Voice boost for video — music and effects held back

Cinema is mixed for a room with a wide dynamic range: effects and music sit far above the lines, and that is on purpose. At home a second step spoils it — the multichannel track is folded down to stereo and the center channel carrying the dialogue is blended into the music. Turning the whole thing up does not help: the explosions get louder by exactly the same factor. Only changing the balance helps, and that is what this tool does.

The ideal case is a 5.1 track: the dialogue lives in its own center channel and the tool takes it whole, with nothing to guess. In stereo the center has to be computed, and the outcome depends on the mix. If music and effects are spread to the sides — which is how most films and series are mixed — the background drops away clearly. If the bass and drums sit dead center next to the speech, they cannot be told apart spectrally: the tool then works as a speech EQ that ducks the background under the lines. The “Original” button and the live meter show you which of the two you got.

A booster multiplies everything, so the ratio between speech and music never changes. Here the track is first split in two — voices and everything else — and then recombined in a different proportion: speech louder, background quieter, and the background ducked further while a line is being spoken.

While watching, no — the source file stays byte for byte the same and the processing lives only in the browser audio path. Press “Save” and a new file appears next to it; the original is still untouched.

Saving lays the whole audio track out in the tab's memory, so length is capped — the tool works the exact limit out from your device's memory and shows it in the save panel. Watching is not capped: there the audio is processed on the fly, in small pieces, and length does not matter.

Two different repairs. Lifting the voice pulls speech forward and makes quiet delivery audible, but it lifts whatever shares the same frequencies — room noise included. Lowering the background pushes music and effects down and leaves the voice where it was, which suits a mix that is simply too loud behind the dialogue. The four presets are combinations of the two; the separate sliders are there for when a preset gets you most of the way.

Yes — load an .srt or .vtt file, set the size and shift the timing if the subtitles do not match your copy of the video. The offset is remembered per file.