This review page is supported in part by the sponsors whose ad banners are displayed below

Once you’re faced with discrete tracks of a recording, say the drums, various guitars and vocals, they must be fixed in relation to each other. This process is called mixdown. Its central tool, no surprise, is the mixing console. The result of a mixdown is a final stereo or surround track which in theory could go straight to physical medium or a download portal were it not for the intermediary step called mastering (which technically really is premastering) which works over said mixdown. We’ll get to that on the last page.


Level matching. Fundamentally that’s very simple indeed. Primarily the mixdown determines which track gets how loud. Enter the signature vertical faders of the mixing console. Those work exactly like a volume control on your hifi system. Down is quieter, up is louder. With that covered, now we just need some filler paragraphs to hit fairaudio’s obligatory word count and get published. But wait, there’s rather more to it still. The mixdown is a real art form which can rely on considerable extravagance and highly specialized tools though level matching does remain at the heart of its task. The problem is that basic adjustments of amplitude aren’t the only parameters which need editing. A virtual stage relies on mapping its three dimensions. The relative loudness of a given sound is a complex sum of spectral and time-domain elements which lead human hearing to perceive its loudness relative to simultaneous competing signal. This includes the loudness of surrounding elements, overall loudness, personal experience and sympathy or aversion to specific sounds. For obvious reasons our ear/brain always treats the human voice with particular emphasis.

One of the level-matching tools are the so-called faders which might exhibit enormous range to enable very precise settings. But this needn’t be static at all. An early piano accompanying solo vocals would go on your nerves if its level remained fixed once the entire ensemble entered. Hence modern mixing consoles and DAWs rely on automation tied to fader positions and other parameters for easy editing via mouse click. For simplicity’s sake or to create later access to overall groupings, one might assign subgroups whereby for example all guitars can be edited together. And there’s also a mute function to kill certain microphones completely, say the one on the drums, in the room for ambiance or the one aimed on the  guitar amp, perhaps even for the oboe whose inclusion in the ballad took a heavy toll on the production budget. Certain signal simply needs room. Other signal may have to step back or get completely muted.


Frequency domain edits. What does a normal music consumer do when his system suffers bass boom? He reaches for the equalizer. So does the sound engineer when an electric bass seems a bit thick in the mix to interfere with other signal; or if the vocals are loud enough but lack presence and incisiveness. Even so a mixing console—never mind an external equalizer or plug-in—offer far more functionality than the basic treble or bass controls we remember from hifi. Those tone controls hinge on a fixed frequency above which (treble) or below which (bass) the signal is cut or boosted. On a mixing console their equivalent is called a shelving filter but often the possible adjustments include the precise filter frequency. Occasionally one might want something more drastic than a mere cut for certain spectral ranges. Now high-pass and more rarely low-pass filters enter with variable corner frequencies and attenuation slopes. And yes, that’s exactly how a loudspeaker crossovers performs.

Here the most common tool is the parametric EQ. Its alternative name of bell filter describes its operation of adjustable-width cut/boost for the so-called Q factor which centres on a variable frequency. Bell filtering allows for narrow notching, perhaps to remove resonance modes from a drum. Frequency filtering thus can alter a signal very strongly. And in general the goal of most productions is not the most natural signal but the adaptation of it for the best final mix. That’s supposed to sound good as a whole, not as unedited individual tracks. More than one acoustic guitar would sound shockingly thin, bass shy and one-dimensional if it were called up with its solo button yet it no longer competes with the more important main vocals. Hence by emphasizing or diminishing certain spectral elements, signal is steered. By shifting components in the single-digit kilohertz band, a bass drum for example might sound round, warm and cozy or wiry, aggressive and dominant – whatever the song calls for.

Dynamics. Level matching and frequency edits are important tools and relatively easy to understand and manipulate. Things get more complex when one addresses dynamics though the core principle is again simple. Dynamic effects edit the level differences of a track. Here the compressor becomes the most important tool. It minimizes the differences between quiet and loud passages. High levels are cut, low levels are boosted to increase density and average volume. This makes it easier to control overall levels but there are many more parameters which are manipulated in this fashion.


Thickening or greater volume density isn’t applied across the board but only when signal eclipses a predetermined value called threshold. Above it the relationship between input and output dynamics is  compressed at a ratio that’s usually adjustable. Typical  compression values are 2:1 and 4:1, i.e. the output signal’s dynamic contrast is halved or quartered. Theoretically this could be accomplished manually or with automated fader adjustments but likely not fast enough. Many compressors have astonishing response times and can be set for how quickly after a threshold trigger they kick in and let go.


Enormously important is the trigger window over time. Since a tone’s transient or leading edge determines much of our perception of it, it makes a big difference whether the compressor squashes the leading edge or lets it pass. A bass drum thickens up if instant compression diminishes its attack. Should you be surprised that amplitude cut can make for positive results, reduced dynamic range also moves up the average level without having its peak redline and distort. A very sudden very powerful dynamic editing tool is the limiter with an 8:1 or higher ratio which is used primarily to remove peaks. Particularly in the final mastering this limiter sees regular use.


[At left, original signal on top, compressed signal in the middle, additionally limited signal at the bottom.]


Compressors not only have a technical but also aesthetic purpose since they can colour the signal. Particularly valve-based compressors or certain vintage transistor designs are very popular and mostly applied to the human voice. But even very subtle and gentle machines or plug-ins alter or remake the original signal. The ‘colour shifts’ of compression aren’t always advantageous, particularly with purely unplugged acoustic styles. Here it can be fascinating to listen to an Original Dynamics recording on the BIS label over a system that’s fully capable of exploiting uncompromised dynamic range.


Certain dynamic editing tools reverse the action of compressors but expanders are more rarely used. A special form of dynamic limiters called noise gate is more common. Once a signal misses a minimum threshold, those block it completely. As the name suggests, these tools avoid the meaningless hum of channels which momentarily don’t carry any useful signal. Noise gates then are a type of automated mute switch whose various parameters are hyper adjustable. A special feature of dynamic editing tools are assignable trigger sources. A standard compressor responds to the input signal. The higher the level, the higher the compression. If you wanted to limit the entire drum set, the bass drum becomes the biggest influencer since it accounts for the highest percentage of the group signal. With a high-pass filter inserted into the trigger detection path, this effect gets diminished. But a compressor can also be triggered by rerouted signal. If a guitar track compressor is programmed to take its cues from the vocal track for example, the guitars are attenuated whenever the vocalists get quieter as a form of sync’d auto leveler. These basic examples barely scratch the surface of what’s possible.

The virtual stage and delays. To assign a central localization between the two speakers of a stereo system relies on a basic level adjuster called pan which essentially mirrors a vintage balance control. To move a signal back on the virtual stage requires a bit more help. The most important tools here are level and ambient room data. Obviously if an instrument or singer is supposed to occur forward in the mix, one first has to define the acoustic rear or stage end. Artificial reverb generates spatial contents with occasionally enormously complex processors or even ambience that’s synthesized from actually sampled venues. Hall sound also fulfills other purposes that needn’t suggest an actual room. Echo trails, reflective walls or their digital samples are very popular. Hall sound can colour, texturize and homogenize a signal. Pay attention for example when George Michael sings an ‘s’ or ‘t’ to create a silvery flash of decay trail. Reverb can also ‘kit’ signal into discrete groups. The mixing console uses special aux-send connections to feed reverb processors with various signal. Decay is rather basic by contrast. A simply delayed copy added to the original signal can be quite effective already. Such an echo can fake up a larger venue without congealing the overall signal with heavy reflections.