6 min read
Location sound recording: a checklist that prevents post production pain
Almost every post production emergency traces back to a setting nobody checked at call time. Here is the list.
Location sound is a discipline of preparation. The recording itself is minutes of attention per take; everything around it is setup, verification and file management.
Before the first take
- Sample rate and bit depth set to the project spec, normally 48 kHz and 24 bit.
- Frame rate and timecode mode matching camera, jammed and verified on the device.
- File naming set to production, roll and scene, not the factory default.
- Track names entered on the recorder: Boom, Lav_Marie, Lav_Sam, Mix L, Mix R.
- Two cards recording in parallel where the machine supports it.
During the day
- Peaks around minus 12 to minus 18 dBFS, nothing near zero.
- Listen to every take, not just the meters.
- Log notes as you go, they never get written later.
- Rejam timecode after lunch and after any battery change on a sync box.
Before you leave a location
Room tone, thirty seconds minimum, same microphones, same positions. Wild lines for anything the noise floor destroyed. A quick note in the report saying what you captured.
End of day
Copy the cards to two destinations with verification, export the sound report, and check a few files: sample rate, bit depth, channel count, track names and start timecode should all be exactly what you set at call time.
That final check is the cheapest insurance in the department. An inspector that reads the metadata locally in a browser takes seconds and needs no upload, which matters when the material is confidential.
If post asks for mono stems
Split the poly files without touching sample rate, bit depth or channel order, and let the recorder's track names drive the output file names. Deliver the originals as well, always.