Video summary

Audio Compression for Dialogue & Vocals: EASY Guide to PRO Results

Main summary

Key takeaways

Technology

Summary (Technology/Tools Focus)

The video explains why vocal recordings can sound “flat and thin” and argues that the main culprit is typically incorrect (or missing) dynamic range compression. The goal is to help dialogue/vocals sound more polished, rich, and forward in the mix—so it “pops” and stands out like professional podcasts/TV/hip-hop productions.

Key Concepts: What a Compressor Does

  • A compressor reduces the loudest parts of an audio signal (“squeeze them down”).
  • It then applies makeup gain to raise the overall level back up, since compression otherwise makes the signal quieter.
  • The video emphasizes that proper compression helps control wide dynamic range (e.g., whisper → yell in one take).

Walkthrough: Stock Pro Tools Compressor (Core Parameters)

The presenter demonstrates using the stock Pro Tools compressor (with a vocal preset) and reviews the main controls:

  • Threshold: when compression starts

    • Set lower → more content gets compressed
    • Set too low → too much compression; set too high → little to none

  • Ratio: amount of “squish”

    • ~1.8:1 shown as gentle
    • Higher ratios approach limiting behavior

  • Attack: how quickly compression clamps after crossing threshold

    • Too fast can remove vocal “attack” / clarity, smearing consonants (“smooshed over”)
    • Suggested vocal starting point: ~5–10 ms

  • Release: how quickly compression stops after dropping below threshold

    • Too abrupt can cause audible “pops”
    • Suggested starting point: ~100–200 ms for transparency

  • Knee: how smoothly compression transitions into action

    • “Zero knee” = sharper transition at threshold

The video also uses waveform comparisons (raw vs. processed) to show how compression raises quieter sections (whispers) relative to louder ones.

Tool Comparisons / Examples

1) FabFilter Pro-C 2 (spoken word vocal preset)

  • Demonstrated with a preset like “spoken word squeeze.”
  • Claims it’s transparent but expensive.
  • Waveform inspection suggests it likely uses stronger makeup gain, raising whisper sections more noticeably.

2) AutoComp (custom plugin from the creator)

  • The presenter’s plugin is positioned as making compression “easy,” with only a few controls.
  • The video says they created two plugins, but focuses mainly on AutoComp (a channel strip-style compressor/leveller).
  • Reported UI/controls:
    1. Lift: boosts low-level signals that would otherwise stay below threshold
      • Solves the issue where whispers don’t get compressed unless you lower the threshold (which then over-compresses louder parts).
    2. Comp: one knob controlling multiple underlying compression behaviors (automatic makeup gain, ratio, knee, etc.)
    3. Trim: output gain aimed at target loudness for platforms
      • Example given: YouTube -14 LUFS target
    4. Safety limiter (optional/turn-off)

Tests are performed at default settings (example: 50% compression, lift on, limiter on) and then at 100%, with the claim that AutoComp brings whispers up close to loud sections more than the other compressors tested.

Listening comparisons include:

  • Raw
  • Pro Tools stock compressor
  • FabFilter Pro-C 2
  • AutoComp

Takeaway: AutoComp most effectively levels whisper-to-loud performance while remaining relatively transparent (with some “rubbery” coloration mentioned for at least one compressor).

Practical Guide: Dialing In Compression for Vocals/Dialogue

The video provides a step-by-step approach using a compressor with metering/visualization:

  1. Set threshold first

    • Adjust so the threshold “break point” sits around the peaks of the vocal signal
    • Too low = overcompression; too high = no compression

  2. Set ratio

    • Suggested range for vocals: ~2:1 to 5:1
    • For aggressive hip-hop vocals: possibly ~8:1 to 10:1

  3. Aim for moderate gain reduction

    • Rule of thumb: 3–6 dB gain reduction
    • More may be needed for aggressive sounds; emphasizes “trust your ears”

  4. Use knee

    • Use a ramp/smoother knee to ease into compression (instead of abrupt onset)

  5. Set attack/release targets

    • Attack: around ~5 ms
    • Release: around ~150 ms

  6. Depth (optional)

    • Example: allow only a certain max reduction (e.g., -6 dB), otherwise leave it off

  7. Mix/Wet control

    • If blending dry with processed (“mix” not 100% wet), you can retain some dynamics for naturalness
    • For diagnosis, set to 100% wet

  8. Makeup gain / loudness matching

    • Use makeup gain both to compensate for level lost due to compression and to approach loudness targets (Spotify/YouTube/Netflix/Amazon, etc.)
    • More compression typically requires more makeup gain

Common Mistakes

  1. Too much compression (threshold too low, ratio too high)

    • results in loss of dynamics; everything becomes exaggeratedly present (breaths/mess)

  2. Not enough compression

    • quiet parts get buried; loudness/ear safety issues

  3. Attack too fast

    • loses consonant percussiveness / clarity

  4. Not enough makeup gain

    • compressed audio stays too quiet and doesn’t “pop”

  5. Compressing everything (music, dialogue, SFX all equally)

    • causes messy mix and listener fatigue
    • recommends compressing dialogue most, leaving music/SFX more dynamic

Mentioned Product/Offer

  • AutoComp is promoted as a paid plugin (price: $50) on the creator’s website.

Main Speakers / Sources

  • Primary speaker/source: The video creator
    • Introduces and demonstrates Pro Tools, FabFilter Pro-C 2, and their plugin AutoComp.
  • Products/tools referenced as sources:
    • Avid Pro Tools (stock compressor)
    • FabFilter Pro-C 2
    • The creator’s AutoComp plugin

Original video