Inserts vs Effects Sends – Which to use for what
The topic of inserts and effects sends came up in a recent post on the message board here, so I thought I’d write a permanent article to explain the difference and how they are typically used in recording or mixing applications.
Many young engineers, and home recording enthusiasts, have never worked with an analog console before, and the only mixer they have ever used is that built into their DAW software of choice. They may have had no formal training, and thus may be unfamiliar with the way channel inserts and effects sends, or aux sends, are traditionally used, especially in the big studios with big analog consoles. While there is no right or wrong way to do anything, and modern DAW software and computers have enough power to do almost anything you want, there are reasons why the more traditional methods of using inserts and sends were established.
On a traditional consol, an Insert is a set of connections in a channel path that would allow you to “insert” an external piece of audio processing hardware into the signal path. Of course, you don’t actually insert the physical piece of hardware! It’s simply a set of audio connections, usually wired to a patchbay in big studios, that allow you to send the signal out to the external hardware, and then bring the processed audio back into the same signal. If nothing was plugged into the insert, the audio proceeds through the channel normally. As soon as you plugged something into the insert, the normal signal path was broken, and the audio was routed out to whatever you patched into the insert, and then returned back to the channel signal path. Usually there is a pair of connections, an “insert send” for sending the audio out to the external gear’s input, and then an “insert return” for patching in the output of the external gear back into the channel. On some lower cost home consoles, the insert jacks use one TRS style connector, which requires a special “insert cable” that breaks that connection out to separate send and return connectors on the other end, for patching to the inputs and outputs of your external gear.
In terms of signal flow, the channel insert connections usually come right after the microphone pre-amps for that channel. So, any external gear you insert will come before the channel EQ, Fader, and Pan. On some consoles, there are switches that allow you to change the EQ, or at least the hi and low pass filters, to come before the channel inserts in the signal chain.
If you wanted to insert more than one piece of external hardware into the signal path, you simply daisy chained them, connection the output of the first piece to the input of the next piece, and finally taking the output of the last piece of hardware back into the channel insert return.
Now with modern software based DAW systems, we have plenty of power and flexibility to work with. Most DAW systems have multiple inserts for each channel that allow you to add software based “plug-in” processors. Some of the newer DAW systems now even allow you to insert external hardware into the software chain using special insert plug-ins that you configure to route audio to and from your choice of audio connecctions on your audio interface, and then the software will allow you to do a “ping” of the channel path to compute, and automatically correct for, the latency of the connections (the round trip time it takes for the audio to go out through your audio interface, through the hardware, and then back into the audio interface… since all audio interfaces have a certain amount of processing delay time that the introduce).
Some DAW programs have a fixed number of insert slots on each channel, while other software allows you to add plug-in/insert slots as needed for the channels. Programs like Nuendo/Cubase, for example, have a fixed number of insert slots, with most of them being configured as pre-fader, and the last couple of slots set up as post fader. Some software will allow you to configure whether the inserts are pre or post fader, while others do not give you the option.
That’s all well and good, BUT, you ask, What Do I Use An Insert For?
In the traditional analog world, inserts were used for hardware processors where you only wanted to hear the resultant processed audio, with none of the original, un-processed, audio passing through. These are typically dynamic processors and equalisers or other enhancers. Compressors, Expanders, Noise Gates, Transient Shapers, outboard parametric or graphic Equalisers, Sonic Maximisers, and Aural Exciters are common examples. These types of processors don’t have any dry/wet blending features because you typically only want to hear the fully processed output, with none of the unprocessed original audio, when you are using these types of processors to modify your audio.
Of course, there are no rules, and sometimes you may want use some of these types of processing in a more extreme way as an effect, and blend the processed audio back in with the original. I’ll discuss that a bit later after I discuss effect sends.
EFFECTS / AUXILIARY SENDS:
Effects and Auxiliary sends can be used for a variety of purposes, including creating monitor mixes for the musicians headphones during recording, or on-stage monitors in a live performance. However, for this article I’m going to concentrate on how effects sends are used for audio processing while recording or mixing.
Again, going back to our traditional analog recording setups, most big analog studios had a fixed number of processing hardware to work with, especially things like reverb, delay, and other “time based” effects processors (such as chorus and flange). In fact, in the really old days before digital hardware reverb and delay processors, the only reverb you had available was what you could get out of the room while recording, or through special acoustic reverb chambers that were setup elsewhere in the studio facility (these were basically a small, but long, chamber with a speaker at one end and a microphone at the other, and you changed the reverb time by moving the microphone closer to the speaker, as well as sometimes a movable wall behind the microphone). Some studios also had analog hardware reverb units, such as spring or plate reverbs.
So, with a limited number of reverbs, delays, and other processors, what happens if you want to use reverb on more channels that you have reverb units? If you only had one reverb processor (analog or digital), you couldn’t use channel inserts to put it on more than one channel, since you could only be connected to one insert at a time. This is where the effects send and return structure came in handy. Instead of using inserts, we use an effects send bus to allow us to route a controllable amount of signal from multiple channels to the same effects processor. You take the output of the effects send bus and patch it to the input of your effects processor, and then you take the output of that processor and patch it into an open set of channels on your mixing console, or patch it to dedicated effects return channels on some mixers. Then, using the corresponding effects send knob on each channel, you can control how much of each channel you want to “send” to the processor. This send control is in parallel to the channel, and doesn’t change the level of the audio through the channel itself. So, if you wanted a lot of reverb on a vocal, for example, but just a little bit of reverb on the snare drum, you would turn up the effects send (that corresponds with the patched in reverb) on the vocal channel quite a bit, while turning up the effects send on the snare channel just a little bit. The output of the reverb comes back into the mixer on its own channels, or special effects return channels, and you can use those channel faders/knobs to control the overall amount of the returned reverb you hear in the mix. The original channel faders for the vocal and snare, in our example, still control the overall “dry” level of the snare and vocal. So, by setting the sends, return channels, and “dry” channels, you can control the blend of the dry and processed audio in the mix. In most applications, the effects sends are set up to be POST fader, so the signal level to your effects is split off AFTER the fader. The reason for this is that you want your ratio of dry to wet signal to stay the same as you ride the faders for each channel during the mix. For example, once you have the blend of dry and reverb levels set up the way you want for a vocal track, you don’t want the reverb level to stay constant if you pull down the level of the vocal in the mix… what you want (most of the time) is for the level of the reverb to come down the same amount as the vocal level comes down when you pull down the vocal channel fader. By setting up the reverb send POST fader, when you pull down the vocal channel fader, the amount of vocal sent to the reverb will be reduced by the same amount, thus preserving your dry/wet blend that you worked so hard to set up!
Almost all mixing consoles have more than one effects send bus so that you can use more than one type of reverb, delay, or other effect during your mix, and share these with multiple audio channels. For example, if you have multiple reverb processors (or chambers) available, you may set one up for a long reverb, one for a short reverb, and maybe even one somewhere in between. You may also want to set up one or more delay processors, maybe one long echo type delay, and one short slapback type delay. Each one of these effects would need to have its own separate effects send bus and return channels. So, you can see how you can easily need 5 or more effects sends during a mix.
Now back to the modern world of powerful computers and DAW software. With plug-in reverbs, delays, and other effects processors, you can have as many instances of these plug-ins running as your computer can handle before you max out the processing power. If you’ve got a powerful computer and are using effects plug-ins that take up relatively little processing power, there is nothing to stop you from putting a plug-in reverb or delay, or other time based effect, into the insert slots of every channel that you want to have those types of effects. If you did it that way, you would need to use the plug-in’s own wet/dry control to set the blend between the dry and processed signal for each effect.
Unless you are setting different effects for every audio channel, though, setting up reverbs, delays, and other time based effects, as inserts is usually a waste of processing power and a bit time consuming as well. In addition, the best sounding reverbs are usually very processor intensive, and you’ll quickly max out your computer’s processors if you try to insert a high quality reverb on every channel.
It’s still usually best to set up processor intensive time based effects, such as reverb, as an effects send and return type of effect. When you set it up this way, you would want the effect set to 100% wet, so that none of the dry signal passes through to the effects return, since you’ll be controlling the dry levels with each track’s channel faders, and you’ll control the overall level of the reverb (or other effect) in the mix with the return level, of the FX channel level. Different DAW programs configure send and return type of effects differently. For example, in Nuendo and Cubase you add an FX channel to your project, which looks like an ordinary channel with several insert slots where you put the effects. When you first create the FX channel in Nuendo or Cubase, it asks you to select the effect you want to use from a drop down list of your available plug-ins. Once you have the FX channel set up, it will appear in the drop down lists for each FX send control on the right side of the channel views in Cubase and Nuendo. With analog mixers, if you connect a reverb to effects send bus 1 on the mixer, that reverb would be associated with effects send 1 of every channel on the mixer. However, with software DAW programs, such as Nuendo and Cubase, you can choose which FX channels each one of your channel sends is associated with, and those are independent from the same numbered effects sends on other channels (although, you can certainly set them all the same way, if you are using the same effect on multiple channels, to keep things easy).
In the analog world, some effects sends on a mixing board are set up to be pre-fader, and others set up to be post fader, while sometimes several of the effects sends would have a switch to change them between pre and post fader. Pre-Fader sends were usually used for creating headphone mixes for the artist that wouldn’t change if the recording engineer changed fader levels… thus, giving the artist and the recording engineer independent control over what they want to hear while recording. In most modern DAW software, you can usually change each individual send to be either pre or post fader. For almost all effects processing situations, you would want to leave them at the default post-fader configuration so that the levels of the reverb for each channel change as you change the fader levels of the channels (as described in more detail earlier).
In most situations, you don’t need a totally unique reverb or delay, or even chorus/flange, type of effect for every channel. You may want a long/lush reverb, a shorter room type ambient reverb, and maybe one or two in-between reverbs, and similary a long echo type delay, and one or more shorter rhythmic or slap-back type delays. It makes more sense, and saves processing power and setup time, if you set these up on FX channels and use sends and returns to share those effects between all the channels that you want to use each one on, rather than inserting a new instance of each effect in an insert slot on each channel that you want it on.
GOING AGAINST THE “STANDARDS”:
Just because the above described methods are the established “standard” method of setting of different types of effects and processing, there are often times when you’ll want to set things up differently to achieve certain special effects, or just to control things in a different way. I’ll discuss just a couple of them in this section.
One popular method of mixing drums is to use “parallel compression”. Instead of inserting compressors on every drum channel that you want to compress, or even routing all the drum tracks to a stereo group channel or buss and compressing that, you can set up a compressor in an effects send and return type configuration. You treat it like you would a time based effect, using the sends on each channel to control how much of each channel gets sent to the compressor, and then using the FX or return channel faders to control how much of the output of the compressor you hear in the mix. The original, uncompressed, drum tracks are still in the mix and controlled by their own faders, but now you can blend in some of the compressed signal as well, and vary the blend of compressed and uncompressed signal throughout the song. This style of parallel compression is typically used when you are using extreme settings on the compressor to get a really heavily compressed, or even distorted, sound, and you blend in just enough of it to give your drum sound a little more “balls”. You can bring in a lot more “balls” during heavy sections of the song, and then back it way down in the quiet, or more sparse, sections.
It doesn’t have to be a compressor, though. You could do the same parallel thing with a distortion processor to add a controllable amount of distortion, or something like a transient modifier to bring out more attack. In fact, that’s one of my favorite mixing tricks now is to set up my transient modifier to really bring out a lot of extra attack. I put that on an FX channel in Nuendo, and only send the kick, snare, and some toms to it. In the quiet verse parts of songs, I usually have that FX channel set pretty low, or even totally off. When things get really heavy and dense, such as when lots of heavy distorted guitars come in, I’ll crank up the FX channel to add a lot of extra attack to the drums to help them cut through the wall of guitars.
It also doesn’t have to be drums that you use this for. You could do the same thing with vocals, or any other instrument.
Another example of going against the “standard” way of doing things is to set up a reverb on a PRE fader effects send. With it set up this way, the amount of signal from a channel sent to the reverb remains at a constant level, no matter what you do with the channel fader. In this way, you can completely pull down the channel fader, removing ALL of the dry signal, but the reverb signal will remain. This can make a cool fading out into the distance type of effect. You gradually fade down the regular channel fader, which slowly lowers the dry signal level while the reverb level stays the same, making it sound like that audio is getting farther and farther away as it slowly fades to reverb only. That’s about the only time I can think of using an effects send to a reverb in a pre-fader setup, but maybe someone else has found other creative uses for pre-fader sends on things like reverb.
Just remember, there is no right or wrong way to do things. I’m just explaining these established “standard” methods of doing things to help save you some time (and processing power) when you are mixing. But, in the end, it all comes down to using your ears and doing what sounds right to you and right for the song!
© copyright 2000-2010, DBAR Productions, LLC
This content may be downloaded for personal use only, and may not be reprinted in part or in whole in any form without the express written consent of Stephen Sherrard and DBAR Productions, LLC