In this article
You’ve sent two takes and the producer says one feels louder. When you open them, their loudest peaks are almost identical. That’s possible. The peak is one instant. The producer is hearing the level across the words.
Run a simple listening test
Take two versions of the same sentence and compare the level of the steady speech between strong syllables. If one version has been heavily compressed, the gap between the louder and quieter words may be smaller even when the biggest peak is the same. Listen to both versions at a similar perceived playback level before deciding which sounds more natural.
Check which meter the brief is talking about
Peak normalization turns the whole file up or down to a target peak. It leaves the relationship between the loud and soft parts intact. An integrated loudness target refers to measurement over the specified audio, and a true-peak ceiling is another measurement. Follow the producer’s instructions about what to measure and where in the workflow to measure it.
If the deliverable is a raw track for an editor, ask for their preference before adding compression or limiting. A voice file that will be part of a finished mix can have a different specification from the mix itself.
Find the problem before changing the gain
If you’re struggling to hear the last word of a sentence, check the read and the setup. Did the speaker turn their head or run out of breath? Is there a fan masking the softer words? A repeatable home recording setup makes it easier to work out which element is changing.
Play the export away from the session
Check the file you’ll actually upload. Listen to the start, a quiet phrase, the loudest passage, and the end. Compare any replacement lines with the original take. The checklist before sending an audition covers the final naming and file checks to do at the same time.
One response to "Why two voiceover files with the same peak can sound different"
Nice, simple, explanation of some basic and important editing rules!