Master

Goal

Refine the tonal balance, dynamics, stereo image and loudness of the premix to make it competitive, cohesive and translatable across all systems and platforms, without losing what makes the mix musical. Processor details are in the guides "Mastering: the complete chain", "Mastering: EQ and dynamics", "Loudness by platform", "Limiter and true peak".

Step-by-step procedure

1. Prepare the mastering project (FL Studio)

• New project, same sample rate as the premix. Import the premix into the Playlist as an audio clip; route it to a MASTER CHAIN mixer track (or to the Master).
• Import 2–3 references on a separate track that outputs without passing through the chain (or bypass the chain when you listen to them).
• Place the meters after the chain: Emphasis (LUFS, true peak) or Youlean, Wave Candy (spectrum, vectorscope), Spreader (phase).

2. Listening without processors

• Listen to the whole premix and take notes per section: tone (dark/bright, mud, harshness), low end (bloated, weak, uneven), vocal (too far forward/back), stereo, dynamics.
• Measure: integrated loudness, maximum short-term, peak, PLR, correlation.
• If you find problems with a single instrument (sibilant vocal, 808 that booms on one note, snare too loud), go back to the mix or ask for the stems.

3. Align the references

• Bring the references to the premix's loudness (by lowering them) to compare the tone; then, at the end, compare the master with the references at the same loudness.

4. Corrective EQ

• HP 20–25 Hz if there's rumble; broad cuts of 0.5–1.5 dB on mud (200–400 Hz) or harshness (2–5 kHz); a narrow cut of 1–2 dB on an obvious resonance.
• Compare at equal loudness after every change.

5. Corrective dynamics

• Dynamic EQ or multiband only where needed (low end that grows in the choruses, harshness in loud passages, residual sibilance): 0.5–2 dB.

6. Glue

• Compressor 1.5–2:1, attack 10–30 ms, auto or tempo-synced release, HP in the sidechain 60–120 Hz, GR 0.5–2 dB. Optional: if the mix is already cohesive, skip it.

7. Color

• Color EQ: air +0.5/+1.5 dB at 10–16 kHz, weight +0.5/+1 dB at 30–60 Hz.
• Light saturation (tape/tube/console), barely perceptible.

8. Stereo

• Mono bass below 100–150 Hz if it isn't already; Side +0.5/+1 dB above 8 kHz only if the mix is narrow; correlation unchanged or better.

9. Loudness

• Clipper (optional, loud genres): 1–2 dB on the transients.
• True peak limiter: raise the input until you reach the desired loudness for the genre; GR 1–4 dB (peaks up to 6 in the loudest genres); ceiling −1 dBTP (−2 dBTP above −14 LUFS). In FL Emphasis: Versatile mode (or Loud for extreme genres), Hardness 2–4, True Peak on, 4× oversampling, Output −1 dB; use Level Matching and Delta for comparisons.

10. Final comparison at equal loudness

• Bypass the entire chain with Level Matching (or by lowering the output by as much as you gained): the master must sound better, not just louder.
• Comparison with the references normalized to the same loudness: tone, low end, vocal, width, punch.
• Listen to the beginning, the densest sections, transitions, the ending: no distortion, pumping or changes of character between quiet and loud sections.

11. Listening on multiple systems

• Monitors, headphones, phone (mono), car, Bluetooth, at low and high volume. Fix only what repeats.

12. Album/EP

• Consistency between tracks: perceived loudness (not necessarily identical integrated loudness), tone, width. Track order, gaps, fades.

Target values

Loudness as a choice

Spotify normalizes to −14 LUFS (with −11 and −19 options), Apple Music to about −16 LUFS: a master at −6 LUFS is turned down by about 8 dB on Spotify and can sound smaller than a master at −10 LUFS with lively transients. Push only as much as the genre requires for its character.

Link with mix and pre-production

• If the master requires EQ beyond 2–3 dB or heavy limiting: go back to the mix.
• Uneven low end between notes: sound design of the 808 or the bass.
• Chorus without impact: it's often the arrangement (little contrast with the verse), not the master.

Common mistakes

• Comparisons at different volumes.
• Limiter working beyond 6 dB.
• Ceiling at 0 dBFS without true peak.
• Master judged only on headphones or only on monitors.
• Dither active in the chain and then also at export.