What it is
Serial compression means placing two or more compressors in a row, each with a different job and a moderate amount of reduction, instead of asking a single compressor to do 10–12 dB of reduction. Each stage works less, the artifacts (pumping, distortion, loss of attack) stay low and the final sound is controlled but natural.
It is the standard technique on pop, rap and R&B vocals, on bass and on many buses. The most famous combination is fast FET + slow opto (1176 → LA-2A): the first catches the peaks, the other levels the phrase.
Why it works
- A compressor doing 12 dB of reduction has to choose between a fast attack (controls the peaks but flattens) and a slow one (lets the peaks through). Two stages of 3–6 dB let you have both.
- Each type of compressor has its own character: FET is aggressive and bright, opto is soft and program-dependent, VCA is precise, vari-mu is dense. In series the characters add up.
- Total reduction is the sum: 4 dB + 4 dB = about 8 dB, with less perceived distortion than 8 dB done by a single unit.
Typical orders
| Source | Stage 1 | Stage 2 | Stage 3 (optional) |
|---|---|---|---|
| Pop/R&B vocal | FET (1176): ratio 4:1, attack 3–10 ms, release 50–100 ms, 3–5 dB on the peaks | Opto (LA-2A): 2–4 dB constant | Limiter or light clipper: 1–2 dB on the residual peaks |
| Rap vocal | FET 8:1 or 12:1, attack 1–3 ms, 5–8 dB | VCA 3:1, 2–3 dB, auto release | — |
| Rock vocal | FET “all buttons” in parallel + opto in series | Opto 3–4 dB | — |
| Electric bass | Opto 3–4 dB (levels the notes) | Fast FET or VCA 2–3 dB on the attack peaks | Limiter 1 dB |
| Acoustic guitar | Soft VCA 2:1, 2–3 dB | Opto 1–2 dB | — |
| Mix bus | VCA (SSL) 2:1, attack 10–30 ms, 1–3 dB | Limiter or clipper 1–2 dB (for the premix) | — |
Peaks first or level first?
- Fast then slow (the most common choice): the first stage removes the peaks and the next stage sees a more uniform signal, so it works consistently and musically.
- Slow then fast: the first stage gives shape and body, the next stage catches only what slips through. Useful when you want the character of the slow compressor up front (bass, soul vocals).
- Between the two stages you can insert an EQ: cutting the mud before stage 2 keeps the compressor from reacting to 200–300 Hz.
Placement of the other processors
Typical vocal chain: pitch correction → corrective EQ (HP, mud) → compressor 1 (peaks) → de-esser → compressor 2 (level) → color EQ and saturation → sends to reverb and delay. The de-esser between the two compressors keeps stage 2 from exaggerating the sibilance (compression raises the relative level of the consonants).
Gain staging between stages
Compensate the output gain of each stage so that the input of the next one is at the same level (about −18 dBFS RMS). If the first stage comes out louder, the next stage compresses much more than you think.
Settings in FL Studio
- Stage 1: Fruity Limiter in COMP mode (ratio 4:1, attack 2–5 ms, release 50 ms, knee 0) or a 1176 emulator.
- Stage 2: Fruity Compressor (ratio 2–3:1, attack 10–20 ms, release 100–200 ms, “Vintage” type) or an LA-2A emulator.
- Alternatively Maximus in single-band mode with two stages (pre-compression and limiter).
Common mistakes
- Two compressors each doing 8–10 dB: it is just excessive compression split in two.
- Identical stages (same type, same timings): you lose the advantage of dividing the tasks.
- Not compensating the gain between stages.
- Putting the de-esser only at the end, after two compressors have already exaggerated the sibilance.
- Releases that are too short on all stages: distortion and vocal “breathing”.
Starting settings
| Parameter | Value | Why |
|---|---|---|
| Reduction per stage | 3–6 dB | Each stage works lightly and transparently. |
| Total vocal reduction | 6–10 dB | Constant level for modern pop. |
| Stage 1 (peaks) | FET, ratio 4:1–8:1, attack 1–10 ms, release 50–100 ms | Catches only the peaks. |
| Stage 2 (level) | Opto or VCA, ratio 2:1–3:1, attack 10–30 ms, release 100–300 ms | Levels the phrase. |
| Level between stages | About −18 dBFS RMS | Compressor 2 works as intended. |
| De-esser | Between stage 1 and stage 2 | Avoids exaggerating the sibilance. |
Frequencies to watch
| Range | What to do |
|---|---|
| 100–250 Hz | Vocal proximity effect: cut before compressor 2, otherwise it reacts to the mud. |
| 5–9 kHz | Sibilance exaggerated by compression: de-esser between the stages. |
| 2–5 kHz | Harshness of FETs with fast attack: check after the first stage. |
Example chain
| Step | Tool | Settings | Why |
|---|---|---|---|
| Corrective EQ | Fruity Parametric EQ 2 | HP 80–120 Hz, mud −2/−3 dB | The compressor does not react to rumble and mud. |
| Compressor 1 | Fruity Limiter (COMP) / 1176 | 4:1, 3–5 dB on the peaks | Peak control. |
| De-esser | Fruity Limiter with filtered sidechain / dedicated de-esser | 5–9 kHz, −3/−6 dB | Sibilance. |
| Compressor 2 | Fruity Compressor / LA-2A | 2–3:1, 2–4 dB constant | Phrase level. |
| Color | EQ and saturation | Presence and air | Final tone. |
| Limiter (optional) | Fruity Limiter / clipper | 1–2 dB | Residual peaks. |
How to check the result
- Vocal crest factor after the chain 8–12 dB (modern pop/rap 7–10 dB); above 14 dB the vocal stays uneven.
- Short-term loudness of the vocal: variation between verse phrases within 3 LU.
- Presence band (4–8 kHz): relative energy must not rise by more than 2 dB between raw and compressed track (sibilance under control).
- Track peak after the last stage between −10 and −6 dBFS.