01Format and sample rate
WAV is the usual master for editing, because it is uncompressed. MP3 is for review copies and for platforms that ask for it.
For video work, 48 kHz is the standard sample rate, and 24-bit keeps headroom for the mix. For audio-only work such as podcasts or audio guides, 44.1 kHz is common. For phone systems, 8 kHz is typical.
02Loudness targets
Loudness is measured in LUFS or LKFS. Each destination has its own target, and meeting it is part of the delivery.
- European broadcast (EBU R128)
- −23 LUFS, true peak no higher than −1 dBTP.
- US broadcast (ATSC A/85)
- −24 LKFS, true peak no higher than −2 dBTP.
- Web and streaming
- Platforms commonly normalise between −14 and −16 LUFS, so −14 LUFS is a usual target.
03Splitting and naming
Decide how the files are split: one file per line, per scene or one full mix. Splitting per line makes corrections easier, because one line is one file.
Name files to your convention, with the language code and a version number, for example DE_scene03_v2.wav. Changing names later is where mistakes enter a pipeline.