How to add voice over to video

1. Prepare a recording that is easy to explain

Choose one task to demonstrate. Close unrelated windows, remove sensitive details and make the important text readable before you record. Move deliberately enough that a viewer can identify the screen you are describing.

If your recording already contains speech, decide whether the new narration should replace it. Two voices explaining the same step are difficult to follow. Keep a copy of the original recording so you can revise the explanation without losing the source.

2. Write for the action, then the outcome

Watch the recording once before writing. Note the moments when the viewer needs an explanation: a new screen, an important selection or a result. Write one connected passage for each meaningful action rather than describing every cursor movement.

The second version tells the viewer what changed and why it matters. Use the same names for controls that appear on the screen. An attractive sentence is less useful if it describes an option that does not exist.

3. Generate speech and check the pronunciation

With a text-to-speech workflow, choose a voice and generate a first take. Listen for product names, abbreviations, numbers and technical terms. Rewrite ambiguous wording when it helps the voice pronounce a sentence naturally.

You can also record your own narration and combine it with a video in an editor that supports audio tracks. The important choice is whether you want to record the voice yourself or generate it from editable text. Frameberry’s current workflow uses the latter.

4. Fit the narration to what is on screen

Listen while watching, rather than judging the audio in isolation. The words should describe the screen currently visible, not the previous one. If the narration runs past a scene, shorten the explanation or give the scene more time where the tool allows it.

Avoid speeding up the voice just to squeeze in more words. Remove repetition first. Brief pauses help viewers understand a change. In Frameberry, narration timing is checked during generation; edits can require a new preview before the sound matches your latest script.

5. Add captions and keep music in the background

Captions help viewers follow the explanation when they cannot hear it. Review spelling and line breaks, and position captions so they do not cover the controls being demonstrated. If you add music, keep the voice easy to hear and confirm you have permission to use the track.

Frameberry supports optional captions and background music. Library attribution requirements remain relevant after you download the finished video.

6. Review the exported video

Open the downloaded MP4, not just the editor preview. Check the start, each important transition and the ending. Listen once with headphones and watch once without sound. Confirm that the video still makes sense, text is readable and the explanation matches the final footage.

• Names and visible actions are accurate.
• No private information appears in the recording.
• Speech does not overlap an unrelated scene.
• Captions are readable and music does not overpower narration.
• The downloaded file plays as expected.

How this works in Frameberry

Import a recording, optionally explain what viewers should learn and create a first draft. Review editable narration beside the video, adjust the wording and generate an updated result. Download the finished MP4 after checking it.

Open Frameberry, bring a recording and follow the steps with your own project. Make your video or compare the plans.