When the delay buries the vocal

The problem is timing, not volume

A delay is a copy of the vocal, so it occupies exactly the same frequency range as the vocal. Turn it up enough to feel and it starts fighting the thing it came from.

Most people respond by turning it down until it stops being a problem, which also stops it being an effect. The better answer is to keep it loud but get it out of the way while the singer is actually singing, and let it bloom in the gaps.

That is what the Movement module is for.

Ducking

The DUCK knob compresses the delay whenever the source is present, so the repeats step back under the vocal and return between phrases.

Four controls shape it:

• DUCK sets how much gain reduction is applied. Watch the gain-reduction meter on the left rather than guessing.
• HARD or SOFT. Hard uses more aggressive compression, which suits dense material like rock and metal. Soft is gentler and more delicate, which suits sparse arrangements where hearing the delay move would be distracting.
• DUCK THRESHOLD decides what counts as "the source is present". Set it too low and quiet breaths duck the delay; too high and only the loudest notes do.
• RELEASE controls how fast the delay comes back. This is the one that decides whether ducking sounds musical or like a pumping effect.

Set RELEASE to SYNC. It locks the recovery to your project tempo, so the delay returns in time with the song rather than at some arbitrary rate. That single choice is what makes ducking feel like part of the arrangement instead of a compressor doing its job.

Focus and Soften

If the delay is still muddying or sharpening the mix after ducking, the next stop is the Automatic Clarity Enhancement pair:

• FOCUS removes mud from the delay signal.
• SOFTEN removes sibilant harshness.

Both have gain-reduction meters beside them, so you can see how much is happening. These act on the delay only, which is the point: you get a cleaner repeat without touching the vocal that fed it.

Reach for these before you reach for EQ. A static EQ cut has to be set for the worst moment in the song, so it is wrong for the rest of it. These move with the material.

EQ the delay, not the vocal

Ducking fixes when the delay is heard. EQ fixes what is heard, and it is usually the faster fix for a delay that feels like it is burying the vocal. Clarity Delay has a full parametric EQ built in for exactly this: high and low pass filters, high and low shelves, and three bell nodes.

Three cuts solve most of it:

• High pass around 200-400 Hz. The low mids in a vocal are what make repeats feel thick. Take them out of the delay and the same delay level suddenly sits underneath the performance instead of on top of it.
• Low pass around 6-8 kHz. The delay stops competing with consonants, and the repeats read as distance rather than as a second singer.
• A narrow bell cut wherever the vocal has its own resonance. Whatever pokes out in the vocal pokes out in every repeat.

Working the EQ: high and low pass filters turn on by double clicking a filter icon, or by clicking and dragging it. Right click a filter icon to set its slope - 6, 12, 18, or 24 dB/oct. Shelves read orange when active and gray when bypassed; the three bell nodes read purple when active. Hold Shift and drag a node up or down for Q, or type a frequency or Q straight into its popup box. Ctrl/Cmd + click a node to reset it.

PRE or POST

The buttons in the top left choose where the EQ sits.

PRE filters the signal going into the delay, which means it is inside the feedback loop. Use it to keep the repeats themselves clean and smooth - a high pass here stops mud compounding pass after pass, rather than just masking it at the output. This is the one to reach for first on a long, high-feedback delay.

POST filters the delay after the fact, and leaves the repeats alone. Use it to shape tone and vibe: a dark filtered wash, a thin lo-fi return, a tail that fades out of the top end. It is also the safer place for a boost.

Cut rather than boost on PRE. Anything you boost before the delay gets fed back around the loop and amplified on every repeat, which is the mechanism behind runaway feedback. If you want a boost, put it POST.

The SOLO button in the EQ section follows the same switch: with PRE selected it plays only the dry incoming signal, with POST selected it plays the signal with the delay applied. That makes it easy to hear exactly what each EQ is doing. It blinks while engaged - leave it on by accident and you will wonder where your mix went.

Width, not level

If the delay is crowding the vocal in the center, the fix may be placement rather than volume.

• Pan each of the three delay channels independently. A repeat that is not on top of the vocal does not compete with it.
• Wide I and Wide II add mono-compatible width, so the delay opens up without collapsing when someone listens on a phone speaker.
• Ping-Pong moves repeats between the sides, which clears the center entirely.
• The SPATIAL WIDTH slider is worth automating: widen into a chorus, narrow for tension.

For more on the width modes and what changed in 1.1, see the Clarity Delay 1.1 update guide.

If it is still too wet

Two quick checks before changing anything else:

1. Solo buttons. A headphone icon left on in the channel mixer, or SOLO in the EQ section, will leave you hearing almost nothing but delay. Check both PRE and POST EQ.
2. Global Dry/Wet. Below 50 percent for an insert. If you are using Clarity Delay on a send or aux, the opposite applies and you want it fully wet.