Effects sends and returns are popular routing options in both hardware and software (DAW) mixers, affording us independent control over our “dry” unprocessed tracks and our “wet” return channels. In the case of modern digital audio workstations, where our track counts can be exceedingly high, the creative potential of effects sends and returns is practically unlimited.
To take full advantage of the creative potential in our effects returns, we must understand how we can separate the dry and wet signals, how we can then process the “dry” and “wet” signals differently, and how we can utilize complex routing to achieve specific and experimental effects.
In this article, we'll dive deep into the creative use of effects sends and returns and how you can achieve incredibly powerful effects using these mixer tools.
Disclaimer: This article includes affiliate links. If you click one of them, I may receive a small percentage of the sale at no extra cost to you (which I'm very grateful for, as it helps me produce more free content here at My New Microphone). You can see the list of my partners here and my ethics statement here. Thank you for your support!
Understanding Effects Sends And Returns
Before exploring creative applications, it's crucial to grasp what effects sends and returns are.
Essentially, they allow you to send a portion of a track's signal to an auxiliary channel for processing, independent of the track's main output. This technique is particularly useful for time-based effects like reverb and delay, as well as for parallel processing techniques like parallel compression and parallel saturation.
The “send” is the option to actually send audio from a track to a separate auxiliary channel.
A send will have a specific bus (the signal path) for the signal to be sent to, along with a potentiometer to control the amount of level being sent. Additionally, the point at which the audio is sent from the original track can be selected between a few different options:
- Pre-fader: the audio signal is sent directly after the channel's inserted effects but before its pan pot and volume fader. In other words, the fader and pan positions of the track have no influence over the send level.
- Post-fader: the audio signal is sent directly after the channel's fader but before its pan pot. In other words, the send level is dependent on the track's fader level (as well as the send level) but not on the track's fader position.
- Post-pan: this option isn't always available (Logic Pro X offers it), but it's a point at which the audio signal is sent after the fader and the pan pot.
- Pre-FX: this is another option that isn't always available (Cuckos Reaper offers it), but it's a point at which the audio signal is sent before the track's inserts (the raw input audio) and certainly before its fader and pan pot.
Auxiliary sends are awesome because they're independent of the output of the track. We can choose a send bus and route audio to some place other than the output (typically in addition to the output).
A bus is simply any kind of audio signal path that allows the combination or “summing” of different signals to be routed/processed together.
A single track can have multiple be sent to multiple buses, and multiple tracks can be sent to the same bus.
The “return channels” are fed by such busses (their outputs are set to receive the audio being “sent”), and the outputs of these channels are generally routed to the mix bus or stereo output.
The real fun comes when we begin sending audio from certain return channels to other return channels, creating a chain of events with our effects or perhaps weaving complex routing networks, sometimes even with feedback loops.
That's what we'll be diving into in this article.
I wanted to give a brief primer before we got into the more advanced concepts, so we won't go into more detail on what these sends and returns are. However, if you would like more information, I put together an in-depth article on the subject:
Check out Mixing/Recording: Auxiliary Tracks, Sends & Returns for more information.
With that, I should state that the creative use of effects sends and returns opens up a world of possibilities in music production. By experimenting with these techniques, you can add depth, texture, and dynamics to your mixes, pushing the boundaries of music production.
Remember, the key to mastering these techniques lies in experimentation and practice. So, dive into your digital audio workstation, try out these strategies, and discover the vast sonic landscapes you can create.
Other Ways Of Achieving Split Processing
We will be focused on utilizing mixer sends and returns to our advantage largely because they're so universal. However, I should briefly mention that there are other ways of achieving this “split processing”.
For example, Image-Line FL Studio has the FL Patcher that allows for incredibly complex routing with MIDI, virtual instruments and effects.
As another example, Ableton Live allows for parallel processing inside its Effects Rack, which is a nice touch.
That's all well and good, and if you prefer to use those tools, by all means use them! However, the standard way of setting up effects returns is via auxiliary tracks in the mixer, so that will be the default mode of explanation for this article.
The Basics: Time-Based Effects Returns
Effects returns are perhaps most commonly used for time-based effects, namely delay and reverb.
These effects are commonly chosen to give a sense of dimensionality to the mix but can also be used for more creative sound design and special effects as well.
Having independent control over our dry signals (the tracks feeding into the reverb or delay return channel) and the “wet” return channel itself is a major benefit of setting up our routing that way.
Note that if the primary effect on the return channel has a dry/wet mix control, we generally set it to 100% wet since the track(s) sending to the return are 100% dry — that gives us the dichotomy. Secondary effects (those effects used to colour or otherwise alter the primary effect) can be mixed at whatever wet/dry mix sounds the best.
As an example, we could have a delay as a primary effect on a return channel (100% wet) and then a saturation plugin to help colour it. 100% wet on the saturator may be a bit too distorted, so pulling it back to mix some of the “dry” delayed signal might work best — always experiment and trust your ears!
Of course, we could insert our delays and reverbs directly on our tracks and adjust the wet/dry or mix controls ourselves. However, this way of processing has two major drawbacks:
- If we wanted several tracks to be sent to the same reverb or delay, we'd have to duplicate the effect across multiple channels. This takes additional effort, time, and computer resources and whenever we want to make any changes to the effect, we'd have to change each inserted instance of the effect.
- If we wanted to make changes to the tonality of the reverb or delay, we're really limited by what's inside the plugin itself (we won't be able to add additional processing that only targets the reverb or delay signal unless that processing is included in the effect itself).
And so, in general, it's best to set up our time-based effects on return channels and have independent control over them and what track or tracks feed into them.
The Basics: Parallel Processing
Parallel processing is set up pretty much the same as the aforementioned time-based effects, except for the fact that we'll generally be processing these returns with effects such as compression, saturation, distortion and others that do not really alter the timing of the audio.
These parallel tracks are then mixed back in with the original signal, generally as a way to improve the overall tone or “glue” of the original track or tracks feeding them.
Again, we could opt to set up our effects directly on our audio tracks and dial in the wet/dry or mix controls (if applicable to the effect in question). However, the same drawbacks I mentioned above apply to parallel processing as well.
And so parallel processing is also best incorporated by utilizing auxiliary sends and returns.
Parallel compression is perhaps the most common style of parallel processing. As the name suggests, it involves setting up an effects return, applying heavy compression, and then mixing it back underneath the dry track(s).
Parallel compression helps to thicken up the sound of our tracks and sounds awesome on drums and vocals.
Remember that if the primary effect on the return channel has a dry/wet mix control, we generally set it to 100% wet since the track(s) sent to the return are 100% dry. Secondary effects (there to colour the primary effect) can be set with varying wet/dry mixes if applicable.
Beyond The Straightforward Wet/Dry Mix
I touched on this already, but I think it's worth revisiting in its own section of this article: setting up a proper effects return gives us complete, independent control over our effects versus our dry signals.
Mixing is all about balance.
In some cases, you'll have your dry track(s) mixed as they should be before testing out some effects — in this case, you may need to shape up the return channels with a bit of EQ or maybe some saturation or some other effects. Having a dedicated effects return channel allows you to add such additional effects before or after the “main effect” (be it delay, reverb or some parallel-style processing).
In other cases, you'll be utilizing return channels as part of the production process to add character to your sounds — in this case, having total control over the processing of the “wet” audio will allow you to dial things in to your exact liking.
Panning Mono Effects Returns
Alright, getting into actually putting our effects returns to use, let's start simple with panning.
I especially love panning my effects returns when dealing with mono signals (both those being sent and the return channels themselves).
While there are nearly endless possibilities for panning our return channels relative to our dry tracks, there are two main ideas I want to touch on. I'll discuss both as if there's only one track being sent to the return, though you can expand on the routing if you'd like.
The first is same-side processing, where we'll pan our effect return channel to the same position as the track feeding it. Setting things up this way helps to solidify the position of the original track by adding depth to it in the same direction (with reverb or delay) or by adding tone to it (with parallel processing).
The second is opposite-side processing, where we'll pan our effect return channel opposite that of the track feeding it. This helps add width and dimension to the overall mix and the original track. It may not sound as “natural” as same-side processing, but it really helps fill out the stereo mix in a subtle way.
Personally, I find myself writing, producing or mixing a lot of material with two guitars. I'll almost always pan each guitar to its own side of the stereo panorama (often hard left and hard right, though not always). In some instances, I'll utilize same-side processing to help “lock in” each guitar's position. In other instances, I'll set up opposite-side processing to help glue them together a bit while also giving some more stereo information.
Beyond single-channel panning, with modern DAWs having high numbers of return channels available, we can opt to set up two sends for any given track(s), pan each send to its own side (hard left and hard right) and alter the processing slightly to achieve a widening of the otherwise mono effects.
EQing Effects Returns
EQing our effects returns is something else I touched on earlier that deserves its own section.
EQ is one of our most important processes, helping to shape our sounds and make everything fit together within the mix.
Any time we set up an effects return and mix the resulting audio, we're effectively adding frequency content to the mix. This may be obvious, but it's worth mentioning, as any new content we add has the potential to mask other frequencies in the mix and cause phase issues with other tracks.
While every situation is different, it's common for frequency masking to become excessive in the “muddiness range” between about 100 and 500 Hz (this range is where most of the fundamental frequencies and the first few harmonics of our instruments and vocals reside).
These low-mids are also where a lot of power resides in the mix, and so I'll often do my best to maintain the energy of this range in my dry tracks and actually cut/reduce it in my time-based effects. Delay and reverb can easily washout this range, so a bit of subtractive EQ can work wonders.
Another important range worth considering is the low-end, which is sensitive to phase cancellation, especially in the stereo field. Therefore, it's often advised to only allow the elements with musical information in their low-end to occupy this range while also keeping things in mono to reduce stereo phase cancellation and ensuring the best mono compatibility.
High-pass filtering our effects returns with EQ can help to keep the low-end clean and powerful within the mix.
As an additional point, I would suggest utilizing linear phase EQ any time to have parallel-style processing on your return channel. It helps maintain the phase relationship between the “dry” and “wet” signals, even with rather drastic EQ moves.
However, in nearly all other cases (notably for our time-based effects), I would recommend minimum phase or “normal” EQ since it sounds more natural than linear phase EQ.
These are simply my suggestions and what generally works for me — always do what sounds best to you!
Dynamic Time-Based Effects Returns
Dynamic processing, and particularly compression, is another important style of processing in music production.
We've already discussed parallel compression, which is a form of dynamics processing.
However, in this section, I want to cover a cool effect that incorporates sidechain compression.
As a quick primer, all compressors have a sidechain. It's the signal that the compressor reacts to in order to apply gain reduction. In most cases, the sidechain signal is the same as the input signal.
As the sidechain signal level surpasses a set threshold, the compressor's gain reduction is triggered according to its ratio, attack and release controls. Here are the quick definitions of these controls:
- Threshold: a set amplitude limit that dictates when the compressor will engage and disengage. As the sidechain signal exceeds the threshold, the compressor kicks in (with its given attack time). As the sidechain drops back down below the threshold, the compressor disengages (according to its release time).
- Ratio: the ratio of the number of decibels the sidechain signal is above the threshold to the number of decibels the output signal is above the threshold. In other words, it is the relative amount of attenuation the compressor will apply to the signal.
- Attack time: the amount of time it takes for a compressor to engage/react once the sidechain signal amplitude surpasses the threshold.
- Release time: the amount of time it takes for the compressor to disengage (to stop attenuating the signal) once the sidechain signal drops below the threshold.
The big difference is that in “sidechain compression”, the sidechain signal is taken as something other than the input signal.
And so we can insert a compressor on an effects return channel, for example, and have it reduce the gain of that signal according to the signal level of a different track.
This brings us to what I call “dynamic time-based effects return”, where we can send a track (a vocal, for example) to a delay or reverb return and then compress that return with the original “dry” vocal as the sidechain.
I'm using a vocal track as an easy example because vocals tend to have good phrasing with space between the vocal lines.
So, with our return channel, we can dial in our time-based effect to our liking, generally as a way to give some dimensional characteristics to the vocal.
However, we may find that mixing it at an appropriate level is difficult — we want it to give a sense of dimensionality without necessarily drawing attention to itself or, more importantly, taking away from the vocal in terms of frequency making and washout.
Sidechain compression to the rescue! With sidechain compression, we can automatically reduce the level of the delay or reverb when the vocal is present and then have it swell up slightly during the breaks in the vocal phrasing. This way, the time-based effect is always present without getting in the way of the dry vocal and it can still do its job in yielding a sense of space as it fills the space between the vocal lines.
Of course, you can swap the vocal with any other track, especially if there is good phrasing in the track.
I have a detailed article on compressors that you can check out here: The Complete Guide To Audio Compression & Compressors
Noise gates are another dynamic process. They work by “opening the gate” (allowing signal to pass through) as an input signal surpasses a set threshold and “closing the gate” (not allowing signal to pass through) as the input signal drops below the set threshold. This is done according to time controls, notably the attack time, hold time and release times:
- Threshold: the sidechain level at which the gate will open and let sound pass through.
- Attack time: how quickly the gate opens.
- Hold time: how long the gate stays open once it's opened (even as the sidechain drops below the threshold).
- Release time: how fast the gate closes.
You may have noticed the word “sidechain” in the points made above. Indeed, noise gates can be triggered using a sidechain signal other than the input signal.
That brings us to the effect of “gated reverb” which is exactly what it reads like: an effects return with a reverb inserted followed by a gate. However, instead of relying on the reverb level to trigger the gate, we'll use the dry signal feeding the return as the external sidechain signal.
This is most commonly a snare track — you can hear lots of gated snare reverb in music from the 1980s.
Obviously, we can use tracks other than snares, but they make the best examples because they're so transient (they have short peaks in their audio tracks followed by lower-level signal that is unlikely to trigger the gate). Additionally, we don't necessarily have to gate reverb; we can use other processes on our effects returns, too.
But getting back to the example at hand, we'll send the snare track to the reverb return, insert a noise gate after the return and use the original snare track as the sidechain. From there, we can adjust the gate threshold and time settings to have it open at the right times.
When I opt to use gated reverb, it's generally on a relatively simple beat. For example, snares of the backbeats (2 and 4). I'll generally have the gate open as the snare hits and have it stay open (by adjusting the hold time) until the following downbeat (1 or 3).
By setting things up this way, we can get the sense of a massive-sounding reverb without the extended tail and washout.
Colourful Delays And Reverbs
When I say “colourful delays and reverbs”, I simply mean adding additional effects not typically offered within delay and reverb units.
This could be a certain distortion or modulation effects, or a combination of several effects — anything to give the delay or reverb return a different tonality from the signal(s) feeding it.
Note that the signal chain (the order of inserted effects) can have a big impact on the sound of the return channel.
For example, as a guitarist, I always have my pedalboard setup as follows:
- Utility pedals
- Dynamics, EQ and pitch-shifting pedals
- Distortion pedals
- Modulation pedals
- Time-based effects pedals
That's my typical setup because it works to give me the most optimal tone in the widest range of scenarios. However, that's not to say that re-ordering the effects wouldn't yield better results — just different results.
It's super easy to re-order plugins in your DAW's return channels, so have some fun experimenting with colouring your reverbs and delays. Try different distortions before or after the effects, have effects before and after the time-based effects — get creative with it!
Mid-Side Processing On Effects Returns
Mid-side processing, in general, refers to having independent control over the “mid” information (the information that's equal in the left and right channels) and the “side” information (the information that's different between the left and right channels).
There are a variety of mid-side processes, including mid-side EQ, that we can utilize to process the stereo image.
However, what I really want to bring the concept behind the mid-side miking technique into our return channel.
Let's quickly go over the mid-side miking technique to get a better idea of what I'm talking about. It involves one unidirectional/cardioid microphone pointed toward the sound source and one bidirectional/figure-8 microphone pointed to the sides.
In the mixer, we keep the cardioid mic panned to the centre. The bidirectional mic, however, is duplicated, with one copy panned to the left and the other phase-inverted and panned to the right.
That way, we have a more ambient capture of the intended source (remember that the bidirectional mic points to the sides) that sounds super-wide in stereo (maximum phase differences) and completely cancels out if the mix is summed to mono.
Mono compatibility is really important for professional mixes, and while this style of processing eliminates the effect when a mix is collapsed to mono, it also doesn't lead to any unintentional worsening of the mix either. So it's not always appropriate, but it's a useful tool to have.
Now, moving onto our effects returns, there are a few ways to set up these “side” channels:
- Send the mono signal to the effect return, process it with a mono delay or reverb, split it to stereo within the mic (keeping the left and right channels the same), and then flip the polarity of the left or the right channel.
- Send the mono signal to two different mono return channels, apply the same effects to both, pan one to the hard left, the other to the hard right, and flip the polarity on one of them.
This will result in a wide stereo image for the effect when the mix is in stereo, and complete cancellation when the mix is summed to mono, just the mid-side miking technique.
Widening Effects Returns With Doubling
Speaking of widening our effects returns, beyond utilizing stereo effects like stereo delays and reverbs, we can also utilize widening effects to gain extra width and dimension.
Doubler effects and stereo wideners can be put to good use on our effects returns. We can maintain much of the original image from the dry track(s) while getting some extra width from the return.
I really enjoy utilizing doubler effects, which basically combine stereo delay and detuning to push copies of a signal to the sides of the stereo image. I think it's a great way to add colour (distortion or modulation, primarily) to the sides of the mix.
When used in conjunction with a dry signal panned to the centre, this strategy can add a subtle (or not-so-subtle) width to the original signal.
If we're dealing with a stereo effect, we can consider utilizing stereo-widening effects as well. I typically stay away from such effects, especially on the mix bus or even the subgroups, but for effects, they can help push things further out without necessarily being detrimental to the overall phase relationships of the tracks within the mix.
Shimmer effects are among my favourites. Shimmer is generally produced by combining pitch-shifting (an octave up) with reverb, though multiple octaves can be used, as can delay effects.
The results can be described as ethereal, celestial and, of course, shimmering. The key to this effect is in the seamless melding of the two effects, typically with any pitch-shifting (which often causes artifacts) coming before the reverb or delay in the inserts (which can often mask such artifacts).
Shimmer effects can be used to add top-end information and texture to a sound and, when automated, to accent certain moments in the song. Because pitch-shifting effects generally work by shifting the harmonics, we can get more harmonically-rich results from simple shimmer effects.
I love using shimmer effects and sometimes even cascade different shimmer effects together, a routing technique we'll touch on later in this article.
Automating Effects Returns
Automation is key for a professional, dynamic mix. Automating our effects returns can really help bring them to life with dynamic movement throughout the mix.
First of all, we can automate the send levels from the track(s) sending to the effects return. This can be gradual or abrupt, the latter being in line with effects throws, which we'll get to next. We can also automate the muting and unmuting of a given send to properly achieve such throws.
So, with that information, we can alter the balance of a given track's sends to a variety of different return channels. We can even automate whether what's being sent will be pre- or post-fader if we really want to (though this may require automating the levels of one pre- and one post-fader send to a common return channel).
Getting to the return channel itself, we can automate the muting of an effects return channel to bring it in and out of the mix entirely. This is somewhat similar to effects throws.
Following the simplicity of the mute button, we can automate the fader levels of an effects return channel to make the desired effect more or less prominent throughout the song.
- Perhaps a reverb is overbearing in a certain section: automate its level down.
- Maybe a special effect is needed for a build-up: automate its level up throughout the build-up and then bring it down at the beginning of the next section.
- Let's say you only need the effect for a brief, periodic moment; consider automating the levels up and down rather than muting and unmuting.
We can do the same with panning, as well, if need be.
What's more is that we can actually automate the parameters of the inserted processes on the return channel. It's here that options become limitless.
Expanding into the topic of effects throws, we know that we can trigger them by unmuting and muting our effects sends to specific return channels.
The term “throw” means to suddenly send a brief part of an audio track to an effects return, which will “throw” the effect back (ultimately to the mix bus).
A throwback channel typically has some sort of time-based effect on it (delay or reverb) in order to give some dimensionality, or at least timing differences, between the original signal and return. It makes the throw more obvious compared to on-time parallel processing, which often acts to increase the level of the dry track with some alterations to the tone.
That's not to say that you can't throw to typical parallel processes but that it's much more common to throw to a delay or reverb.
Throws can be used to accentuate certain notes or phrases within the song, helping to create a more sonically engaging production.
I have a video going into greater detail on vocal throws, in particular, if you're interested:
Chaining Multiple Effects Returns Together
Here's where things start to really get fun and experimental.
Beyond sending our “regular” tracks to specific aux return channels, we can also send those return channels to other return channels.
The easiest way to conceptualize this is with cascading delays:
- Send an audio track to an effects return (Return Channel 1)
- Insert a delay plugin on Return Channel 1, set the wet/dry to 100% wet, and set the feedback to 0 so that there's only a single repeat.
- Send Return Channel 1 to a second effects return (Return Channel 2)
- Insert a delay plugin on Return Channel 2, set the wet/dry to 100% wet, and set the feedback to 0 so that there's only a single repeat.
- Repeat the process as you see fit.
In this simple example, we're cascading delays together. All else being equal in the delay effects, it would probably sound like a simple delay, with each successive repeat getting quieter.
However, we can opt to alter each return's processes to our liking. For example:
- We can alter the delay time on each delay to stagger the repeats.
- We can insert additional effects to colour each successive repeat — note that the effects from the previous returns will be carried to the successive returns.
- We can increase the feedback of the delay plugins and have multiple repeats of the first trigger even more repeats from the second, and so on.
- We can pan the returns differently and opt to send post or pre-pan to the following returns.
- We can automate the return channels and their respective effect parameters as we see fit.
As a related method, we can send a single track to multiple return channels, each with a different delay time (spread out by quarter notes, for example), for a cleaner repeat with true independence of sonic character.
And that's just an example with delays. Imagine what can be done with various different effects. Here are a few options worth considering:
- Reverb return into delay return
- Delay return into reverb return
- parallel saturation return into delay return
- delay return into shimmer reverb return
- parallel saturation + delay return into shimmer delay return
The options are endless!
Effects Return Feedback Loops
My big disclaimer here is to be careful of overloading when experimenting with feedback loops. Positive feedback loops, like those in live microphone setups, can quickly get out of hand and can potentially damage your equipment!
In the previous section, we discussed chaining a sequence of return channels together. In this section, I want to discuss feeding two or more return channels into each other to create and appropriately manage a feedback loop.
A feedback loop happens when an audio output is effectively routed back into its input.
A good example of this is the delay effect we've been discussing throughout this article: the feedback parameter controls how much of the delayed output is routed back into the input of the delay line and, therefore, controls the number of repeats.
We can achieve this within our effects returns as well by simply feeding one return channel into another and then feeding that return channel back into the first.
The cleanest way to do this is with delay returns since we can hear the repeats stack on top of each other. However, any effects can be used in this setup to create feedback loops.
Once again, I have to warn you to be careful when setting these up. They can quickly add up and overload the system.
However, when controlled, feedback loops in our effects returns can yield interesting self-oscillating results. We can then opt to automate the fader(s) of the involved return channels to create interesting results in our music production.
Be careful with one, but have fun experimenting with feedback loops. You can find some really interesting stuff that close to danger!
Call To Action
Take any and all of these ideas and make them your own. Get in the digital audio workstation and experiment for yourself.
Let me know in the comments what you come up with — I'm always interested to hear about how people expand their creativity with the technologies available to them!
New To Mixing? Check Out My FREE In-Depth Article:
Make the right decisions at the right times to craft professional mixes!
What is a subgroup? A subgroup is a mixer channel fed by a group of tracks/channels combined onto a single bus. Similar tracks are often grouped together (i.e., drum tracks bussed to the drum subgroup). A subgroup offers control over its grouped sources, often with inserts, sends, pan pot and fader, and output routing.
For more information, check out my article What Are Subgroups? (Audio Mixing, Recording, Live Sound)
What is a VCA group? A VCA group is an independent fader that doesn't pass audio but allows us to control the signal levels of multiple channels simultaneously without altering the channels' faders and without having to route the channels to their own subgroup.
Have any thoughts, questions or concerns? I invite you to add them to the comment section below! I'd love to hear your insights and inquiries and will do my best to add to the conversation. Thanks!
This article has been approved in accordance with the My New Microphone Editorial Policy.