Synth Science

Loudness

Loudness normalisation and the end of the loudness war

Streaming platforms turn your master down to a target. Every decibel of dynamic range you destroyed to be louder is simply gone.

Updated 2 min read 11 citations Evidence strength 3/5

What normalisation changed

For decades, a louder master sounded better in direct comparison, so mastering pushed level upward at the cost of dynamic range. The logic was sound while playback was uncontrolled.

Streaming services now measure integrated loudness and adjust playback toward a target. A master crushed to be maximally loud is simply turned down further. You paid in dynamics and received nothing — which inverts the whole incentive.

What the research says about compression

The effects of compressor plug-ins on dynamic range have been measured [3], and loudness perception has been studied in relation to dynamic range and interpolation [4]. Work relating dynamic range compression to the semantic descriptors listeners use [5] is particularly relevant to production, because it connects a processing parameter to the words people use about the result.

Compression research from hearing devices is also informative about the perceptual consequences of aggressive dynamic processing [2], and independent dynamic range compression of separate sources has been studied [1] — which maps closely onto multiband and per-stem processing.

Compression and perceived dynamics

Practical targets

Approximate streaming normalisation behaviour
SituationWhat happensImplication
Master well above the platform targetTurned down on playbackLost dynamics for no loudness gain
Master near the targetLittle or no adjustmentDynamics preserved, competitive level
Master well below the targetMay be turned up, or left aloneDepends on platform; headroom preserved
Club playout from fileNo normalisationThe one context where a hot master still competes

The club case is the genuine exception and the reason blanket advice fails. A track destined for a DJ set played from file is not normalised, so the old incentives partly survive there.

Compression is still a tool

None of this is an argument against compression. It shapes envelopes, controls transients, glues elements and creates the pumping that is a stylistic feature of much electronic music. Those are musical decisions with audible consequences, and the semantic-descriptor research [5] is precisely about how listeners perceive them.

What ended is compression used solely to win a volume comparison that no longer happens.

Common questions

What loudness should I master to?
Close to your primary platform's normalisation target, prioritising how the record sounds at that level over the number itself.
Will a quieter master sound weak?
Not after normalisation, which is the point. Compared at matched loudness, the more dynamic master usually wins.
Does this apply to club tracks?
Less so. Club playout from file is not normalised, which is the real exception.
Is limiting bad?
No. Limiting to control peaks is routine. Limiting to gain 3 dB of level you will lose to normalisation is what stopped making sense.

References

Every citation below links to the original peer-reviewed record on PubMed or via DOI. Nothing here is a substitute for medical advice.

  1. Effects of Independent Dynamic Range Compression of Speech and Music Signals on Sound Quality Rallapalli V, Huang Y, Corey R · American Journal of Audiology · 2026 · Journal article DOI
  2. Multiple-source dynamic range compression in hearing devices: What can we learn from music mixing? Corey R · The Journal of the Acoustical Society of America · 2023 · Journal article DOI
  3. THE EFFECTS OF COMPRESSOR PLUG-INS ON THE DYNAMIC RANGE IN MUSIC PRODUCTION ERDAL K · SOCIAL SCIENCE DEVELOPMENT JOURNAL · 2023 · Journal article DOI
  4. Loudness Perception and Dynamic Range Depending on Interphase Gaps of Biphasic Pulses in Cochlear Implants Pieper S, Brill S, Bahmer A · Ear & Hearing · 2020 · Journal article DOI
  5. Dynamic Range Compression and the Semantic Descriptor Aggressive Moore A · Applied Sciences · 2020 · Journal article DOI
  6. Invited Article: Optical dynamic range compression Jiang Y, Zhao S, Jalali B · APL Photonics · 2018 · Journal article DOI
  7. Loudness constancy in healthy older adults: Effects of sound production and visual cues of music playing Honda A, Yasukouchi A, Sugita Y · The Journal of the Acoustical Society of America · 2018 · Journal article DOI
  8. Sound production shows robust loudness constancy when visual cues of musical performance are available Honda A, Yasukouchi A, Sugita Y · Psychology of Music · 2018 · Journal article DOI
  9. Dynamic Range Across Music Genres and the Perception of Dynamic Compression in Hearing-Impaired Listeners Kirchberger M, Russo F · Trends in Hearing · 2016 · Journal article DOI
  10. The academic training of music therapists: Chances of normalisation and specialisation Wosch T · Approaches: An Interdisciplinary Journal of Music Therapy · 2015 · Journal article DOI
  11. Quality and loudness judgments for music subjected to compression limiting Croghan N, Arehart K, Kates J · The Journal of the Acoustical Society of America · 2012 · Journal article DOI