A function offered by many compressors is the ability to equalize your compressor’s sidechain signal… But why would this be beneficial and when might you need to use this function? In this article, we’ll break down the reasons to equalize your compressor’s sidechain signal.
Understanding a compressor’s sidechain
First, let’s briefly look at what a ‘sidechain’ is. When a signal enters a compressor, a copy of the signal is fed to the compressor’s ‘sidechain’ circuit. The sidechain analyses this copy of the signal to determine when the compressor should apply compression. Some compressor’s allow you to feed a signal from a different track into the sidechain to create sidechain compression. Regardless of which signal feeds the sidechain, many compressors offer you the ability to equalize the signal that the sidechain sees using built in filters.
When equalizing your compressor’s sidechain signal, it’s important to note that applying this equalization does not work in the same way that inserting an EQ on a track does. We do not hear the EQ move directly, although a change in tonality is often likely as we will see later on. Instead, these EQ moves only alter the version of the signal that the sidechain sees.
Equalizing your compressor’s sidechain creates a more uniform response across the frequency spectrum
So why might you choose to equalize your compressor’s sidechain signal? One reason is that compressors are generally more responsive to low end frequencies. So with many instruments, the lower the note, the more the signal excites the compressor.
Let’s look at an example of a vocal track. The lower notes being performed will be compressed more than the higher notes. The same is true of bass guitars and many other instruments. This has the potential to make things sound somewhat unbalanced. Further still, compression which is triggered by low frequencies is applied to the signal’s entire frequency spectrum. For signals which contain a broad range of frequencies, this can have a dulling effect as the compression being triggered by the low end will compress the high end too.
In these instances, equalizing your compressor’s sidechain signal allows you to cut some of the low end that the sidechain sees. Taking away some of this low end which would normally trigger larger amounts of compression makes the compressor respond in a more unified way across the frequency spectrum. The compressor will still apply compression to all of the track’s frequencies, but you are making the compressor less sensitive to some of the low end frequencies.
Equalizing your compressor’s sidechain allows you to attenuate problem frequencies
Another use of sidechain equalization is to make your compressor respond more to selected frequencies than others. You do this by boosting the frequencies in the sidechain signal. This is often used to create a ‘de-esser’ for vocals. To do this, you boost the band of frequencies where the sibilant sounds are occurring. The sibilant sounds are the harsh S and T sounds. With these sibilant frequencies boosted, the compressor will react most to those frequencies. This helps to reduce the track’s sibilance. The whole frequency range of the signal will still be compressed, but you are making the compressor more sensitive to your selected frequencies.
Conclusion
So as you can see, equalizing your compressor’s sidechain signal allows you to take full control over the way your compressor reacts to a signal’s frequency content. This allows you the ability to really tailor the way that your compressor reacts to different sounds.
Do you equalize the sidechain signal on your compressor? If so, what do you use this process for?
2 thoughts on “How and why to equalize your compressor’s sidechain signal”
Such interesting explorations of compression! Yes, I equalise my side chain (in Fabfilter Pro C2) for vocal drama work. A bump in lower mid range causes a more realistic balance between warm conversational tone and harsh shouting.
I think the shouty stuff has a higher peak value, so tends to be held too low. Maybe an option of rms sensing, or a combo of that with peak, would be a better way. Unsure of Fabfilter’s design in that respect, and when I asked they wouldn’t quite say!
Hey Howard, thanks for your comment. That’s really interesting. I’ve only occasionally worked on audio for motion picture / voice actor type stuff, but on the occasions that I have, I’ve certainly found recordings of conversations to need different treatment to recordings containing shouting. My approach was to use automation to lower the compressor’s ratio, and sometimes to raise the compressor’s threshold during the moments that contained shouting. That way, the shouting underwent less compression, which seemed to retain the realism whilst still keeping things under control.