How Compressed Is a Hit Vocal?

How much compression do hit vocals use?

One number answers it: a gated crest factor of about 16 dB. Crest factor is the gap between a vocal's peaks and its average level while it is actually singing. Lower means more compressed, higher means more dynamic. Half of all sung hit vocals land between 15.5 and 16.7 dB, and 71% land inside a 2 dB window from 15 to 17. That window holds in every genre, in every decade since 1990, and at every chart position.

It is not a ratio or a threshold. Two vocals can both arrive at 16 dB with completely different settings: a gentle 2:1 with a fast attack on a steady singer, or a 4:1 with a slower attack on a dynamic one. What the hits agree on is the result, and the result is measurable on your own vocal.

Crest factor by genre

Sung vocals only (rap and sparse-vocal stems are gated out of this set). Window share is the percentage of that genre's vocals that land inside 15 to 17 dB.

The spread between the most compressed genre and the least is 1.2 dB. Gospel and rock vocals are squeezed the hardest, R&B the least. Hip-hop is measured separately because most rap stems fail the "sung" gate; run through the same pipeline on every song, rap vocals sit around 17.6 dB, the most dynamic on the chart. A rapper's transients survive because the vocal rides above the beat instead of inside it.

What the number means for your compressor

1. Measure before you set anything. A vocal above 17.8 dB is more dynamic than 90% of hits. Below 15.1 dB, it is more squashed than 90% of hits. Between those you are already in professional territory and the compressor's job is tone, not control.
2. Get there in two stages. A slower first compressor that catches the loud phrases, then a faster one that shaves the peaks, arrives at 16 dB with less audible pumping than one compressor doing 8 dB of gain reduction. The compressor settings post works through the ratio, attack and release that produce this window.
3. Don't chase the number with the threshold alone. If the vocal reads 19 dB, the problem is usually a level that jumps between phrases, and that is a riding job. See the next section.

Compression or level riding?

Crest factor has held in the same window since 1990, but a different measure of dynamics has not. Dynamic range measured the other way, loud sections against quiet sections across the whole take, fell from a median of about 35 dB in the 1990s to about 25 dB from the 2000s on. Modern vocals are not more squashed peak to peak. They are more level, phrase to phrase and verse to chorus. That is fader riding, clip gain and automation, not a harder compressor, and it is the part most home mixes skip. The waveform comparison shows what it looks like across five #1 vocals, and magic.RIDE Lite does the riding automatically so the compressor only has to do the last few dB.

Vocal compression FAQ

How much compression should I put on vocals?

Aim for the result, not a ratio: a gated crest factor of about 16 dB, where 3,054 hit vocals land. Half of all sung hit vocals sit between 15.5 and 16.7 dB and 71% inside 15 to 17 dB. Above 17.8 dB you are more dynamic than 90% of hits; below 15.1 dB, more squashed than 90% of hits.

Do #1 hits compress vocals harder than other hits?

No. The 498 #1s have a median crest factor of 15.99 dB; the other 2,552 songs, 16.04 dB. Chart position has no relationship with vocal compression.

Are modern vocals more compressed than 90s vocals?

Peak to average, no. What changed is level consistency between sections, which dropped by about 10 dB between the 1990s and the 2000s and has been flat since. That is riding and automation, not compression.

Will 16 dB read the same on my meter?

No. Crest factor depends on the gate, the peak percentile and the frame size, so other meters give different absolute numbers. The free vocal analyzer uses the identical pipeline that measured the hits, so its number is directly comparable.