Classical Loudness Before Mastering: 918 Tracks Measured

2026-11-11 ·

The median track that authors mastered with the Classical preset measures −18.6 LUFS on input, and 1.9% of such tracks are louder than −9 LUFS.

This is a measurement of 918 tracks over the 180 days to 11 November 2026, run in Magic Master. The style here is not a genre: it is the mastering preset the author chose. So “Classical” means tracks that authors mastered with the Classical preset, not every track that sounds like classical music. All styles are compared in one table in the music loudness 2026 hub.

Loudness before and after mastering

Measure p10 p25 median p75 p90
Input: integrated loudness, LUFS −29.3 −23.5 −18.6 −15.9 −13.7
After mastering, LUFS −18.2 −18.0 −16.2 −14.1 −14.0

On input, 1.9% of tracks are louder than −9 LUFS, 10.3% sit between −14 and −9 LUFS, and 87.8% are quieter than −14 LUFS. Of those, 39.9% are quieter than −20 LUFS.

Measure Before mastering After mastering
Crest factor, median, dB 18.9 16.1
True peak, median, dBTP — −1.3
LRA, median, LU 8.6 —

LRA in this measurement is given for the input only.

What this means for the author

Classical arrives quiet almost by default: a median of −18.6 LUFS, 87.8% of tracks below −14 LUFS and 39.9% below −20 LUFS. Only 1.9% are louder than −9 LUFS.

Mastering changes the picture little: the median output is −16.2 LUFS, and three quarters of outputs are not louder than −14.1 LUFS. Dynamics fall more than in any other style, the crest factor going from 18.9 to 16.1 dB.

A quiet track is not necessarily a flawed one, but a platform will not always lift it. Normalization raises quiet tracks only within the peak headroom, and a high crest factor means the peaks reach the limit before the loudness reaches the target. See the article on quiet tracks for the mechanism.

Comparison with the overall measurement

Across all 28,646 tracks the median input is −13.8 LUFS, 14.1% are louder than −9 LUFS and 47.0% are quieter than −14 LUFS. The Classical median input is quieter than the overall median (−18.6 against −13.8 LUFS), and 1.9% against 14.1% are louder than −9 LUFS.

After mastering the median is −16.2 LUFS against −14.0 for all tracks. 80% of outputs in this style fall between −18.2 and −14.0 LUFS, against −14.4 and −11.1 overall. Crest factor before and after: 18.9 → 16.1 dB, against 14.3 → 13.9 dB overall.

How to check your own track

Upload your file to the LUFS analyzer: it shows integrated LUFS, true peak and loudness range. Compare the result with the median of your style in the table. If your track is louder than −9 LUFS, master toward about −14 LUFS so that a platform does not have to turn it down. To master the track, open mastering; it takes about a minute.

Method

  • Window. 180 days to 11 November 2026.
  • Sample. Only mastering jobs that finished successfully. Only aggregates are published: medians, percentiles and shares, with no data on individual tracks or people.
  • Style. The preset the author chose when mastering, not a genre classification.
  • Measurements. Integrated loudness in LUFS following ITU-R BS.1770; true peak in dBTP; loudness range (LRA) in LU; crest factor in dB.
  • Percentiles. p10 means 10% of tracks are quieter than this value, p90 means 10% are louder. The median (p50) splits the sample in half.
  • Rounding. To 0.1; shares are percentages of tracks in the style.

Read next

Questions and answers

What is the loudness of a Classical track before mastering?

The median is −18.6 LUFS: half of the tracks are quieter and half are louder. 80% of tracks fall between −29.3 and −13.7 LUFS, out of 918 tracks over 180 days.

How many tracks in this style are louder than −9 LUFS?

1.9% of them. Another 10.3% sit between −14 and −9 LUFS, and 87.8% are quieter than −14 LUFS.

What does style mean in this measurement?

The preset the author chose when mastering in Magic Master. It is not a genre: the author made the choice, and the measurement did not classify the sound.

What should I do if my track is louder than −9 LUFS?

Master toward about −14 LUFS with a true peak no higher than −1 dBTP. A normalizing platform turns down a track louder than its target, and it does not end up louder in the feed.

Try Echoteka

Upload a track — the same mastering engine, prices in euros.

Open Echoteka →AI mastering onlineLUFS analyzer