side chains
side chains
I would like to start a thread discussing side chains. I feel like side chains are the cousins at family reunions that you don't know anything about and you see them once every 5 years. I am open to hearing how everyone uses them, what they use them on and when it is pratical to use a side chain. I personally don't have much knowledge on them, so Im seeing what everyone else has to say.
-
Professor
- ghost haunting audio students
- Posts: 3307
- Joined: Wed May 07, 2003 2:11 pm
- Location: I have arrived... but where the hell am I?
Well let's just start with the basics of what a side chain is - that might make some applications pretty obvious, and explain why they appear on some items but not others.
Put simply, a side chain is an additional input into a device that is used to trigger the process in that device. The typical unit that uses them would be a compressor. In a normal compressor, the audio signal you intend to process goes in, it activates the compression circuitry which processes that signal and sends it back out. The side chain allows that same audio signal to enter, be processed, and leave the compressor, but uses a different signal to control the process.
From there, it helps to have an example. Let's say I'm compressing my lead vocals. The vocal signal enters the compressor, is processed and leaves. When the signal gets loud, the compressor holds it back, and when it's soft, it brings it up - what a compressor is supposed to do. But let's say I split the vocal track and feed one side of the split into the compressor as usual and then send the other split through an equalizer and into the side chain input. If the EQ is neutral then nothing should be different because the same signal is controlling the process. But if I start cranking up a peak in the upper mid-range between say 7-9k then I'm increasing the volume of the sibilant sounds, the "S" and "T" sounds on the voice. Because those sounds are louder as they are fed into the side-chain, they activate the compressor circuitry faster (because the compressor responds to the volume level entering the circuit) and so it will turn down the output until the signal goes away. That's what we call a "de-esser" because it removes "S" sounds.
That's also about the most common approach to side-chaining that there is, splitting the signal, routing through an EQ circuit, and using the altered signal to control the response of the compressor. We don't want to hear the signal feeding the side-chain in our mix(which is why there is no side-chain output except to link multiple compressors) although we might listen to it briefly to know what we are cutting or boosting. And if you see a compressor with a built-in EQ/side-chain, it works the same way.
The same arrangement can work in reverse too. Consider dipping a frequency that you would like to stand out - that means higher or lower frequencies will be held back as they happen, but the one you want to stand out will be pushed a little harder before the compressor starts to squeeze it. Like if you want the brighter sounds of an acoustic guitar to pop out of a mix, you could turn them down a little in the side-chain. Then when the guitar line is playing in the lower range it will be compressed more, but as it moves higher it will be compressed less, and able to ring through the mix.
That one is a lot less common, and you have to be careful that the signal running through the compressor doesn't overload things as it goes.
The really cool thing though is that any old signal can be plugged into the side-chain to activate the compressor. If you feed a constant sine wave in, then you'll have a constant amount of compression... which kinda defeats the purpose of an automatic volume control, doesn't it? But consider the possibility of pushing the bass a little harder every time the kick drum hits. You simply feed the bass through the compressor as usual and set the controls so it is being push just a little bit beyond where it is sitting naturally. Then feed the kick into the side chain, and maybe make an adjustment to the threshold, and you can give the bass an extra push whenever the kick drum lands so the tone & harmonic foundation aren't lost down there.
Or try the reverse of that and setup a compressor on an acoustic guitar track that compresses with little or no makeup gain. Then plug a split of your lead vocals into the side-chain and whenever the vocals are going, the guitar backs off a little, but comes right back up as the vocals drop out. We call that one a "ducker" because it ducks the volume on one source to give priority to another. It's actually used all the time in courtrooms, boardrooms, etc. so that when the judge or chairman speaks into his microphone, everyone else takes a back seat. Though the settings there are usually a little more severe, like a 10dB or 20dB duck when someone is speaking while you might use only 3dB in a mix. Oh, you'll notice these at restaurants too when the music dips as they call somebody's name over the PA.
All sorts of fun stuff to try.
-Jeremy
Put simply, a side chain is an additional input into a device that is used to trigger the process in that device. The typical unit that uses them would be a compressor. In a normal compressor, the audio signal you intend to process goes in, it activates the compression circuitry which processes that signal and sends it back out. The side chain allows that same audio signal to enter, be processed, and leave the compressor, but uses a different signal to control the process.
From there, it helps to have an example. Let's say I'm compressing my lead vocals. The vocal signal enters the compressor, is processed and leaves. When the signal gets loud, the compressor holds it back, and when it's soft, it brings it up - what a compressor is supposed to do. But let's say I split the vocal track and feed one side of the split into the compressor as usual and then send the other split through an equalizer and into the side chain input. If the EQ is neutral then nothing should be different because the same signal is controlling the process. But if I start cranking up a peak in the upper mid-range between say 7-9k then I'm increasing the volume of the sibilant sounds, the "S" and "T" sounds on the voice. Because those sounds are louder as they are fed into the side-chain, they activate the compressor circuitry faster (because the compressor responds to the volume level entering the circuit) and so it will turn down the output until the signal goes away. That's what we call a "de-esser" because it removes "S" sounds.
That's also about the most common approach to side-chaining that there is, splitting the signal, routing through an EQ circuit, and using the altered signal to control the response of the compressor. We don't want to hear the signal feeding the side-chain in our mix(which is why there is no side-chain output except to link multiple compressors) although we might listen to it briefly to know what we are cutting or boosting. And if you see a compressor with a built-in EQ/side-chain, it works the same way.
The same arrangement can work in reverse too. Consider dipping a frequency that you would like to stand out - that means higher or lower frequencies will be held back as they happen, but the one you want to stand out will be pushed a little harder before the compressor starts to squeeze it. Like if you want the brighter sounds of an acoustic guitar to pop out of a mix, you could turn them down a little in the side-chain. Then when the guitar line is playing in the lower range it will be compressed more, but as it moves higher it will be compressed less, and able to ring through the mix.
That one is a lot less common, and you have to be careful that the signal running through the compressor doesn't overload things as it goes.
The really cool thing though is that any old signal can be plugged into the side-chain to activate the compressor. If you feed a constant sine wave in, then you'll have a constant amount of compression... which kinda defeats the purpose of an automatic volume control, doesn't it? But consider the possibility of pushing the bass a little harder every time the kick drum hits. You simply feed the bass through the compressor as usual and set the controls so it is being push just a little bit beyond where it is sitting naturally. Then feed the kick into the side chain, and maybe make an adjustment to the threshold, and you can give the bass an extra push whenever the kick drum lands so the tone & harmonic foundation aren't lost down there.
Or try the reverse of that and setup a compressor on an acoustic guitar track that compresses with little or no makeup gain. Then plug a split of your lead vocals into the side-chain and whenever the vocals are going, the guitar backs off a little, but comes right back up as the vocals drop out. We call that one a "ducker" because it ducks the volume on one source to give priority to another. It's actually used all the time in courtrooms, boardrooms, etc. so that when the judge or chairman speaks into his microphone, everyone else takes a back seat. Though the settings there are usually a little more severe, like a 10dB or 20dB duck when someone is speaking while you might use only 3dB in a mix. Oh, you'll notice these at restaurants too when the music dips as they call somebody's name over the PA.
All sorts of fun stuff to try.
-Jeremy
Another relatively common use of a side chain is the "keyed gate." It's much the opposite of the "ducker." Run a guitar signal into the main input of a noise gate or expander with a side chain (rarely called "key") input. Run a percussive type of sound, say a submix of the kick and snare, into the side chain input. With a little tweaking of threshold and release settings you've now got an extremely versatile "tremelo" effect on the guitar. It will follow the overall beat of the song, or can be used to create quite complex pulse patterns.
Run a white noise signal through the gate, and hi-hat or snare into the sidechain to add a little extra "sizzle" or complexity to the top end of that instrument.
Run a white noise signal through the gate, and hi-hat or snare into the sidechain to add a little extra "sizzle" or complexity to the top end of that instrument.
Works wonders on kick drum too with a low frequency sine wave to add a little more low end to your bass drum sounds.ashcat_lt wrote:Another relatively common use of a side chain is the "keyed gate." It's much the opposite of the "ducker." Run a guitar signal into the main input of a noise gate or expander with a side chain (rarely called "key") input. Run a percussive type of sound, say a submix of the kick and snare, into the side chain input. With a little tweaking of threshold and release settings you've now got an extremely versatile "tremelo" effect on the guitar. It will follow the overall beat of the song, or can be used to create quite complex pulse patterns.
Run a white noise signal through the gate, and hi-hat or snare into the sidechain to add a little extra "sizzle" or complexity to the top end of that instrument.
-
Professor
- ghost haunting audio students
- Posts: 3307
- Joined: Wed May 07, 2003 2:11 pm
- Location: I have arrived... but where the hell am I?
Yeah, I hadn't written about gear with side chain inputs other than compressors, but gates have some pretty cool uses too.
The kick drum + sine wave one is one of my favorites. I process and mix in the kick drum as usual but if it seems like it needs a little more tone or if it needs a little more resonance and ring because it was dampened too much, adding a tone helps. I kick on a signal generator, set it to a sine wave and find the frequency I'm looking for in about the 70-100 range. I try not to go any lower since that would add a frequency which many systems can't reproduce, though I would consider something like one sine wave at 70 with another a little softer at 35 or something like that. Either way, the tone is fed out its own channel and passed through a gate. The real kick track is then split and fed into the side-chain or "key" input for the gate and the threshold is set so the kick drum hits trigger the gate to open and let the tone through into the mix, but with some care to make sure bleed from the snare or toms doesn't trigger it by mistake.
From there, the attack and release times will sculpt the sound to create a more realistic or completely fake sounding drum. I find that somewhere in around 3-6ms on the attack will allow the real kick to be the first sound heard, and then I just set the decay time as long as I'd like the "drum" to sustain. When I hit it right, it's all but impossible to notice that the tone is there until I turn it off and the kick turns into a dry slap of a sound.
A similar trick can be used to give you the classic 80s gated snare sound. Here a ridiculous & over the top reverb patch is fed into the gate and the snare is fed into the mix and split into the sidechain. Then each snare hit opens the reverb which adds lots of density and complexity to the snare sound, but it is then closed back off before the length of the decay gets messy or muddy. Again, the attack and release times will sort out how clear & clean the attack is and how long the effect stays in the mix.
Fun times all around.
-Jeremy
The kick drum + sine wave one is one of my favorites. I process and mix in the kick drum as usual but if it seems like it needs a little more tone or if it needs a little more resonance and ring because it was dampened too much, adding a tone helps. I kick on a signal generator, set it to a sine wave and find the frequency I'm looking for in about the 70-100 range. I try not to go any lower since that would add a frequency which many systems can't reproduce, though I would consider something like one sine wave at 70 with another a little softer at 35 or something like that. Either way, the tone is fed out its own channel and passed through a gate. The real kick track is then split and fed into the side-chain or "key" input for the gate and the threshold is set so the kick drum hits trigger the gate to open and let the tone through into the mix, but with some care to make sure bleed from the snare or toms doesn't trigger it by mistake.
From there, the attack and release times will sculpt the sound to create a more realistic or completely fake sounding drum. I find that somewhere in around 3-6ms on the attack will allow the real kick to be the first sound heard, and then I just set the decay time as long as I'd like the "drum" to sustain. When I hit it right, it's all but impossible to notice that the tone is there until I turn it off and the kick turns into a dry slap of a sound.
A similar trick can be used to give you the classic 80s gated snare sound. Here a ridiculous & over the top reverb patch is fed into the gate and the snare is fed into the mix and split into the sidechain. Then each snare hit opens the reverb which adds lots of density and complexity to the snare sound, but it is then closed back off before the length of the decay gets messy or muddy. Again, the attack and release times will sort out how clear & clean the attack is and how long the effect stays in the mix.
Fun times all around.
-Jeremy
Who is online
Users browsing this forum: No registered users and 49 guests