Are modern vocals more compressed than older ones?

Peak to average, no. The crest factor of hit vocals has stayed in the same 15 to 17 dB window since 1990. What changed is level consistency: the median dynamic range between loud and quiet sections fell from about 35 dB in the 1990s to about 25 dB from the 2000s on, and has been flat for 25 years. Modern vocals are ridden and automated to a constant level, not crushed by a compressor.

What dynamic range should a vocal have?

Measured as the loudest 10% of the vocal against the quietest 10% over a 90-second segment, hit vocals from the 2000s onward have a median of about 25 dB. Modern pop and rap #1s can sit under 10 dB, like Doja Cat's Paint The Town Red at 8.9 dB. Ballads stay far more dynamic.

Why does Blinding Lights have 52 dB of dynamic range?

Because it was produced to 1980s conventions on purpose. The vocal sits in a big reverb and moves with the song instead of being pinned to one level, which is why it is the most dynamic modern pop #1 in the dataset.

30 Years of Vocal Compression, in Waveforms

Five iconic Billboard #1 vocals, stacked at the same scale. The more brick-shaped the waveform, the more compression was applied. Whitney in 1993 looks like a mountain range. Doja Cat in 2023 looks like a wall.

Loading waveforms…

Finding

A 1993 #1 vocal had 32 dB of dynamic range. A 2023 #1 has under 9 dB. Listen to each and the difference is subtle — but look at the waveform and it's obvious.

Context

Dynamic range is a measure of how much the vocal level varies between its loud and quiet sections. Heavy compression squashes that variation to a near-constant level — reads as a brick in the waveform.

Outlier

The Weeknd's Blinding Lights sits at 52 dB — the most dynamic modern pop #1 in the dataset. That's because it was deliberately produced with 80s retro conventions, not modern pop conventions.

See the compression trend across all 3,054 hits →

Dynamic range by decade, across all 3,054 hits

The five waveforms above are the extreme cases. Here is the same measurement, the loudest 10% of the vocal against the quietest 10% over a 90-second segment, for every hit in the dataset, grouped by decade.

Decade	Songs	Median vocal dynamic range
1990s	819	35 dB
2000s	611	24 dB
2010s	907	26 dB
2020s	713	25 dB

The drop happened once. Between the 1990s and the 2000s the median fell by about 10 dB, and it has barely moved in the 25 years since. Whitney's 32 dB in 1993 was typical for her decade. Eminem's 8.9 dB in 2003 and Doja Cat's 8.9 dB twenty years later are the same number, from the same era of vocal production.

Compression, or level riding?

This is the part the waveforms alone can't tell you. On the compression page I measured crest factor, the gap between a vocal's peaks and its average level while it is singing, and it has stayed inside the same 15 to 17 dB window since 1990. So the modern vocal is not more squashed peak to peak. It is more level: quiet phrases brought up, loud phrases brought down, verse and chorus held at the same height. That is fader riding, clip gain and automation, applied before the compressor ever sees the signal. The brick-shaped waveform is the fingerprint of a vocal that has been ridden flat, then compressed the same amount Whitney's was.

It also explains the outliers. Hillsong's Oceans at 67 dB and The Weeknd's Blinding Lights at 52 dB are not less compressed. They are records where the vocal was allowed to move with the arrangement instead of being pinned in place.

How to get the modern shape without crushing the vocal

Ride first. Level the vocal phrase by phrase before compression. Done by hand this is clip gain and automation; magic.RIDE Lite does it automatically and for free.
Then compress into the window. With the level already even, a compressor only needs to shave the last few dB. Land the crest factor near 16 dB, where the hits are, instead of using 10 dB of gain reduction to fake consistency.
Check the result, not the settings. The free vocal analyzer measures your vocal's dynamics with the same pipeline used on these hits and tells you which side of the window you are on.

Waveform FAQ

Are modern vocals more compressed than older ones?: Peak to average, no. Crest factor has stayed in the same window since 1990. Level consistency between sections is what changed, by about 10 dB between the 1990s and the 2000s, and it has been flat since.
What dynamic range should a vocal have?: By this measure, hit vocals from the 2000s onward have a median of about 25 dB. Modern pop and rap #1s can sit under 10 dB. Ballads stay far more dynamic, and none of these numbers are targets on their own. Match the genre you are making.
Why does Blinding Lights have 52 dB of dynamic range?: Because it was produced to 1980s conventions on purpose: the vocal sits in a big reverb and moves with the song instead of being pinned to one level.

Analyze your own vocal →