Finish a Dark-Pop Single in 6 Hours: My 5-Plugin, 3-Pass Finishing Rack for Logic Pro

Horia Stan is a music producer and sound engineer at The One Records, Bucharest.

My promise

I finish full dark-pop singles in under 6 hours. Not because I rush. Because I force decisions early. I cut options. I use five plugins that do heavy lifting. They save time and keep mixes coherent. If you want long, vague chains, stop reading.

The principle

Fewer plugins means fewer decisions. Fewer decisions means faster progress. I reject the idea that more plugins equal better results. A tight, repeatable chain keeps tonal intent and loudness predictable. I design that chain to survive stems, codecs, and mastering checks.

The five plugins and why each earns a place

1. FabFilter Pro-Q3 - surgical, transparent shaping. I use minimal bands. Typical moves: HPF on buses - 35-45 Hz (slope 24 dB/oct) depending on genre. Cut 200-400 Hz with a wide bell -3 to -6 dB to clear mud on synth beds. Use dynamic mode on 2-4 bands for problem resonances. I avoid linear-phase on bus inserts to keep transient life.

2. Oeksound Soothe2 - selective resonance control. I place this early on dense buses: main synth bus, pad bus, and vocal bus. Settings: 0.5-1.5 strength depending on source, 200-2k focus for synths, 1.0-1.5 depth for vocals. Soothe reduces masking so I can be decisive with Pro-Q3 instead of chasing dozens of peaks.

3. Soundtoys Decapitator - tone and harmonic glue. I use it subtly on the glue bus and sometimes on a parallel drum bus. Choose A or E model for darker color. Drive 2-4, mix 20-40% when I want grit. It replaces time I would spend automating distortion with multiple plugins.

4. Waves SSL G-Master Buss Compressor (or Waves Abbey Road TG if you prefer slower glue) - timing and bounce. I set attack 30-50 ms, release auto, ratio 2:1, threshold for 1-3 dB gain reduction on peaks. This gives pocket and keeps the transient picture intact for mastering.

5. Valhalla VintageVerb (plate or chamber) - ambience that reads on small speakers. Use a short plate on the vocal send: 0.9-1.4 s decay, pre-delay 12-25 ms, mix 10-18%. Use the same reverb algorithm on a very subtle stereo return for the whole mix - mix 4-7%. One reverb family keeps space coherent and saves countless A/Bs with different reverbs.

I know these choices are opinionated. They work for dark-pop. You can swap colours, but keep the roles: shape, tame, tone, glue, space.

My 3-pass finishing process

I rarely go back after pass 3. The song either works or it needs a rewrite. Mixing should not be the rewrite tool.

How I wire this in Logic Pro

I use a single master buss with the five-plugin rack in insert slots. Then I use five buses under that master: drums bus, bass bus, synth pad bus, vocal bus, FX bus. Each bus gets the same Pro-Q3 instance for local cleanups. I keep returns limited to three: reverb, delay, saturation. Track Stacks save states fast.

I use Track Alternatives to A/B structural versions. I use Logic Project Alternatives to store whole arrangement variations. I avoid loading more than two major instance-heavy instruments at once. When I need Keyscape, I freeze the instance after committing MIDI.

For export I choose 32-bit float WAV, 48 kHz. I export individual stems soloed, and the full mix with the five-plugin rack active. I also export a stereo dry bounce with only corrective EQ and no glue compression for mastering to reference.

Tuning numbers that matter in 2026

• Target loudness for single pre-master checks: -14 LUFS integrated. This lets me A/B against streaming targets without killing dynamics.
• True peak headroom: -1 dB TP on exported mix. That avoids inter-sample overs on converters and streaming encoders.
• Typical glue compression on master: 1-3 dB gain reduction. More means arrangement or mix problem.
• Reverb sends: lead vocal 10-18% wet. Global reverb 4-7% wet.

Decision saving techniques I use every session

• Preset folders for the five-plugin rack. I keep three presets per song: Fat, Neutral, Thin. I try each and pick one within 90 seconds.
• Commit noisy instruments early by rendering and replacing. That trims CPU and forces arrangement decisions.
• Use Soothe2 on busses, not every track. One instance across grouped elements tames masking faster than 20 instances on every track.
• Use Decapitator on a parallel bus. Blend until it supports the vocal, then reduce other saturation plugins.

Common pushbacks

"You need more plugins to get unique tones." No. You need fewer, used decisively. "But different songs need different chains." True. But the roles stay the same. Shape, tame, tone, glue, space. Fill those roles fast and you finish.

"What about mastering?" I do a quick chain for previewing only. I print a clean version for mastering with corrective EQ only. Mastering gets the -14 LUFS stem and the full mix at -1 dB TP.

Example timeline - 5.5 hours

• 0:00-1:00 - Pass 1: arrangement and rough balance
• 1:00-3:00 - Pass 2: tonal sculpting and automation
• 3:00-4:30 - Pass 3: quick master checks and prints
• 4:30-5:30 - Metadata, exports, and quality checks

If something takes longer, I stop and decide: rewrite or outsource.

Final words

This method is not about shortcuts. It is about focus. The five-plugin rack removes choice paralysis. The three-pass structure forces decisions early. The numbers -14 LUFS and -1 dB true peak keep my mixes ready for mastering and streaming. Build the rack. Commit by pass 2. Export stems at -14 LUFS, -1 dB TP. You will finish more singles and keep quality high.