A GUIDE FOR VOICEOVER ARTISTS

Loudness specs for voiceover spots

The numbers that come up on commercial work, and which ones are yours to hit. Loudness targets in LUFS and LKFS are for the finished spot. The dry read you hand over is judged on peaks, a steady level and a quiet room.

If a producer, station or network sends you a spec, theirs wins. Everything below is for when nobody has.

The finished spot

Integrated loudness over the whole spot, and the highest true peak allowed. Whoever mixes the spot hits these, which is you only if you deliver it produced.

The CALM Act covers television only, so a radio spot has no legal loudness target in the US. Stations run their own processing on air, so ask for the station’s delivery spec.

The dry read

A dry read goes to someone who will edit and mix it, so leave it unprocessed and don’t normalise it to −24 or −16 LUFS: that is the mix’s job. What helps the producer is room above your loudest word, an even level from start to finish, and silence between lines. These are the limits the SendScripts.to recorder checks every take against, until you set your own:

Peak

≤ −3 dBFSHeadroom for the producer’s processing.

Average (RMS)

−24 to −14 dBFSLoud enough to work with, not squashed.

Room noise

≤ −60 dBFSThe gaps between lines should be quiet.

These sit close to the audiobook rules on ACX, which many voice artists already know: an average of −23 to −18 dB RMS, peaks no higher than −3 dB and a noise floor no higher than −60 dB. Auphonic on the ACX rules

The terms

LUFS / LKFS

Loudness units relative to full scale: how loud the whole spot sounds, measured the same way under both names (ITU-R BS.1770). Europe says LUFS, US television says LKFS.

LU

A loudness unit, the step on that scale. One LU is one decibel.

dBFS

Decibels below full scale, the loudest a digital file can go (0 dBFS). Peaks and averages are read on it.

dBTP

True peak: the highest level once the audio is turned back into a waveform, which can land a little above the highest sample. A meter that shows sample peak, as the recorder does, can read lower than true peak, so leave a little room when a spec is in dBTP.

RMS

The average level over a stretch of audio. ACX and the recorder’s Average check read it.

Integrated

Measured over the whole spot from start to finish, rather than moment to moment.