Sidechaining

Sends 2: Side-chaining

What is Side-chaining?

Most effects, choruses, flangers, phasers, eqs, distortions etc react in exactly the same way regardless of what audio you feed them. There are effects, however, whose behaviour changes according to various qualities inherent in the audio they are processing. These include compressors, expanders and noise gates. When you feed audio into a compressor for example, the compressor only acts when the audio exceeds a certain volume, turning it down by the designated proportion.

there are many applications where we want to have this type of effect impose their changes on another piece of audio entirely however. It is possible to use a compressor to fade background music down to make way for a voice-over for example. If one sets the threshold such that most of the talking will trigger the compressor, by having it act on the music instead of the voiceover, the music will rise and fall in time with the cadences of the speaker. The process by which this is achieved requires sends to implement and is called side-chaining. This is because instead of the sequence being audio, compressor result, in a horrizontal line, the compressor is connected across the audio so that its input connects to audio A and its output connects to audio B.

the nitty gritty: How to Make it Happen

For this guide, we'll stick with the music-voice-over example and create what is known as a ducker.

Step 1: create two tracks in Reaper, naming them according to their respective purposes - voice-over and music and import the relevant media.

Step 2: Select the voice-over track and create a send so that the audio on that track is sent to your music track:

• Press i to activate the routing dialogue for the voice-over track.
• Tab to "Add New Send" and select the music track and press tab so that the selection takes.

Step 3: If you are confident with adding regular sends, feel free to skip this step. If you wish, you could exit the dialogue, solo the music track and play the project to confirm that the send has worked. If it has, you should be able to hear the voice-over and the music fighting with each other in the mix. Unsoloing the track will mean that the voice-over becomes louder, but this may cause clipping.

step 4: In order to achieve the desired effect, we must make the voice-over inaudible when track 2 is soloed by sending it to channels three and four of our music track, though, by default, tracks in Reaper only have two channels.

If Using a Keyboard with an Applications Key:

1. Inside the routing dialogue for the voice-over track, Tab until you find the delete button for the send to the music track. Paying attention to which send you are on and labeling tracks well becomes important in larger projects.
2. Press the applications key, which will bring up a context menu.
3. Cursor down to "Destination Audio Channel" and press enter, which will invoke another menu.
4. Cursor to "New Channels on Receiving track" sub-menu, and open it.
5. Cursor to "3/4" and press enter, which will not only change the music track to a 4-channel track, but direct the sent output of the voice-over track to that channel pair.

If using NVDA and a desktop computer layout without an applications key

1. From inside the routing dialogue for the voice-over track, Press NVDA numpad 8 three times to get to the top level of the routing window.
2. Navigate to the information about the correct send by pressing numpad 9. Paying attention to which send you are on and labling tracks well becomes important in larger projects.
3. Having found the correct send, navigate to the first instance where it says "Audio:/ 2e 1/2"
4. Navigate by word, numbed 6 until you reach "1/2". Then press NVDA numpad enter, which will bring up a context menu.
5. Cursor to "New Channels on Receiving track" sub-menu, and open it.
6. Cursor to "3/4" and press enter, which will not only change the music track to a 4-channel track, but direct the sent output of the voice-over track to that channel pair.

When soloing the music track, you should no longer be able to hear the voice-over.

Step 5:Add recomp, the compressor that comes with Reaper, to the music track.

Step 6: Ensure the Reaper UI for the plug-in is exposed. Whether or not the UI appears by default will depend on your preferences.

Step 7: tab until you find the "Detector input" combo box and press cursor down to highlight "Auxillary input.

Step 8: You will then need to configure the compressor to taste. An explanation of how to configure Reacomp is beyond the scope of this article but the crucial parameters with which to experiment are threshold, ratio, attack, release and precomp. An attack of 50 ms, release of 500 ms and precomp of 10-20 ms may be a good starting point, but these are very rough figures.

If the operation has been successful, you should hear the music rising and falling along with the speech. Even if you solo the music, you should hear it fading up and down even though speech is inaudible.

Musical Applications of Side-chaining

If you set the parameters of your compressor correctly, you can make your compressor into an invaluable tool for uncluttering the low end of a mix or even turn it into a musical instrument.

Producers often use side-chaining to duck a bass-line slightly to make room for the kick drumb, to avoid an over-cluttered low end, which produces a muddy sound. Producers of techno have taken this technique to extremes and make the kick sound even more powerful by causing it to duck the rest of the instruments.

Two excellent examples of this are Good Feeling by avicii, in which the piece starts with instruments being rhythmically ducked by something we cannot actually here, and Titanium by David Guetta, where the kick starts by ducking the filtered sweep at the end of the first verse and really gives the song its power after the first chorus has finished.

Side-chaining with Gates

We have seen how compressors can be used to fade something up and down based on the behavior of something else. If you use a noise gate instead of a compressor, you can use one audio source to make another cut in and out as desired, perhaps in rhythmic stuttering patterns.

simply find an audio source that will trigger your gate in the desired pattern at the desired tempo and link the output of your gate to the audio which you would like people to hear in the same way as described in the compressor example above.

Con Spirito by CJ Bollandis an example]]

Keyed gating, as this side-chained method of gating is called, doesn't have to be so dramatic though. Gated reverb, A more subtle technique, is described by [Roey Izhaki in his book, Mixing Audio: concepts Practices and Tools. It was reputedly discovered during the recording of Peter Gabriel's third album.

In a more refined form, the gate is inserted after the reverb and configured such that it opens for the initial reverb burst but closes to trunkate the tail of the reverb to remove inaudible and therefore useless mud from a busy mix. A compressor is put between the reverb and the gate in the effects chain to increase the portion of the reverb that falls above the threshold, thereby changing its decay envelope. Since this might make the reverb tail too long, the uncompressed version of the audio is fed into the input of the gate and the compressed reverb is tamed at the gate's output so that its exaggerated quality is retained, cutting through the busy mix, whilst not cluttering it by hanging around too long. [Roey Izhaki, 'Mixing Audio: Concepts Practises and Tools', Taylor & Francis, 2 May 2013.]

These are just a couple of examples of the applications to which side-chaining can be put and, in turn, the uses of sends detailed in this courses wiki only scratch the surface. the best way to internalize the information and build sends into your work-flow is to practice and experiment.