Writing voiceover for a 10-second scene

Each scene's picture is 10 seconds long. A Vietnamese voiceover at a natural pace covers about 3.8 words per second (female voice) to 4.15 words per second (male voice), so 10 seconds fits about 38 to 41 words.

Why 15 to 45 words

• Under 15 words: the voiceover ends early and the rest of the scene is silence.
• 15 to about 38–41 words (depending on the voice): the voiceover is shorter than or just fits the picture, which plays at normal speed.
• From there up to 45 words: the voiceover is longer than the picture, so the picture is slowed down to match. At 45 words it plays at about 84–92% speed and the scene runs about 11 to 12 seconds.
• Over 45 words: the system won't accept it. Move the extra idea to the next scene.

On the episode editor, the voiceover box shows the current word count. Past the comfortable fit (38 words for the female voice, 41 for the male voice), the page also tells you how much the picture will slow down, as a percentage.

Cutting words without losing meaning

• Drop filler like “really”, “you could say”, “as we all know”.
• Say one thing per scene. Save the second point for the next scene.
• Round long numbers: “almost two thousand” instead of “one thousand nine hundred and eighty”.

If it is still too long after cutting, the cleanest fix is to split it into two scenes. An episode has up to 10 scenes, plenty of room for that.