A narrated slide deck needs a script that tells listeners what matters while each slide is visible. Start with one audio segment per slide, then connect the segments into a coherent explanation. The practical deliverable is a small production sheet: slide identifier, visible information, spoken words, and the event that tells the editor when to advance.
Give every slide one speaking job
Write the slide's purpose before writing narration. A title slide identifies the subject; a comparison slide explains a choice; a demonstration slide describes an action. If one slide has several unrelated speaking jobs, either divide it or explicitly signal the change in the script. This keeps the audio from sounding like a list of disconnected labels.
For a slide with three project timelines, a useful purpose is “explain why option B finishes earlier.” Read the dates needed for that comparison and explain the dependency. Reading every date gives the listener more work without answering the slide's question.
Keep stable slide IDs such as S010 and S020 alongside the visible titles. Titles change during review, and slide numbers shift when an introduction is added. A stable ID lets a reviewer request a replacement without accidentally replacing a neighboring segment. Carry that ID into the audio filename and the final handoff.
Separate the screen text from the spoken explanation
Use four columns: slide ID, screen content, narration, and assembly notes. The narration column should contain only words intended to be heard. An instruction such as “highlight the second bar” belongs in assembly notes; otherwise it can accidentally enter the generated audio.
Write a complete explanation that survives a listener looking away. Replace “as you can see here” with “the shorter timeline removes the second approval step.” For complex figures, use the approach in describing charts in audio lessons. The goal is to preserve the useful relationship, not describe every decorative element.
Avoid making narration a perfect duplicate of dense slide text. Decide which information belongs in the narration and which belongs in a companion reference. If a slide presents an exercise, speak the task and explain how to pause. Leave the full dataset available as readable text.
Measure timing with the real audio
Generate a draft for one representative slide before completing the deck. Include a slide with a difficult term or a complex visual, because a short title slide tells you little about practical pacing. Listen while advancing the slide manually and record where the explanation feels rushed.
Do not estimate slide duration from word count alone. Listeners also need time to locate a chart, compare options, or try a step. Add those intervals in your presentation or editing software. VocalCopyCat can supply downloaded narration; the slide application handles playback placement and visual timing.
Microsoft's presentation recording guidance explains narration and timing controls. Check the controls available in your installed version. If the deck changes from a self-running video to a live presentation, reconsider the timing assumptions rather than reusing every automatic advance.
Deliver a deck someone else can update
Package the final script, downloaded audio, slide file, and a short assembly sheet together. Mark each segment approved, revision needed, or intentionally silent. Record the voice used and any pronunciation decisions so the next update can match the existing narration.
Review the assembled deck from beginning to end. A correct audio file can still be assigned to the wrong slide, cut off by an advance, or followed by a confusing silence. Check the opening words of every slide and the transition into the next one.
Provide text alternatives for the audio lesson alongside the finished presentation. For a practical first pass, create a three-slide sample in the voice studio, assemble it in your presentation tool, and approve the workflow before producing the remaining deck.
