Compression I: Peak Reduction

In this first example, we discuss the very basic process of reducing peaks in a waveform.

Compression effects are frequently used to reduce loud spikes in a track, and this can help a sound fall back into the overall mix less obtrusively. The snare drum in this example track is particularly loud. In most cases, you are going to measure the loundness of one track against the whole, but for the sake of illustration, we are going to reduce the volume of the snare in the drum track alone. Generally, drums are supposed to have a lot of dynamic range, this is what allows them to cut through a mix with punchy bass and crackly attack, but let’s say the track needs to fall in line a bit. We can use some compression to drop the volume of the snare without affecting the overall volume of the drums by reducing dynamic range.


In order to reduce the snare alone, and not the overall volume of the drums, I used the following settings:


The threshold of a compression effect determines at what point the signal is going to be effected. (Threshold = the “where” compression happens.) Everything above the threshold will be compressed according to the ratio you set. In this example, I have dropped the threshold back to -13.5db. This means that everything above -13db in the drum track will have a compression rating according to the ratio. A smart way to set the threshold for downward compression is watch the level meter for the source material, and set the threshold just above the average volume of the whole track. The result will be a track that “falls back” into the mix.


The ratio setting determines how many decibels of input amplitude are required to give 1db of rise in output (plus the gain we add in the end). (Ratio = the “how much” compression happens.) The ratio sets the angle of slope the source waveform will take as it rises. In this case we do not require a lot of ratio to give the desired effect, so I only use a 4:1 ratio above -13.5db. Adjusting the amount of ratio in downward compression will eventually result in a “limiter” style of effect (as we will see in the next post). This means that as the ratio approaches infinity, the volume simply will not rise above the threshold at all, and you have effectively “limited” the track to the threshold you have chosen.


Attack determines the time it takes for the compression to kick in. (Attack is the “when” the compression happens.) This setting will be extremely important in determining whether you want more or less cut-through for your tracks. So, “attack” is aptly named because that is essentially what it will add. In this example, we are trying to remove the audible attack of the snare, which is a fast instrucment. So we reduce the “attack” level so that it happens as immediately as possible. If we set a longer attack, the initial peak above the threshold (in this case the snare hit) that triggers the compression remains the same, but the effect does not kick in until the attack amount has been reached. This would be a god option if you wanted to add dynamics to the mix on instruments without natural dynamics. Large string instruments or an overcompressed electric bass, for example, could benift by adding more attack, or time before compression.
Other effects may use a slider for attack from “fast” to “slow”. Fast/less attack results in immediate compression, whereas slow/more attack adds a time delay on when the compression occurs.


Gain sets the volume needed to add back to the track after compression. This is why it is sometimes called “make up gain” as opposed to “input gain” which can also sometimes be adjustable. In most cases in which you are using compression for peak reduction, you will want to adjust the gain setting to get close to the original volume because the reason you have compressed in the first place was to change the dynamic peaks, but not the overall volume of the rest of the track.


By compressing the track at a 4:1 ratio above -13.5db using a fast attack, the volume of the snare drum in this drum track was reduced, seemed to recede into the mix. Notice how the compression has the additional effect of adding more “room” sound as well. Not only was the punch of the snare gone, but by flattenting out the sounds above 13.5db, we were also able to give more attention to the snare transient sounds like reverb and ring.

Leave a Reply

Your email address will not be published.