INTRODUCTION

1- Theory:

MAIN PARAMETERS OF DYNAMIC COMPRESSION:

A- THRESHOLD:

In practice, the compressor is active only from the time the audio signal reaches a certain level called the threshold, it is expressed in dB (decibels) and is adjustable using the button "Threshold"

Below this threshold, the compressor is not running, the sound enters and exits the compressor without any change in the level. The dynamics of the audio signal exiting the compressor via the output (Output) is the same as that of the signal that enters the compressor through the inlet (Input)

What happens if the input signal exceeds the threshold?

The compressor becomes active and can then perform the compression of the dynamic as we will see below.

B- THE COMPRESSION RATIO:

The compression ratio, or ratio, expressed as a fraction x: y represents the amount of compression that will suffer the signal. The more this fraction will be larger, the more the sound will be compressed.

The ratio x: y means that whenever the threshold is exceeded by x db, the compressor will let only y db.

Formula:

Let I, O, S, R, x and y such that:

I = Input level

O = Output Level

S = Threshold

Ratio = x: y

The formula that expresses the relationship between the various parameters of the compression is as follows:

O− S= y(I− S) / x

Practical examples:

Exemple 1:

Threshold: -20 dB and ratio: 2:1

A 2:1 ratio means that whenever the signal exceeds the threshold of 2 dB, the compressor will let only 1 db. If the input level is equal to -18 dB, the compressor will reduce the output to -19 db. If the input level is equal to -10 dB, the compressor will reduce the output to -15 db. If the input level is 0 dB, the compressor will reduce the output to -10 dB

Exemple 2:

Threshold: -10 dB and ratio: 4:1

A 4:1 ratio means that whenever the signal exceeds the threshold of 4 dB, the compressor will let only 1 db. If the input level is equal to -6 dB, the compressor will reduce the output to -9 dB. If the input level is equal to -2 dB, the compressor will reduce the output to -8 db. If the input level is 0 dB, the compressor will reduce the output to -7.5 db

C- ATTACK:

In milliseconds, this parameter represents the time required for the compressor to go into action. It will have a great influence on the attack (and thus the timbre) of the recorded sound. Try to experiment ...

Too short Attack time in some cases will tend to flatten the signal, and to diminish the percussive aspect, whereas an attack time too long will affect the quality of the compression ...

D- RELEASE:

In milliseconds, this parameter represents the time required for the compressor to recover the stand by once the threshold is no longer reached. It will have a great influence on the maintenance (and thus the timbre) of the recorded sound. Try to experiment ...

E- MAKE-UP:

Also called "gain compensation" and often with an adjustable knob, it can increase or decrease the level of the output signal (after compression) this setting is indisponsable at major compressions.

2- Practice:

COMPRESSION IN ADOBE PREMIERE PRO:

In Adobe Premiere Pro ( in any editing software as well), it is surprisingly easy to apply compression to an audio track, just click on the audio track you want to compress, and apply a compression effect (via the bin "Audio Effects", see photo below)

Various options are available: Mono, Stereo or 5.1. By selecting "Stereo" for example, we access all stereo audio effects available, just choose your favorite compressor. For information, hundreds of free VST compressors are available on the net.

For this article I chose a basic compressor called "Dynamics", by clicking and dragging to place it on my track, I can access the settings of various parameters.

COMPRESSOR INTERFACE:

Here I will focus only on the compressor section (middle section), the other parts, AutoGate and Expander sections may be seen in futur articles.

The compressor is enabled by default, if this is not the case, activate it by checking the "Compressor"

And here we find the different commands we have seen in the theoretical part. As a reminder, Threshold is the level at which the compressor becomes active, ratio corresponds to the importance of our compression, the attack correspond to the reactivity of the compressor.

A small horizontal bar, graduated from 0 to -40 db indicates the importance of the compression (and thus: the reduction of the level) ..

Ok, but where to start?

Take the case of a voice-over, I guess it was accurately recorded, ie that the microphone was located at a safe distance, not too close to avoid distortion, the sounds of mouth and other side effects, and not too far away to avoid ambient noise, and to keep a voice as clear as possible.

Depending on the type of microphone used, the power of the voice and location of the sound, the microphone should be set at a distance between 20 and 40 cm (more or less)

Another key parameter is the gain of my record, it is controlled by the device that I use for the sound (camera, portable recorder, mixer ...). The basic rule is: "a too high gain will produce an unrecoverable distortion, a too low gain will produce noise (breath noise)"

To compress this track, I will first set the Threshold at -20 dB, the ratio to 2:1 and the Attack at 0.1ms.

This represents a starting point quite correct. Depending on the outcome and the desired result, I will change my settings, and especially I will not hesitate to experiment with extreme settings, because it is the best way to understand the action of each parameter of the compressor.

I listen carefully to what happens:

- If the compression is not sufficient, ie, if I still have too much differences in levels, I lower the Threshold to -25 db and / or I increase the ratio to 4:1 , and so on ...

- If the compression is good but you find the sound too flat, it's because the Attack is too short, it will gradually increase the setting until i have the desired sound.

- If the compression is excessive, increase the Threshold to -15 or -10 db, and / or decrease the ratio ... etc.

CONCLUSION:

A compressor is not intended to resolve all audio problems, in some cases it is more appropriate to combine it with an equalizer, or even to refrain from using it, because if not properly set it can cause a whole series of problems that can deteriorate the quality of your recording.

To use the compressor properly, it is important first to identify the problem to treat, was the voice recorded correctly? was the microphone at the right distance,... etc.

After this, correct speakers and a good pair of ears will be your only guide to properly parameterize your compressor.

Voili voilou, I hope that this article has introduced you to the dynamic compression.

Hicham.

N.B: This is the online assisted traduction of the french version...Sorry for the style and for the mistakes