Spoken narration from typed words in Video by Zaweba

Not every video has someone available to record the narration. A draft needs a guide voice before the real one is booked, a tutorial needs a calm explainer, a series of short episodes needs dozens of lines read out. Video by Zaweba can speak typed words or script voice lines using the offline voices from Voice by Zaweba, and put the results straight onto the timeline as takes.

What Voice by Zaweba is

Voice by Zaweba is a separate Zaweba app: an offline voiceover and singing voice maker. Its voices run on your own PC, which means your script isn't sent anywhere to be spoken. Video by Zaweba uses those voices, so you need Voice's voice pack installed for this to work.

From words to takes

There are two ways in. You can type words and have them spoken, which suits a one-off line or a quick title read. Or you can speak the voice lines from a script, which is where it shines in the Episode Builder, where a script with numbered shots and voice lines is already part of the project.

The spoken lines land on the timeline as takes. That's the same idea as recording your own voiceover take after take: you can have several versions of a line and pick the one that works. If one reading puts the stress in the wrong place, make another and compare.

Jobs it's good for

Guide tracks come first. Editing to a spoken guide is far easier than editing to silence, because you can cut pictures to the rhythm of the words. Later, a real voice can replace the guide line by line.

Tutorials and explainers are another. A clear, even voice reading steps out loud works well over screen recordings.

Series built from short AI clips are the third. When you're assembling episodes from a script, having the voice lines spoken offline means a whole episode can be put together without a recording session.

Getting natural sounding results

Write for the ear. Short sentences read better than long ones. Spell out anything that might be misread, and use commas where you want a breath.

Make more than one take of important lines and pick the best, just as you would with a human reader.

Listen to the whole narration in context before you settle on takes. A line that sounds fine on its own can feel rushed or flat next to the pictures, and hearing it against the cut tells you which version fits the pace of the scene. Names and unusual words deserve a special check, since they're the most likely to be read in a way you didn't expect.

Treat the spoken lines like any other dialogue in the sound pass. Tag them as dialogue, match loudness with the other speech, and let the music duck under them. A voice EQ or a touch of room sound can help a generated voice sit with the rest of the mix.

Where it fits with recorded voiceover

Video by Zaweba also records voiceover straight onto the timeline. The two complement each other: spoken lines for a fast first version or for projects with no narrator, your own recording when a personal voice matters. Since both arrive as takes on the timeline, swapping one for the other later is a small job rather than a re-edit.