Production guides · Meme Plugins · 27 October 2026 · Updated 2 November 2026
An insert processes audio along a track’s signal path. A send feeds audio to a separate path, often a return or auxiliary channel, whose output is blended with the mix. Choose the routing by the job: one source’s individual treatment, a shared effect, or a separate parallel texture.
The same plugin can serve either role. “Reverb belongs on a send” is a useful common setup, not a law about every reverb sound.
Follow the audio before choosing the control
- Insert: source → effect → later processing and output. The plugin’s own dry/wet control may combine unchanged and processed audio inside that path.
- Send: the original source continues to its output while a feed goes to another channel → effect → return to the mix.
Ableton’s return-track manual explains how one return can receive audio from numerous tracks, while an effect on a clip track processes that track. Check the routing in your own DAW, including group tracks and output assignments.
1. Use an insert when the treatment belongs to this path
An EQ correcting one recorded part or a compressor controlling that part commonly works as an insert. Each later processor receives the result of the earlier one. If a reverb is an essential sound-design layer for only one synth, its insert blend may also be the clearest arrangement.
Do not assume every insert must be fully wet. A dry/wet control can deliberately preserve part of the source, depending on the processor and task.
2. Set a shared reverb return deliberately
Create an effect return, place the reverb on it, and feed the selected tracks through their sends. For a conventional reverb-only return, set the processor fully wet so the original dry signals remain on their own routes. FabFilter documents a 100% Mix setting for Pro-R used as a send effect.
Start the return quietly, raise one source’s send until the space is useful, then add the next source. Keep the return’s output going to the intended mix destination, without a route back into itself unless you deliberately want feedback.
Example: two vocal parts, one room
Imagine a lead and a backing-vocal pair that should feel like they share a small room. This is an authored routing example, not an auditioned session.
- Keep both parts on their normal dry outputs. Feed one shared fully wet reverb return from their sends.
- Give the lead a smaller send than the backing pair if it should feel more forward. The sources can feed different amounts while sharing the same reverb settings.
- Mute the return. The dry arrangement should remain; restoring it should add the intended space.
- Compare a version with separate reverb inserts if the parts need different spaces or timing. Keep that version only for a useful musical difference.
A shared return makes one decay change affect both feeds. Separate inserts let their settings diverge. Neither design automatically makes the mix better.
3. Know which level you are changing
A send level changes one source’s contribution to the effect. A return fader changes the combined effect channel. The source track’s fader may also change its effect feed, depending on whether the send is taken before or after that fader.
Check the documented pre/post setting before automating a dramatic fade. This guide chooses the route; a detailed pre-fader versus post-fader comparison is a separate decision.
4. Diagnose accidental extra dry signal
If adding a return mainly makes the source louder or changes its tone unexpectedly, check whether the effect is also passing dry audio. Two routes carrying the source can change the balance, and timing differences can complicate their combination.
Mute one path at a time and verify the actual plugin blend. Use matched-level comparisons rather than judging a routing change by its volume jump.
5. Check what reaches the export
Include the intended returns and tails in the final render. If you are delivering grouped files, document where shared effects belong using the stems-versus-multitracks guide.
For effect choice, read reverb versus delay. For the tail itself, use pre-delay and decay. Routing and effect settings solve related but different questions.
For a dynamics example using two audible paths, follow the parallel-compression workflow. Keep the original and compressed paths clear, avoid an unintended extra dry copy, and check the timing and final sum.
If moving the source fader changes the effect relationship unexpectedly, check pre-fader versus post-fader sends. Compare the new effect feed separately from the tail already in the return and from the return fader itself.
To compare effect order with separate branches, follow serial versus parallel processing. Trace which signal reaches each processor and check the dry copies and final sum.