Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
(Visuals make these concepts easier to understand - check out the video!)
Watch the video to learn more >
You might be thinking, I understand compression; I don’t need to visualize it, but trust me for a second - this is helpful.
So, say I have a snare playing every quarter note - let’s say our BPM is 120, so that means a snare hit occurs every 500ms.
Let’s say we set our compressor’s attack time to 20ms. Now, the attack time isn’t instant - once the snare’s transient crosses the threshold, we have 20ms before the snare is fully compressed.
So, a little at 1ms is compressed, more at 5ms, more at 10ms, and the most it’ll compress occurs at 20ms and after until the signal falls below the threshold; the onset of the attenuation almost always occurs in an exponential way; a little is compressed at first, and the compression quickly becomes more aggressive the more time goes by.
After looking at various snare drums, I found the average length of the attack and decay, or the transient, to be about 60 milliseconds. So, in this example, the majority of the snare’s transient is being compressed.
Additionally, right around the time when the attenuation becomes exponential and right after, we have the greatest amount of waveshaping.
Since this region is comprised of high amplitude signals, and the compressor is just now hitting its greatest amount of attenuation, the waveforms around this area will have the greatest amount of change applied to them.
By shaping these waveforms aggressively, we cause audible harmonic distortion, which will amplify the transients, albeit not as much as they were just attenuated.
Now, let’s say from the point right when the snare crosses the threshold, it takes 200ms in total before the signal falls below the threshold. Also, let’s say our release is set to 400ms. What would happen to the next snare hit?
Remember, in our example, a snare hit is present every 500ms. Since it takes 200ms before the signal falls below the threshold, and our release only begins once the signal falls below the threshold, then for 400ms, after that 200ms, compression is still occurring.
This means that by the next snare hit, our compressor is still compressing. So, the first snare hit had some of its initial transient let through, but the second snare hit will be attenuated in its totality, transient included.
This becomes even more interesting when we consider the threshold - in our example, say it takes 200ms before the signal falls below the threshold if our threshold is set to -10dB.
But, if the threshold is set to -20dB, it’ll take longer before the signal falls below it, causing the release to initiate later - meaning compression will occur for a longer period of time overall.
Although using your ears is always best, keep the timing of your track in mind. If you know that the snare occurs every 500ms, like in our example, and you want to retain its transient, be sure to set a quicker release to avoid entangling other transients.
Additionally, you can experiment with the attack time to find what sounds the best - a shorter time will result in more harmonic distortion due to waveshaping, which can sound good and even make the snare a little louder in some instances.
A longer attack, around 50ms, will let just about the full transient through while distorting a little of the transient’s backend. Additionally, The lower the frequency or fundamental of the snare, the greater the amount of distortion.
Of course, there are more factors to consider, and these principles definitely apply to any instrument, not just snare drums, but this should give you ideas to consider before compressing.
Let’s listen to a snare that’s been compressed with attack and release times that significantly impact the transient and then settings that let the transient through.
Watch the video to learn more >
Compression is considered a stand-alone effect—something to use when you need to control the dynamics of a dry or relatively dry signal. Then, other effects are built on top of that dynamically balanced signal.
But think of compression as a dynamics controller for any effect you use.
For example, say I’m reverberating the vocal. If I place a compressor after the reverb, attenuate some peaks, and then amplify post-compression, those reflections are going to become a lot louder.
Since I lowered the high-amplitude signals and brought the level back up with amplification, all of the minuscule reflections that were being masked a moment ago are now loud enough to compete with higher-amplitude signals.
This is especially helpful in a parallel context. If my vocal is spectrally and dynamically balanced, I could send it to an aux track, on which I’ve inserted a reverb.
If I set the reverb to 100% (something that wouldn’t be common if I inserted the reverb directly on the vocal’s channel strip), then any compression I introduce next will only affect the reverb’s reflections.
With that in mind, I could use the various settings, like the attack and release, to control the ADSR of the reflections, likely bringing the quieter details to the front of the mix.
The same goes for delay, chorusing, flanging, and just about any creative effect. If you run it in parallel, you can set it to 100% wet and follow it with compression to affect the dynamics.
Let’s listen to reverb with and without compression - notice how much control it gives us over the dynamics of the effect, and consider how you could use it to shape any and all effects you introduce.
Watch the video to learn more >
Downward compression, with all of its complexities, is fundamentally an attenuator. Its primary function is to reduce the amplitude - ideally, whenever you want it to.
The presence of sound, followed by the lack of sound or a noticeable reduction in the sound’s loudness, are the building blocks of rhythm. Rhythm is exactly this, but repeated until we notice a pattern.
Because compressors periodically reduce the amplitude of a signal, this attenuation can and often does get interpreted as a distinct pattern. This pattern can augment or support the rhythms already present in a song via the instrumentation and composition, or it can interrupt them.
I tell you this to emphasize the importance of closely monitoring your compressors, especially how the attenuation they cause affects the perceived rhythm of the track. This is especially true for bus compression, in which a compressor’s attenuation affects all of the rhythms in a song.
To demonstrate this, I’m going to play a track that consists of all static signals - the instruments don’t have any rhythm or pattern; they have constant notes and noise.
With compressors and ghost notes to trigger the compressors, I’ll create various patterns and create the perception of rhythm by contrasting the signal with lower amplitudes or silence.
I’m going to start the track with all compressors disabled. I’ll keep the volume low, but just a heads up, it sounds awful. Then, I’ll enable the compressors—still not great sounding by any standard, but you can definitely hear how the triggered attenuation caused by the compressors creates the perception of rhythm.