volume lost by the heavy compression. Here’s how to accomplish this.
Fig. 12. To set up a background section as a group, add a 12:4 mixer to your rack, then add an MClass Compressor to the rack; the compressor is automatically wired into the aux sends and returns. Then take the direct outs of each instrument or background voice channel, and connect them to their respective channels in the 14:2 mixer. Finally, set all three instrument Aux 1 sends on the 14:2 to maximum, and crank the Aux 1 return above the 14:2’s master fader. Now all these tracks can be controlled with one fader, and they all go through the same compressor settings.
Fig. 13. We wanted to really squash the horns, but also to give their note attacks some room. Threshold: -30 dB, soft knee. Ratio: 60:1. Attack, 36 ms; release, 100 ms. The MClass Compressor has a very smooth sound, perfect for this type of application.
The massive compression we applied to the horn section as a subgroup helped bring it together as a single sound:
Dynamics processors usually perform their duties based on the audio that’s passing through their circuitry. However, there is a groovy feature that allows you to control the effect a compressor has on a track — by using the audio from a separate track. It’s called sidechaining, and it has a variety of uses, from dance tracks to voiceover work.
The devices in Reason are designed to take advantage of the musical power inherent in sidechaining, and it’s very easy to set it up. If you have Reason installed as well, sidechaining opens up entire worlds of creative possibilities, as you can use signals from just about any instrument or device to control the sound of many other instruments.
Let’s say you’re doing the voiceover for a podcast, and you want to have music playing in the background while you speak. Normally, every time you start to talk, you have to pull down the fader for the music bed. With sidechaining, you can trigger a compressor with your voice, but have the effect apply to the music bed, lowering its volume while you speak. This technique is called “ducking,” and with Reason's Audio Track Devices, it’s very easy to set it up.
Fig. 14. Each Audio Track Device has inputs expressly designed for sidechain input from another device. On the back of the voiceover track, take the Insert FX “To Device” output and connect it to the Sidechain Input on the back of the music bed Audio Track Device. Then turn on the compressor in the music bed’s channel strip and tweak it until you get the ducking effect you want.
Easy and effective ducking:
With exactly the same simple little setup, you can create grooving synth parts by triggering the synth channel’s gate instead of the compressor. Just click on the gate’s “On” button and adjust the threshold.
Drum loop and synth pad with gate off:
Drum loop and grooving synth stutters with gate on:
Maximizing is generally added after the mixing process, when you’re putting the final mastering touches on your tune. Reason's MClass Maximizer operates much like the compressors and limiters we’ve worked with to this point. But it’s got an element of magic to it, too. It’s one of those things that can just make a mix sound better. The Maximizer is automatically patched in to the master channel effects; bypass these effects as you’re doing your rough mix, then add them in once you’ve got your EQ and track balance where you want it. Check it out.
Fig. 15. The MClass Maximizer’s settings are familiar to you from your compressor work. This is the place to actually indulge in the feeling that “louder is better.”
Rough mix of our bluesy horns, before maximizing:
The Maximized Blues. Sounds better already!
The “Good” Button
So far, we’ve talked about how to use compression to solve problems with individual tracks and groups of tracks. There’s a very special part of the mixing process that applies dynamics to all your tracks: mastering. Here at Record U, we’ll be providing you with tons of information on the mastering process, which involves using dynamic processing a great deal, and in very specific and groovy ways.
But even before we get into mastering, let’s check out a very special compressor that’s integrated in to Reason's mixer. At the top of the master channel, above even the mastering suite and insert effects, is the Master Compressor. Reason's mixer is a meticulous software model of the classic SSL analog consoles, and the Master Compressor is a shining example of this attention to detail.
The Master Compressor not only looks just like the famous mix buss compressor on SSL analog consoles, it makes everything sound good at the touch of a button, just like the original.
The Master Compressor is a faithful recreation of the mix buss compressor on the SSL mixing boards, from the look of the VU meter to the circuitry. It was designed and built like a typical mastering compressor to give an entire mix the “radio sound” as soon as it was turned on — even though it does have the standard compression controls, including sidechain keying and a “makeup” gain for increasing the volume following compression. You engaged the original by pressing on the “In” button; many people took to calling this the “Good” button because it simply made things sound better when you clicked it. You can do the same with your music by clicking on the “On” button on the Master Compressor. Try it out!
Dynamics 'til daylight
Reason's dynamic processors sound fantastic, and they’re extremely flexible. On every channel of the mixer you’ll find compression and gating that not only gets the job done, but sounds smooth and warm, too. The MClass effects are very effective. And we didn’t even get to the Comp-01 half-rack compressor!
It’s incredibly easy to try out all the dynamic processing effects Reason has to offer. Experiment often, and soon you’ll have a good sense of what kind of dynamics processing works best for your music.
Use dynamics processing sparingly; even if you can’t hear the compressor doing its work when you solo a track, check the track in the mix. Nothing messes up a mix faster than dynamics processing that goes overboard. Use the “bypass” switch on the dynamic effects to make sure you’re not straying too far from where your music wants to go!
Currently editor-at-large for Keyboard magazine in the U.S., Ernie Rideout writes and performs in the San Francisco Bay Area.