1. Make the words the reference point
Start with the approved narration and the loudest realistic background bed. Lower music and effects before adding processing. If the sentence is unclear at ordinary volume, more loudness is not the same as more clarity.
Keep a clean voice export before mixing. It lets a new editor rebalance the project without regenerating the speech and gives you a trustworthy comparison when the mix changes.
2. Check the environments people actually use
Listen on headphones, a laptop, a phone speaker, and any small speaker that matters for the product. Move between the quietest and busiest moments of the edit. Plosives, sibilance, low-level words, and crowded effects often appear only in one environment.
Do not make a mix universally loud to solve a context problem. A quieter listener may need device controls, while a dense background may need arrangement changes or fewer competing sounds.
3. Export a predictable handoff
Name the master, review copy, and final delivery clearly. Preserve the sample rate and format required by the destination, and keep a lossless or unprocessed source when the platform permits it.
Write down the voice, speed, mix notes, loudness process, and date. This short handoff prevents a future revision from guessing which file was approved.
4. Use a four-pass release check
- WordsListen for names, numbers, clipped endings, and breaths that change meaning.
- BalanceCheck the opening, busiest moment, and final call to action against music and effects.
- DevicesListen on headphones, laptop speakers, and a phone at ordinary volume.
- HandoffConfirm filename, format, transcript, source master, and approval note.
