Scenario 1: Suppose you have a two-band compressor, with the band cutoff at 120 Hz. Both bands of your compressor use the exact same compression settings. So the same attacks, releases, and thresholds are used on each band.
Would the compressor work any differently if you removed the cutoff and just used one band? Since both bands use the same settings, does using two bands really make a difference?
In short, using two bands makes a crucial difference in the way your compressor will operate.
Before we get into the reasons why, there is one small nit to pick. When a multiband effect is used, a set of operations called a “crossover” will be used to separate your audio into bands, and later to stitch them all back together. In this process, crossovers can leave a slight “fingerprint” on your audio, and so a multiband effect will often sound slightly different. This fingerprint is very small and can be further reduced by choosing the right settings, so for the purposes of this question we can ignore the crossover.
So, with the crossover out of the way, how are these two effects different, and where does the difference come from? The difference has to do with the number of compressors running in each effect.
In a one-band design, a single compressor acts on the whole audio spectrum, from the bass up to the treble. This means that if a loud kick were to sound in the bass, the compressor would take action on the entire spectrum at once. Everything in the spectrum, including a quiet sound in the treble range, will be attenuated in response to the bass.
In contrast, the multiband design really represents *two* compressors, each operating on a separate part of the spectrum. The first compressor covers the range below 120 Hz, and the second handles all of the signal above 120 Hz. A loud pulse in the 80 Hz region will register on the first compressor, but not the second. The first compressor may respond to this pulse, if it’s sufficiently loud, but even so it will only attenuate in its own frequency range. The signal below 120 Hz may be adjusted, but the signal above 120 Hz belongs to the second compressor and will not be touched.
This distinction is important when you consider how much energy can be conveyed in the bass frequencies, from high-amplitude kicks to bass guitar. Multiband compression allows you to isolate and treat the heavy bass dynamics on their own, without making changes to your fragile higher frequencies.
Scenario 2: A loud and punchy snare drum is pushing your mix too close to digital clipping. You remedy this by applying a limiter, and clamping down on the very loudest of the snare hits. In general, would you run your limiter in RMS or Peak mode? Would you set fast attack times, or slow ones?
Peak measurements and faster attack times are ideal for limiting the abrupt, dynamic sound of this snare drum.
A limiter, like other dynamics effects, measures the level of an input signal as part of its operations. “Peak” measurements, which are usually the default for a limiter, constantly track the highs and lows reached by a signal. RMS, or Root-Mean-Square measurements, are more complex – they track a figure related to signal’s average loudness, rather than its momentary crests or troughs. While Peak pinpoints the brief crests of a signal, RMS readings correlate well with the overall perception of loudness.
Peak measurements are ideal for detecting a quick, high-amplitude burst of audio, such as we’d encounter in a snare drum hit. RMS readings, on the other hand, might obscure a quick burst by averaging it into the lower values around it. To prevent digital clipping, we want our limiter to catch as many loud snares as possible, so for our purposes it’s best to play it safe and choose Peak.
Similarly, as soon as a loud peak is detected, we want our limiter to attenuate quickly, and hence a faster attack time is best. With a slower attack time, we risk a loud signal ascending to clipping levels before the limiter can react.
Scenario 3: You’ve just completed your very best take ever – the holy grail of kazoo solos – only to discover an intermittent, faint rumbling sound in your mix. (Perhaps your downstairs neighbor was making smoothies for breakfast.) Whatever the cause, this take is too good to throw away. Armed only with an expander, you intend to noise gate your mix and evict the rumble (and perhaps your neighbor).
Part 1: To turn your expander into a noise gate, how should you set the expander’s ratio?
With a ratio greater than 1, an expander can function as a noise gate – this is also known as “downward expansion.” Sounds beneath the threshold level are mapped to quieter sounds, or to -INF dB, depending on how high your ratio is set. In general, the higher the ratio, the “steeper” the gating effect.
Part 2: Part of the appeal of your kazoo solo is that you’ve managed to record in a naturally reverberant setting. After gating the mix, you find to your horror that the gate has cut the tails off your reverbs! They don’t fade away naturally, but rather vanish completely as soon as they reach the gate threshold. Is there anything you can do about this?
Setting a slower attack on the expander may help preserve the natural fades and reverb tails in your music. A slow attack will delay the gating action, so signals that drop below the threshold will have more time to fade away naturally. Note that a slower attack will also give unwanted noises a greater chance of passing the gate. In such difficult cases, a dedicated noise removal effect may be called for.
Scenario 4: Dynamic Range refers to the difference between the loudest and quietest sections of a mix. In general, will compression raise or lower the dynamic range of a mix?
In most cases a compressor (or limiter) will lower the dynamic range of a mix, because it attenuates the highest amplitudes downward, closer to the noise floor. In contrast, an expander can be used to “magnify” a range of levels into a wider range, and thus increase the dynamic range.