Wednesday, February 15, 2017

Side-Chaining and Ghost Notes Tutorial

I have another tutorial video for you.  This video is the last in a series of videos that Urban Francis and I put together, as a techno production project for a track called Hijack.

The track was produced in Ableton Live, and we have all of our Ableton project files available as a free download from Dropbox.

In this specific video, we wanted to explain what was going on with the Side-chain compression.  Eventually, the two of us are going to do a very detailed and complex set of tutorials about compression and side-chaining.  They’re very complicated and moderately advanced topics, and they take a little while to learn to do properly.  However, because we have some side-chaining happening in the Hijack project, we decided that it would be smart to give a brief overview, so any producers who are moderately familiar with Ableton would understand exactly what we’re doing here.

Here's a link to the video:

If you don’t understand what compression is, let’s start there, even though it’s a pretty complex topic.  I tried to do an explanation in my Mastering & Marketing Your DJ Mix video on YouTube, starting at the 21 minute mark.  It’s worth a watch, because of the graphic visuals.  However, let me also try to explain it here.

 A lot of people, when they hear about compression, they get the idea that compression makes things louder.  That’s incorrect.  Compression by itself does not raise the level of an audio signal.  There are several different settings in a compressor, but the main ones are this:

Threshold Value – this is the signal level above which audio is being affected.

Ratio – tells how much the portion of the audio that is above the threshold is getting squished.

Now obviously, if the audio is getting squished, then the audio is getting pushed down from its original volume to a lower volume.  So why, when people talk about compression, do 99% of them think that the audio is getting louder?  That’s a great question, and it’s something that a lot of people don’t understand properly.  First, you need to understand that compression BY ITSELF does not make the volume louder, it reduces the volume.

Now, let’s be honest – in the fight for volume to gain prominence in the listening space, very few people want to reduce the volume of their productions.  So that’s where another setting commonly associated with compression comes into play.  That setting is the make-up gain.

After compression is done, if a producer applies make-up gain to the compressed audio, it brings the volume back up.  A producer may set the make-up gain to the exact amount to bring the peak of the volume back up to where it was before the audio was compressed.  Sounds like a rather useless practice, doesn’t it?  Drop the volume, then raise the volume.

But the key thing is that when the volume is compressed, it is NOT dropped in a linear manner.  Louder parts are dropped more than quieter parts.  Hence the reason that I said that the audio gets squished.  Therefore, the audio now has a narrower dynamic range than before.  The peaks and valleys in the amplitude are not as significant; the changes from the loud parts to the soft parts are less significant.

Since the peak of the volume is unchanged from the original, once the makeup gain was applied, and since the dynamic range from loudest to softest is reduced, this means that the AVERAGE volume of the processed audio is louder.  And that’s the source of the misconception that compressed audio is always louder.  Compressed audio definitely has a narrower dynamic range.  If it is louder, that is only because makeup gain has also been applied.


Moving on to Side-chaining, the way that side-chaining works, is that you’re taking a signal from somewhere else, and you’re analyzing that signal, and you’re using the analysis of the side-chain signal to control the compressor’s actions on whatever other audio is being compressed.  So in other word’s, the compressor’s actions are being dictated by the energy of the side-chain signal (side-chain audio source) that’s being fed into the compressor.

Remember that the signal that’s being fed into the compressor (the side-chain signal) may have nothing at all to do with the audio that you’re processing.  And also, remember that you don’t have to have a signal being fed into the side-chain.  If you don’t have a side-chain signal, the compressor just works at a steady set of values while it’s processing audio.  Essentially, a side-chain compressor without a side-chain signal is just … a regular compressor.

Ok, so imagine this.  Let’s pretend that the signal that is coming in to the side-chain input of the compressor is some sort of signal that is steady and repeating.  High energy/amplitude (lots of signal) alternating with low energy/amplitude (quiet).  A good example might be a repeating kick drum.  If you keep hitting a kick drum in a steady pattern regularly, then the “signal” (amplitude) is essentially a constantly repeating on/off/on/off/on/off/on/off pattern.  If this signal is what is controlling the compressor, then you’re basically turning the compressor on/off/on/off/on/off/on/off in the same timing and pattern.

Many producers DO use a kick drum as a side-chain signal, so that a side-chain compressor keeps turning on and off in time with the kick drum.  The compressor may be processing just about anything else while it keeps turning on and off.  Maybe it’s processing a synth line.  Maybe it’s processing a background pad.  Doesn’t matter.  It’s compressing some sort of audio, and it’s turning on and off in time with the beating of the kick drum, since the kick drum is being used as the side-chain signal.

A producer can use the kick drum from the track they’re working on, as the side-chain.  But they don’t have to.  They could use a different kick drum track as the signal.  And that other kick drum which is being used as a side-chain signal to the compressor does NOT necessarily have to be audible to listeners of the track.  So a producer could easily have two kick drum tracks in their song:  one that is audible that the audience can hear, and a second one that is not audible to listeners, but which is being used as a side-chain signal.

Why bother using two different kicks?  Maybe the producer wants a signal with a fast release for the side-chain (to turn the compressor off quickly), but doesn’t want the listeners to hear a kick that has a fast release.  Or maybe the muted side-chain kick needs to act as a side-chain signal during parts of the song where there is no kick drum playing.

If a producer brings an extra kick drum (or any other type of instrument or midi notes) into a side-chain compressor, but that audio/midi side-chain signal is not audible as part of the project, the side-chain audio signal is called a ghost signal, or ghost notes.

Using a side-chain to turn your compressor on and off allows you to compress audio when needed, to allow for a different part to shine through a mix, but does not compress the audio at other times.  This gives the producer a lot of flexibility.  It’s much better than if the producer’s only option was to permanently decrease the volume of other parts of the mix.

An advantage of using ghost notes or a muted side-chain signal is that you can have it turn the compressor on and off in any pattern or timing that’s convenient to you.

An advantage of using a regular audio source within the song (one that the listeners hear) as a side-chain signal to affect another track is that the timing may be perfect for your needs.  For example, if you need a kick drum to cut through a synth line, you can play the audio of the kick, and you can simultaneously route that auto into the side-chain of a compressor that cuts the volume of the synth, allowing the audio version of the kick to cut through the mix.

By the way, if you want to see the project that we used as the basis for this tutorial, here's the music video (yes, the full project files for this track are available as a free Ableton or stems download):

Hopefully this all made sense.  For more information about the video tutorials that I have to offer, visit this page:

Thanks for your time and interest!

- Jonathan Clark (DJ Bolivia)

Follow Jonathan Clark on other sites:
        Main Site:
        Music Blog: