Finagle My Song · Blog
Why two masters at the same LUFS can feel completely different
Pull up two masters, both reading -9 LUFS on a meter. One hits like it means it — kick lands, snare cracks, the track has weight. The other sits at the exact same loudness and feels flat, like someone turned a dynamic performance into a single sustained tone.
LUFS can't tell you which one you're going to get. Something else explains the difference, and it's a number most producers have never looked at: crest factor.
What crest factor actually measures
Crest factor is the gap, in dB, between a track's peak level and its average (RMS) level. It's answering a simple question: when a transient hits — a kick drum, a snare, a plucked string — how much louder is that instant than everything sitting around it?
A big gap means the transients are doing real work above the rest of the track. That's what reads as punch: your ear catches the spike, and the spike is genuinely there in the signal. A small gap means the peaks and the average have been squeezed close together, often by heavy limiting, so there's nothing left for a transient to jump above. The drum hit hasn't vanished — it's just stopped standing out from everything around it.
The bands, and where "heavily limited" starts
Our analyzer classifies crest factor into rough bands: above 12dB is natural, transients intact and doing their job. 7 to 12dB is compressed — controlled, radio-ready, but still has some bite left. Below 7dB, it's flagged heavily limited: peak and average sit close enough that the track is functionally one continuous wall, whatever the LUFS meter says about how loud that wall is. Push it under 6dB and the analyzer calls it out directly as heavy peak limiting, because at that point you can hear it, whatever the intent.
How this differs from DR
If you've read about dynamic range on here before, this will sound familiar, and it should — they're cousins. DR looks at a track's dynamics over time, comparing the loudest passages to the quietest across the whole song. Crest factor is narrower and more instant: peak versus average, moment to moment, everywhere in the file at once.
That's why two tracks can carry a similar DR reading and still feel completely different to listen to. DR tells you the song has contrast between its sections. Crest factor tells you whether the individual hits inside those sections still have teeth. A track can be dynamic section-to-section and still feel lifeless hit-to-hit, if the limiter is working hard enough within each passage.
Where the gap actually goes
Crest factor drops for the same reason DR drops: it's the cost of loudness. Every dB you push a limiter to catch peaks and raise the floor closes that gap a little further. Get a dense track to -9 LUFS and the limiter has almost nothing left to catch — it's working constantly, and the constant work is exactly what erases the transients that used to poke through.
This is also why the fix isn't always "limit less." Sometimes the real problem is upstream, in the mix — a kick that's already been over-compressed on its own channel arrives with no crest factor to protect in the first place, and no mastering chain can hand back a transient the mix never gave it.
Check the number, not just the meter
Next time a master feels flatter than it should at a loudness you're happy with, look past the LUFS reading. Finagle My Song's free analysis shows crest factor alongside loudness and dynamic range, so you can see whether what you're hearing is a genuine mix problem or a mastering chain working harder than the track needs it to.