Level a vocal with Lume-2A, then catch peaks with Strike-76

For SUS4 Lume-2A and SUS4 Strike-76 1.4.07 min readUpdated 9 November 2026

A vocal that moves from a whisper to a belt is hard to level with a single compressor. Set it hard enough to hold the quiet lines and it flattens the loud ones. Two gentle stages often work better: a slow one that rides the overall level, then a fast one that only catches the peaks.

Lume-2A is an optical levelling amplifier in the LA-2A tradition, and Strike-76 is a fast FET compressor in the 1176 tradition. This guide chains them in that order.

The short version

Lume-2A first: Optical Vocal. Insert Lume-2A on the vocal and load Optical Vocal. It is set to Comp mode with Emphasis/Xfmr on.
Raise Peak Reduction. Loop the loudest phrases and raise Peak Reduction until the gain reduction reads about 3 to 6 dB on them.
Watch the memory. On a held note the T4 CELL MEMORY readout rises and stays up. On a quick staccato line it lets go. That is the opto behaviour doing its job.
Strike-76 second: Vocal Tighten. Insert Strike-76 after it and load Vocal Tighten: 8:1, a fast attack, a 235 ms release and a detector high-pass at 77 Hz.
Set Input for peaks only. Turn Input until the gain reduction only shows on the peaks, a couple of dB at most. If it reads constantly, back off.
Listen to the esses. Compression brings sibilance up. Check the esses with both plugins on and de-ess before the chain if they have got harsh.

Why two stages

Lume-2A grabs in about 10 ms and lets go in two stages. Most of a short hit’s reduction returns in well under a second, but after a long, loud passage the cell stays “lit” and recovers over several seconds. That programme-dependent release is why an opto sounds smooth on a sung line: it follows the phrase rather than the individual syllables.

It is not a peak catcher. Strike-76 is. Its attack runs from 800 µs down to 20 µs and its release from 50 ms to 1.1 s, so it clamps the occasional peak that the slow stage lets through. Putting the leveller first means the fast compressor only has the peaks left to deal with.

Stage 1: Lume-2A, Optical Vocal

Optical Vocal is the reference vocal patch in Lume-2A’s 28 factory presets: gentle levelling at full wet, with a little transformer and valve colour. These are its values, converted to what the plugin displays.

Optical Vocal, as Lume-2A 1.4.0 sets it
Control	Setting	Why
Peak Reduction	−22.5 dB threshold	The dial reads 0 to 100: 41 here. Range −6 to −46 dBFS
Gain	+5.3 dB	Range −24 to +24 dB
COMP / LIMIT	COMP	3:1 with a 12 dB knee. LIMIT is 8:1 with a 6 dB knee
EMPH·XF	On	Detector emphasis and output transformer
TUBE	35%	Valve second-harmonic warmth
MIX	100%
EXT KEY / SC HP	INT / 53 Hz stored	With EMPH·XF on, the detector high-pass is about 800 Hz while SC HP is below 100 Hz
AUTO GAIN	Off

Setting Peak Reduction

Peak Reduction is the main control for how much happens. Loop the loudest part of the song and raise it until the dB GR readout moves by around 3 to 6 dB on the loud phrases, which is the manual’s own starting advice. Then play a quiet verse: the reduction should be small. If it is not, the vocal is being squashed.

Then set Gain so the level feels right, not louder. If you want to judge the sound rather than the level, switch AUTO GAIN on, which matches the output to the input over about 2.5 seconds, and switch it off again once you have decided. The EMPH·XF switch changes the detector so that presence and sibilance lead the ride rather than low-end thump. Try it on and off: on a bright singer, off may leave more air.

Stage 2: Strike-76, Vocal Tighten

Vocal Tighten sits in the Vocal group of Strike-76’s 28 factory presets. Input is not a threshold you type in: it is how hard you drive a fixed threshold, so turning it up increases the gain reduction.

Vocal Tighten, as Strike-76 1.4.0 sets it
Control	Setting	Why
Input	−14.4 dBFS threshold	Range −6 to −30 dBFS
Ratio	8	Buttons: 4, 8, 12, 20
Attack	81 µs	Range 800 to 20 µs
Release	235 ms	Range 50 ms to 1.1 s
Output	0 dB
Analog (XFMR)	Off
Mix	100%
Stereo	LINK
SC HP / EXT KEY	77 Hz / INT	High-pass on the detector only, so plosives do not trigger it
Auto Gain	On	With Auto Gain on, 0 dB Output means as loud as bypass

Setting it to catch peaks

Start with the Strike-76 bypassed. Play the whole song and find the handful of syllables that jump out of the levelled vocal. Switch the plugin on and turn Input until the gain reduction only shows on those, rather than on every word. The needle should spend most of its time at rest. A couple of dB on the peaks is a good target; this is advice, not a factory value.

Because Vocal Tighten has Auto Gain on, the level should stay at roughly what you had, so you hear the control, not a boost. The release stretches by up to about twice after dense compression, like the hardware, so a long chorus recovers more gradually than a single peak. If it sounds as though it is chewing the ends of words, lengthen the release, or lower Input.

Troubleshooting

Harsh esses after compression: both compressors bring sibilance up as they pull the level down. Put a de-esser before the chain. ChannelStrip’s De-Ess Only preset or DynamicEQ’s Vocal De-Ess will do it, and the vocal chain guide covers de-essing in detail.

Vocal flat or lifeless: you are probably asking too much of one stage. Back off Peak Reduction first, because the leveller is doing most of the work. A flat vocal usually means the slow stage is working constantly.

Breaths pumping up between phrases: the slow release brings them forward. Lower Peak Reduction a little, or ride the breaths down in your DAW before the chain.

Plosives still punching through: raise Strike-76’s SC HP, which is a step button that goes through off, 60, 120 and 250 Hz.