Missed the start of the series? Catch up here.
Compressors, limiters, single band, multi band, soft knee, hard knee, ratios, attack, release… on and on and on.
These terms sound more like medical tools/procedures used in gastrointestinal bypass surgeries!
Despite some of their more gruesome sounding names, these are all terms related to, and describing dynamics – and dynamics control in audio. Let’s define a few items of interest.
Core Definitions in Dynamics
“Dynamics” in its simplest and most functional definition means this:
Dynamics: An audio signal’s lowest and highest points (with regards to volume), and the subsequent and continuous raising and lowering of those points, with consideration to everything else in between.
Dynamics: The proportion/relationship between the loudest and softest parts of that audio signal.
A simpler definition is this:
Dynamics: The peaks and valleys of an audio signal.
Now obviously there is much more to it than that last sentence, but essentially what you are talking about is Volume, or actually… the changes in volume of a signal!
Now, one more definition we need before we move on:
Dynamic Range: The actual measurable range between the absolute softest and loudest parts of a signal.
Misc. note: There is A LOT more to dynamic range, but that is way beyond our scope for now. As a matter of fact, a full discussion of dynamic range is more suited for recording/mixing, and mastering discussions. If this subsequent topic is of interest to you, please let us know in the comments!
OK, I lied. One more definition. Sorry!
Dynamic Control: The ability to manually, or automatically control an audio signals dynamics and dynamic range.
THIS is the important one… Remember it!
Now that have we our core definitions, all of this begs the question: How does all of this fit together?
Or, the more relevant question: Exactly how do we manipulate, regulate and control dynamics along with dynamic range?
Eventually, we will ask: Why would we want to?
First things first:
We control and manipulate dynamics with compressors.
Compressors come in a variety of flavors (as do all effects): from the budget-minded stomp box (pedal), to studio spec. rack mount units, all the way to the multi-thousand dollar high-end units used by mastering engineers.
ALL of these compressors do this:
They raise the softest part of a signal, and lower the loudest part of that signal.
The result is a compromise that lets both softest and loudest parts be clearly audible. As a tangential effect, reducing the dynamic range to a narrower and overall consistent field allows the whole signal to be raised in overall volume. This is especially beneficial to mastering engineers, as this is one of the ways that they make mixes sound louder.
There, in the above paragraph, also lies the “why”. With dynamics processing we can regulate individual or many instruments, and “herd” them into a specified range for the sole purpose of being able to hear the softest and loudest parts equally, as well as have all the sounds blend well together. In simple terms, we want our collection of sounds or instruments to play nice together!
That’s it in a nutshell. That is the basic function of a compressor reduced to its lowest common denominator. It is simply a tool to manipulate dynamics and dynamic range.
Compression ear training
Here’s a little taste – listen to the original funk riff, and then the compressed version:
Can you hear how the dynamic range has been reduced? The quiet parts are now louder, and the louder parts quieter, giving a more consistent sound throughout.
Misc. note: Some compressors impart a tonal color of their own, while other compressors are totally transparent… as in, all you hear is the compression with no added tonal altering. Pedal compressors used by guitar players usually color the original tone, and that can be a desirable trait in the overall tone creation. However, with studio compressors or especially mastering studio compressors, tonal alteration is viewed as a negative side effect to be avoided at all costs. And I do mean that fully literal as well, since transparent compressors that only alter dynamics are usually above the $1000 range! Of course, which compressor is right for you is entirely subjective, and dependent largely on your tonal needs for the job at hand.
Uses of Compressors
Here are a few basic uses of compressors:
- Drums: Adding punch, or creating entirely new sounds.
- Guitar: Adding sustain. Balancing out the strings so they are all equally audible.
- Bass guitar: Same as guitar, plus enabling that slap/funk sound.
- Vocals: Gently bring up the softest whispers and tame the peaks.
- Full mixes: Get all the instruments to blend well.
- Mastering: This is a whole separate topic on its own!
We’ll see examples of some of these in the compression ear training examples which follow.
Moving on, of course there are different parameters that allow control over certain aspects of your compressed output. Here are the most common ones. Actually, let me restate that… the most important ones!
Here are the basics of each:
The Threshold is what controls the actual point when the compressor clamps down on your signal and works its magic. The lower the threshold, the more signal gets compressed. The higher the threshold, the less signal gets compressed (since the high setting allows only the top part of a signal to get compressed).
Different threshold points are required for different instruments. For example, slap bass requires heavy compression, so naturally you want the entire signal processed, and hence a low threshold; whereas something like vocals that require very gentle manipulation would utilize a high threshold to maintain a natural, unprocessed sound.
A clean guitar would be a candidate for various thresholds. High threshold for just a bit of compression, and low for full processing and that “everything sounds equally loud” funk tone! By the way, because of the squeezing of soft and high sounds that a compressor does by default, instrument sustain is increased. That is why guitar and bass players love them!
There are many variants in between, but I think you get the idea.
Time to hear it in action! Please realize that the settings used to “demo” the various functions are set to VERY drastic, and to be quite honest, “obnoxious” levels… BUT this is done strictly to make the differences blatantly audible. In real life things would be set much more subtle; with attention to musicality and finesse!
With that said… Enjoy!
Light clean picking, same lick x3. Threshold set to 0, -25dB, and then -45dB.
Note how the sound gets “splattier” as the threshold setting is reduced, allowing for more compression.
Light clean picking with chord, same lick x3. Threshold set to 0, -25dB, and then -45dB. In addition to the “splatty” single notes, see how the chord at the end seems “suppressed” with reduced threshold settings, as if it is being choked.
Note: When the threshold is reduced, more compression takes place and therefore the overall volume level gets reduced. It is common practice to use a parameter called “make-up gain” to replace the volume that has been lost. In clips 1 and 2 this was used.
Here in clip 3 you will hear the EXACT same threshold settings WITHOUT make up gain.
Light clean chord, same lick x3. Threshold set to 0, -25dB, -45dB respectively. Make-up gain OFF.
Yeah, I know… Compression without make up gain is pretty useless!
The Ratio parameter works in concert with threshold, and controls the proportion of the output that gets compressed compared to the input signal. It is the input/output ratio used when signals exceed the threshold.
Common ratio levels are 1:1, 2:1, 4:1, 6:1, etc. Here is how the math works: If you set a ratio of 2:1, that means that if the input signal exceeds the threshold by 2 decibels, then after compression, the output will only be 1 decibel above that same threshold. 4:1, (a guitar favorite) means that when the input signal exceeds the threshold point by 4 decibels, then the output is still only 1 decibel above that same threshold. The difference between the input and output is called gain reduction. In the 4:1 example the gain reduction is a total of 3 decibels. On a ratio of 6:1, it would be a gain reduction of 5 decibels.
Did you notice that a 1:1 setting will have no effect? 1dB change on input goes to 1dB change on output! And no gain reduction.
There is much more to say, but for now this will give you the idea that the threshold and ratio are the controls that govern just how much processing you want.
Now, let’s have a little fun pushing the limits of good taste! Naturally it will be easier to hear the effect with the higher ratios. Once you’ve tuned into these, it should be easier to hear the more subtle lower ratios.
Two chords x5. Ratio settings: 1:1 (i.e. no effect), 3:1, 4:1, 6:1, 10:1.
Note how the higher the ratio, the more the sustain at the end of the last chord swells and “howls” more. Also note how the clip starts to “waver” and distort!
Chord comp x5. Ratio settings: 1:1 (i.e. no effect), 3:1, 4:1, 6:1, 10:1.
Note how the effects of clip 4 are even more prominent with this clip which contains more sonic information. Also notice how the chord comp becomes more fatiguing and noisier with each ratio increase. If there is a lesson here, it is this: Compressors are high powered sonic weapons that need to be handled with care! Truly, a little goes a long way!
The Attack is simply what controls the speed in which the compressor starts to operate. Ranges are usually from a few milliseconds to a few hundred milliseconds – or even more. Here are the relevant facts:
Faster attack means almost instant compression. This means you do not get to hear the initial natural beginning (or attack) of the input signal.
Slower attack times simply means that you get to hear that natural beginning of your input signal. Great for drums, since the attack is the main component of the drum sound. This results in a very, very punchy sound with the control specified by the compressor. Classic studio trick used for loud punchy drum tracks.
Now be warned: these examples will sound OBNOXIOUS!
Chick’n pickin’ lick on bridge pickup x5. Attack settings: Original, .01ms, 2ms, 500ms, 1 second.
This is a peculiar control. Note how on the second time through (the .01msec setting) the signal distorted violently. Very fast attacks always do that. Remember that sonic phenomenon.
Past that, slower attack settings get more and more percussive on the transients of the phrase. How much you want is, well, up to you. But again, tread carefully!
Chick’n pickin’ lick bridge and middle pickup x5. Attack settings: Original, .01ms, 2ms, 500ms, 1 second.
Just a different tone for you to hear. Also, I lowered the threshold to get more compression. Here, slower attack times work better since more signal is being squashed. The slower attack times seem to fill out the sound more, and are a little more musical. Subtle difference but it’s there!
Also the real fast attack (.01ms) distortion phenomenon is still present here as well regardless of threshold settings!
Before I forget, in this clip, the 2ms setting almost distorted, producing a “wavering” sound. Hopefully you are starting to see how all these controls are very dependent on each other!
The Release parameter works in conjunction with the attack control to specify how soon the compressed signal returns to an uncompressed state. Times vary from a few milliseconds all the way to a few seconds. Simply put, faster release times get the signal back to “natural home base” almost instantly, while slow release times let the signal stay compressed longer, and return to its natural uncompressed state in a slow deliberate way.
This where artifacts called “pumping”, or “breathing” can come from. Too fast a release and the change between compressed and uncompressed becomes WAY too obvious. Too slow of a release, and now you’re dragging the compressed signal out too long, which results in a “squashed” sound. Some compressors offer automatic attack and release to take out the guess work, but nothing beats the manual settings for specific surgical control!
Let’s listen to the control over how quickly things get back to an uncompressed state!
Clean sustaining chord x5. Release settings: Original, 50ms, 500ms, 1 second, 5 seconds.
This particular example is subtle, but if you pay attention you can hear the decay of the chord wanting to die out but not being able to. The 50ms setting is the most musical. Even though the other settings aren’t completely out of control, there are still subtle tell-tale signs of too long a release: very slight wavering, distorting, up and down pumping, etc. They are all very slight, but listen closely and you’ll pick them out!
Listen to the tracks above for Threshold, Ratio, Attack and Release. Listen again! Try to tune your ear into how changing each parameter affects the sound. Read the explanations, and try to hear what’s going on in the signal. The examples are going to be tougher next time, so make sure you get the basics of compression ear training down now!
There are many, many more aspects to discuss, but remember, we are laying groundwork here!
Next time we’ll build on this with a look at limiters, expanders, gates, and single/multi-band compressors. These are all direct relatives or variations to compressors with similar control layouts, but their own uses and specialities. We’ll also look in depth at real-world examples of compression and see how to choose parameters to get that particular sound you’re going for.
Any questions? Don’t be shy!