Psycle Wiki


78pages on
this wiki
Add New Page
Talk0 Share

Go back to Main Page > Documentation > Mastering


The first in a series of "Masterclass in Mastering with Angelus" ;-)

If you think that you can contribute something about this subject, please edit this page and write. It does not matter the slight it is, surely that will be useful. And don't forget to put your name/nick in the Credits section, as well to stamp the version into the Page Version Section.


If you want your tracks to sound really professional, it is necessary to handle the compression well. This is one of the most critical and important steps in the Mastering process, but it could ruin a good mix if it isn't well used.

Have you never asked why one song sounds louder than another one that is already normalized? The answer is in the compression.


The best way to explain and understand how a compressor works is with an example: Imagine that you have a song with a drums track, a synth-pad track, a bassline track, synth-riff track and a lead-synth track.

While the song is playing you see that sometimes the volume is over the maximum ("red zone") and this causes the sound to distort. One solution is to reduce the general mix volume (you can do that automatically checking the option in the psycle's minimixer called "Decrease Master Volume automatically when Clipping"). Another one is to reduce the volume of the tracks that cause this distortion (i.e. drums-track) but this one could make the parts too silent in the parts where they didn't distort.

Before, there were sound engineers with a "trained-ear" who could hear where these peaks occured and then they could operate the volume faders, making quick adjustments to them when needed. Needless to say, this was a lot of hard work.

When the compressors entered the scene all this work became easier. They could just set a threshold level to control where the compressor starts to work and reduce those peaks automatically. Then it's possible to turn the track volume up and make it sound much more powerful without causing distortion. (remember though, that analog distortion is very different from digital clipping and was/is still used consciously to make the sound "fatter")

But too much compression is dangerous. The aim is to compress the maximum that you can without losing the dynamic range (that is, the difference between the upper peaks and the lower ones). Excessive compression could make your song too plain & tiring for the ears to listen, (unless that it is what you want) or it could make the background noise gain volume.

The effect of compression on a final master

A. Is the track with no compression applied and the volume normalized. You can see that the average volume is around -6db (that means only half of the available dynamic range!) and only a few peaks go all the way to the maximum. Even a small boost in the volume would cause these peaks to clip nastily.
The track with no compression

B. Is the same track with a nice amount of compression applied. The input was boosted about 6db and a soft knee compressor applied with fast attack & release. This results in a nice boost to the volume but still allowing for some life and variation in the genral volume level.
The track with nice level of compression

C. Is the track with too much compression applied. The volume is again louder, but it's easy to hear how the track gets muffled and in the loudest parts (namely, the beginning of the beat with the kick drum) distorted too, due to to excessive signal shaping by the compressor plug. (yes, depending on the available range of settings in the compressor plugin, they can also be used like an analog "soft" distortion!)
The track with too much compression

Queue all these examples in your media player and compare. Remember that sometimes "too much" compression can sound good, too! Think of some dirty gritty electro for example...


The compressors have also other controllers in addition to the threshold. Taking a look at any soft compressor (Fig.1 and Fig.2) we can see some parameters like "ratio", "attack", and so on.
Fig.1: Kjaerhus's Classic Compressor VST
Fig.2: Arguru compressor Psycle plugin

Compressor ParametersEdit

The most important parameters that we can see in a software compressor (Fig.1 and Fig.2) are the following ones:


Just as it is explained above, this is the level (usually in dB -decibels-) where the compressor starts to work. The signal below this level is unaffected.


This is the amount of compression that is applied to the signal when it exceeds the threshold level. It is expressed in a X:Y kind of relation, where X is the level of the input signal and Y is the amount of level to increase. When the input level is over the threshold level, for each X dB of this signal the compressor outputs Y dB. Usually the Y value is always 1. The higher the ratio, the more compression is applied.

Example: You are applying a compression ratio of 2:1. It means that for each 2 dB of signal input that surpass the threshold level, the compressor only raises 1 dB (Fig.3-A). It is like if you turn the volume down manually in those input signal peaks to the half.

A ratio value of 1:1 is the same as the bypass function (no compression is applied). A ratio value of ∞:1 has the same function as a limiter.


This parameter has the same function as in a synth or a volume envelope (a compressor isn't another thing but a dynamic regulator of the volume level). It controls the speed at which the compressor reacts when the threshold is exceeded.

The amount of attack depends on the signal you are compressing: If you are compressing a bassdrum signal you will need a fast attack because if you apply a slow one it is very probable that the signal of the bassdrum already has finished when the compressor begins to work.


(AKA Decay) In the same way as the attack parameter, it sets the time (usually expressed in msec -milliseconds-) that the compressor takes to return to its rest position.


(AKA Soft Clip, see Fig.2) This is an optional parameter that controls the shape of the attack response. Usually the values it can take are two: HARD KNEE and SOFT KNEE. A hard knee makes that all the compression is applied suddenly. On the other hand, a soft knee makes the compression to be applied gradually ( Fig.3A-B ).

Usually a hard knee is applied for drums or bass to give more punch, and a soft knee is applied to voices or in the final mastering compression.

Output gainEdit

Because the compressor is applying a reduction in the output volume level, we'll need to compensate the loss of volume. With the output gain we can increase the output volume to give us a general sensation of volume increment.
Fig.3: Scheme showing how a compressor works. The continuous white line represents the input signal level and the output signal level without compression (bypass). The discontinuous white line represents the output compressed signal level. [A] Compressed with a hard knee. The compression begins only when the input signal surpasses the threshold level [B] Compressed with a soft knee. Notice that the compression starts to work as the input signal approaches the threshold.

A Practical ApproachEdit

Let's go with an example:
Open Psycle and insert a compressor (for the example I used GeneComp, Fig.4) and the jmdrum psycle plugin (JAZ's Drum Synth). Connect this generator to the compressor, and the compressor to the Master. For the generator I used the default preset (it's a bassdrum).
Fig.4 GENECOMP Mono VST compressor from Buzzroom

Apply the following settings in the compressor:

    Input gain:	0 dB
    Threshold:	       -6 dB
    Ratio:		11.6:1
    Knee:		11
    Attack:		1 msec
    Release:		21 msec or auto
    Ouput gain:	0 dB

Let's see how it affects to a non-compressed signal. The following figures (see Fig.5) represent the rendered bassdrum wave (C-4 note) of a non-compressed signal and a compressed signal with the above settings.
Fig.5: Example of how the compressor affects a signal. [A] Un-compressed signal. [B] Compressed signal. The discontinuous red line represents the curve evolution of the uncompressed signal peaks. The discontinuous yellow line represents the maximun peak of the compressed signal. [C] Representation of the compression settings over the compressed signal.

Now let us analyze the differences:

Threshold -6 dB 
As you know, it is the level where the compressor begins to work. If you take a look at the figures, all the signal waves surpass that level except the last one (this last signal wave is not affected by the compression because it is below the threshold).
Ratio 11.6:1 
The level of compression. In our example for each 11.6dB of signal that surpasses the threshold level, the output signal only increases by 1dB. Look at the 8th signal wave in both figures. If you make an easy calculation (it is a simple relation), as the original signal wave reaches -1dB (take a look at the red arrow in the 8th signal wave in the Fig.5-C, and the same point in the non-compressed wave of the Fig.5.A), the result will be an output signal wave that reaches +0.4dB (that is, -5.6dB) over the threshold (-6 dB).
Knee 11 
It means how the compression is applied. This value means not an absolute hard knee (the compression is not applied suddenly), but this value is near to a hard knee. If you look at the Fig.5-C, the orange discontinuous line takes a softer shape from the starting attack point to the point were all the compression rate is applied (11.6:1).
Attack 1 msec 
The point where the compression starts. It is represented by a vertical discontinuous violet line in the Fig.5-C . As you can see the compression is applied quickly.
Release 21 msec or auto 
The time the compressor takes to stop compressing the signal after it has started to work. In our example, before the release time reaches 21 milliseconds, there's another new wave surpassing the threshold. In fact, there's no release up to the last but one wave.


Compressors have a wide range of aplications, but we could classify them in two big groups:

1. Wideband compressors (AKA Singleband compressors).
2. Multiband compressors.

In addition there are special dedicated compressors for certain purpouses.

Wideband CompressorsEdit

In a wideband compressor, the whole range of frequencies are treated with the same parameter settings at the same time (See Fig.6).
Fig.6: Flow in a single-band compressor

This kind of compressors (i.e., see Fig.1, Fig.2 and Fig.4) are not good for mastering a song or to compress an instrument with a wide range of frequencies (i.e. drum set with bassdrum, hihats, snare, toms...) because low frequencies have more power and dynamic range than high frequencies. Compression will have a collateral effect: regions where low frequencies eat high frequencies (i.e. you can hear how the hi-hats volume decreases when the bassdrum sounds). But maybe this could be an interesting effect for your drum lines ;).

Setting up a Wideband CompressorEdit

Multiband CompressorsEdit

Frequencies are treated separately with the same settings for each band, or with different compression parameters for each range of frequences (See Fig.7).
Fig.7: Flow in a multiband compressor

In addition we can found another important element in a wideband compressor, the cutpoints to select the different ranges of frequencies (See Fig.8.B.[4]). Usually, this kind of compressors have up to 2 cutpoints (in Hz) which determines 3 frequency ranges:

1. Low-Mid cutpoint: Limit between the Low frequency (See Fig.8.B.[3]) and the Middle frequency ranges (See Fig.8.B.[2]).
2. Mid-High cutpoint: Limit between the Middle frequency (See Fig.8.B.[2]) and the High frequency ranges (See Fig.8.B.[1]).

This 3 zones are enough to make a good compression in a master session, but as many ranges you can control as better results you can obtain (if you really know to select the right range of frequencies and work with them).
Fig.8 Two examples of multiband compressors
[A] Arguru's Stardust. The compression parameters (see [1] selection) are the same for all the frequency ranges.
[B] Buzzroom's GranComp3. The compression parameters can be selected for each range of frequencies (see [1], [2] and [3] selections).

As it is shown in Fig.8.B, there's a control of volume output for each band. Then a we can say that a wideband compressor can work as an equalizer too. This is another useful characteristic of this kind of compressors.

Setting up a Multiband CompressorEdit

Dedicated CompressorsEdit

There are dedicated compressors to manage dynamics for specific purpouses:

1. Limiters
2. Side-chain compressors
3. De-essers


Side-chain compressorsEdit



Table.1- Starting values for compression


Threshold (dB)



Attack (ms)

Release (ms)


Stereo Mastering

-5 to -10

1.5:1 to 3:1







3:1 to 10:1


3 to 5


Bassdrum 1 -5 3:1 Hard 1 Auto
Bassdrum 2 -7 8:1 Hard 1 Auto
Hi-hats -3 2:1 Hard / Soft 1 Auto
Snare 1 -4 3:1 Hard 1 Auto
Snare 2 -5 2:1 Hard 1 Auto
Drumroll -4 3:1 Hard 1 Auto
General -3 to -10 2:1 to 10:1 Soft Fast Slow
General -5 to -12 3:1 to 10:1 Hard / Soft Fast Fast
General -2 to -10 3:1 to 8:1 Hard 5 to 10 10
Synth Bass -4 4:1 Hard 4 to 10 10
Real Bass -2 to -10 8:1 Soft 4 to 10 10
General -10 to -14 2:1 to 8:1 Hard / Soft Fast Fast
Pads -5 3:1 Soft 9 8
Lead Riff -5 5:1 Hard / Soft 2 Auto


Here's an example of more extreme compression settings and how they can be used musically

Take note of a few things: I have a dummy plug inserted before the main compressor. This way it's easy to check visually the level of sound that's entering the compressor. If you move down the threshold slider of the main compressor while the tune is playing, you'll start to hear very clearly how the compressor affects sound (at under -20db or so). The higher you set the threshold, the louder the track gets, but the less clearly the big BOOM kick is separated from the rest of the mix. Play around with the attack and release when the threshold is way down to get a grasp of how they affect the compression. If you bypass the compressor for the breakbeats, you'll notice how they disappear into the mix. I've used loud compression and a reverb with a "kick back" effect to really bring the breakbeats to the front. There is also a compressor on the bassline, I have a highpass filter that resonates loudly with some notes that would get out of hand without compression.

It's always wise to have a master compressor before the main generator. This way the sound won't clip and the overall tightness of the mix can be adjusted.



  • Taika-Kim (mastering examples, some spell-checking)
  • TranceMyriad (just a couple of minor grammar/spelling changes)

Page VersionEdit


Ad blocker interference detected!

Wikia is a free-to-use site that makes money from advertising. We have a modified experience for viewers using ad blockers

Wikia is not accessible if you’ve made further modifications. Remove the custom ad blocker rule(s) and the page will load as expected.