Compression is, without a doubt, the tool which people seem to struggle with the most when first getting into audio.

Unlike a lot of mixing tools, techniques and concepts that can often be grasped through watching just a single tutorial, with compressors, it can take years of research, experimentation and practice for everything to finally “click”.

On that note, today, we’re going to try and “fasttrack” your understanding of compression, and hopefully by the end, you’ll be using it in all sorts of creative ways in your mixes. Let’s get started!

First Up, What is Compression?

Just as the name suggests, in its most basic form, Compression is a process used to “compress” (reduce) the dynamic range of an audio signal.

With a raw, “dynamic” vocal performance for instance, the singer’s average volume might fluctuate drastically from section to section, or even from word to word. The verses may be whisper-quiet while the choruses are belted-out-loud, and all throughout, certain words or syllables might be jumping out disproportionately to the rest.

Trying to find a single volume fader position at which this “totally dynamic” vocal can comfortably sit on top of the track at all times simply isn’t going to happen. This is where compression comes to the rescue.

Red: A raw, “overly-dynamic”, uncompressed vocal. Green: The same vocal, compressed with an 1176 compressor at a 4:1 Ratio

As you can see in the example above, we can use a Compressor to somewhat “close the gap” between the loud and soft parts of a signal, to the point that everything is operating within a more confined “volume window” (aka, a narrower dynamic range).

With our green, compressed signal, there are no longer any huge jumps or dips in volume, and we have a nice, consistent average loudness level throughout the duration of the performance.

“Cool. So, how does a compressor work?”

As we mentioned previously, the primary function of a compressor is to reduce dynamic range. However, how you go about setting the 5 primary controls you’ll typically find on a compressor determines the severity and tonal character of said gain reduction:

• Threshold: The threshold control on a compressor determines the minimum input level at which compression will start to kick in. If the threshold is set pretty high, you might only compress the loudest peaks in a signal. If the threshold is set really low, even the quieter parts will be compressed. Anything that remains below the threshold is left as-is and uncompressed.
• Ratio: The ratio control determines the severity of the compression that will be applied above the predetermined threshold. With a 4:1 ratio for example, for every 4dB’s of level that surpasses the threshold, only 1dB of volume increase will be allowed to pass through (-3dB of gain reduction is applied, meaning we’re reducing the dynamic range of everything above the threshold by 75%).
• Attack: The attack time of a compressor determines how long it takes for the compression to fully kick-in after the signal has gone above the threshold.
• Release: The release time of a compressor determines how long it takes for the compressor to “let go” of the signal after it’s gone back below the threshold.
• Makeup Gain: While it may have a fancier name, makeup gain is essentially just gain (volume), and is simply used to bring the signal back up to its original volume after being compressed (if -3dB’s of compression has been applied, you might apply +3dB of makeup gain after-the-fact).

Trying to process all of this information purely theoretically using just your imagination can be confusing. This being the case, let’s dive straight into some visual examples of all of these controls in action, and they’ll probably instantly start making a lot more sense.

Compression, Visualized

Red: An uncompressed 1kHz sine wave with some dynamic variation. Green: The same signal, compressed at a 4:1 ratio (30ms Attack and 100ms Release). Blue line: The threshold used for our compression

In the example above, we can clearly see how the ratio, attack, and release are all affecting the signal based on our predetermined threshold (the blue line).

Let’s superimpose the two signals on top of each other so that we can clearly compare the differences caused by the compression:

Red: An uncompressed 1kHz sine wave with some dynamic variation. Green: The same signal, compressed at a 4:1 ratio (30ms Attack and 100ms Release). Blue line: The threshold used for our compression

So, starting with the obvious: Due to our 4:1 ratio, the huge volume jumps in the uncompressed red signal that went above our blue threshold line are now pretty evened-out dynamically in the compressed green signal, to the point that there’s only a few dB of variation between them.

You’ll notice that the low-level red signal at the very right side is the same level as its green counterpart, as it didn’t surpass the threshold, and was therefore unaffected by the compression (only being somewhat altered by the release as the compressor “lets go”).

At the beginning of each red “volume jump”, as the compression kicks in, we’re now left with transient “volume fade-out spikes” in the green signal, caused by the compressor taking 30ms to fully kick-in based on our predetermined attack time.

Circled: The “Attack Spikes” in the green signal, caused by the compressor taking 30ms to reach its fully compressed state after reacting to each red volume jump

At the end of each red “volume jump”, as the compressor transitions from compressing “X” amount of input signal to the lesser “Y” amount, we now have gradual, 100ms-long “volume fade-in slopes” between the two states, based on our predetermined release time.

Circled: The “Release Slopes” in the green signal, caused by the compressor taking 100ms to “let go” after each red volume drop (from X to Y, for example)

Finally, if we were to apply some makeup gain to this compressed green signal after-the-fact in order to bring it back up to its original volume, then we’ve essentially reduced the dynamic range, while also enhancing (boosting) the front-end attack and back-end sustain characteristics of each note/hit:

Our compressed and uncompressed signals, volume matched using makeup gain

Some Practical Tips On Using Compression

Odds are, actually seeing what compression does to a signal has helped you better understand the process. So to top things off, let’s take a look at some common uses for compression in mixing based on what we’ve discovered so far:

1. Evening-Out Dynamic Sources

Most “raw” microphone recordings will benefit from a degree of compression depending on factors like genre and instrumentation density. Simply put, if you’re dealing with a 200-track orchestral metal track and want your lead vocals to stay upfront and audible at all times, you’re gonna need to use some heavy-handed compression.

Try an 1176 with 10-20dB of gain reduction at the classic “4:1 ratio, attack at 3, and release at 7” settings for great results. Here’s an example (first repetition raw, second repetition compressed):

2. Slow-Attack Compression For Transient Enhancement

While simply reducing the dynamic range of a live kick or snare drum track with some 4:1 compression can be a good idea in and of itself for evening-out hits, doing so with a slow attack time of 10-30ms is a great way of adding a snappy transient boost to each hit, making them sound much more exciting and aggressive.

Try an SSL compressor on the “slow” attack setting with a fast release for this purpose. Here’s an example (first repetition raw, second repetition compressed):

3. Fast-Attack/Release Compression For Sustain Enhancement

By using super-fast attack and release times when compressing drum rooms, you can essentially kill the spiky drum transients in the signal while greatly enhancing the sustain of the room.

The blackface 1176 compressor (or plugin equivalent) at a 4:1 ratio, with the fastest attack and release, and 10-20dBs of gain reduction is a go-to choice for this effect.

Here’s an example (first repetition raw, second repetition compressed):

Conclusion - Go Forth and Compress!

Now that the monster has finally been unmasked, hopefully it’s a lot less intimidating.

While getting to a point where you understand exactly how compression works may still take a while, the quickest way to conquer the beast is to simply charge at it head first.

Use it, abuse it, tweak it, get things wrong, and don’t wait until you’ve “totally got it down” to start using compression in your mixes, because trust me, unless you get hands-on, that day will probably never come…

Until next time - Go forth and Compress!