When to use send channels instead of inserts

An insert replaces the signal. A send copies it. That one difference decides which you should use, and it decides more about how a mix sounds than most plugin choices.

The mechanical difference

Insert: the signal enters the plugin and the plugin’s output continues down the channel. The processed version is the signal now. Mix control, if any, happens inside the plugin.

Send: a copy of the signal is split off at a level you set and routed to a separate return channel. The original keeps going untouched. You now have two independent paths: the dry source on its channel, and the processed copy on the return.

That independence is the whole reason sends exist.

When a send is the right choice

Shared spaces

Put a reverb on a send and feed the vocal, the snare, the guitar and the keys into it. They now occupy the same room.

That is the single biggest reason mixes sound coherent. Four separate reverb instances create four different rooms and the listener hears a collage. One shared reverb creates one place.

It also saves considerable CPU: one reverb instance instead of four.

Independent control of the effect

On a send, the dry signal and the effect are separate faders. You can:

• automate the reverb without touching the source;
• duck the reverb under the vocal while the dry vocal stays put;
• EQ the reverb return without EQing the source;
• compress the return for a denser tail while the source stays clean.

None of that is available when the effect is buried inside an insert.

Parallel processing

Send a copy to a heavily compressed or saturated return and blend it under the original. You keep the original’s transients and add whatever the parallel path provides. This is how parallel compression on a drum bus works.

When an insert is the right choice

• Corrective processing. EQ, compression, gating and de-essing act on the whole signal by definition. A send makes no sense.
• Effects that define the sound. If a delay is the instrument, put it on the channel.
• Anything where you never want the dry signal. A fully wet effect that replaces the source belongs as an insert.
• Per-source settings. Four sources needing four different reverb sizes need four instances, and at that point inserts are simpler.

Pre-fader vs post-fader

Post-fader is the default and the one you want almost always. The send level scales with the channel fader, so when you pull the vocal down, its reverb comes down with it. The wet/dry relationship stays constant.

Pre-fader ignores the channel fader. Pull the source to silence and the effect keeps playing at full level.

Pre-fader is useful for:

• headphone monitor mixes during tracking;
• deliberate effects where the source disappears and its tail continues;
• a reverb that must stay constant while the dry source is automated.

If your reverb gets louder relative to the source every time you lower a fader, you are on a pre-fader send by accident.

Setting up a send correctly

1. Create a return channel and put the effect on it as an insert.
2. Set the effect to 100% wet. This is the step people miss. The dry signal already exists on its own channel — passing it through the return as well duplicates it, which changes the level and can cause phase problems.
3. Route the return to the master, or to a bus if you want it processed with a group.
4. Add a send from each source channel, post-fader, starting at silence.
5. Raise each send until you hear the effect, then back off slightly.

Processing the return itself

The return is a channel, so treat it like one.

High-pass the reverb. Low frequencies in a reverb tail are the fastest way to make a mix muddy. Removing them below a few hundred hertz usually costs nothing and clears a lot. Detail in how to fix muddy reverb.

Low-pass it too. Real spaces absorb high frequencies. Rolling the top off a tail makes it sit behind the source rather than on top of it.

Duck the return from the source. A reverb that steps aside while the vocal is singing and blooms in the gaps is a classic technique — ducking to create space covers how.

Compress it if you want the tail to sustain rather than decay naturally.

How many sends?

Three covers most records:

Return	Purpose
Short reverb	Depth and size without obvious tails
Long reverb	Atmosphere, usually only on a few sources
Delay	Rhythmic space, often tempo-synced

Add a fourth when something genuinely needs a different space. Not because the mixer looks more professional with eight returns.

Sends, buses and structure

A send is not a bus. A bus carries the whole signal for grouping; a send carries a copy for shared effects. Both feed the master, and both are part of the same routing picture — see what is a bus in mixing.

For the reverb itself, SPC Dimensions is built for send duty with ducking and spectral shaping on the tail, which is exactly the processing a return channel benefits from.