Slide Decks, Stories, and Audio Media
Summary
This chapter covers MARP web decks and downloadable PowerPoint lectures with speaker notes, illustrated stories and graphic novels, and sourcing freely-licensed images from Wikimedia and government archives with proper attribution. It closes with ElevenLabs text-to-speech narration and glossary pronunciation buttons. Students will be able to generate a slide deck and an audio-narrated glossary entry after this chapter.
Concepts Covered
This chapter covers the following 17 concepts from the learning graph:
- Illustrated Story Format
- Text-to-Speech Narration
- Freely-Licensed Images
- Graphic Novel Format
- ElevenLabs Voice Settings
- Audio Streaming Playback
- Pronounce Button
- generate-images.py Script
- Wikimedia Commons Sourcing
- Government Archive Images
- Image Attribution
- Image Compression
- trim-padding-from-image.py
- compress-images.py Script
- Overlay Quiz Mode
- PowerPoint Lecture Deck
- Speaker Notes
Prerequisites
This chapter builds on concepts from:
- 1. Foundations of AI, Language Models, and Prompting
- 3. Python Fundamentals for Skill Automation
- 7. Progressive Disclosure and Meta-Skill Routing
- 17. Chapter Structure and Content Elements
- 18. Chapter Content Quality and Review
- 19. FAQs and Curated References
- 20. Glossaries and Quizzes
- 21. MicroSim Anatomy and p5.js Basics
- 25. Text-to-Image Models and the Verified Infographic Pipeline
- 26. Interactive Infographic Overlays
The last stretch of media, then straight into publishing.
Decks for the classroom, stories for a first introduction, and a voice for readers who'd rather listen — this chapter rounds out everything a book can be besides plain chapter text. Right tool, right task!
Downloadable Decks: PowerPoint Lectures
Chapter 25's MARP deck lives in the browser; sometimes a classroom needs a downloadable file instead. A PowerPoint lecture deck is a downloadable presentation file intended for classroom use, generated with structured narrative and presenter guidance, carrying speaker notes: presenter-facing text attached to a slide describing what to say and which points to emphasize — visible to the presenter, invisible to the audience on screen.
Telling Stories with Pictures
Some ideas land better as narrative than as definition. The illustrated story format is a narrative retelling of course material paired with generated pictures, used to introduce ideas through story rather than through a direct explanation. For material that needs more room to unfold, the graphic novel format is a sequential illustrated narrative with panels and dialogue, used for extended storytelling within a book. Both are produced with help from generate-images.py: the program that requests illustrations for a narrative and stores them with their descriptions.
A story is a scaffold, not a shortcut.
An illustrated story works because it gives an abstract concept a concrete situation to attach to before the formal definition arrives — it's instructional scaffolding from Chapter 12, told through character and plot instead of through ordered prose.
Sourcing Images Responsibly
Not every image in a book needs to be generated. Freely-licensed images are pictures whose terms permit reuse in educational material, typically with attribution required. Two reliable sources supply them: Wikimedia Commons sourcing obtains reusable illustrations and photographs from a large repository of openly licensed media, and government archive images are pictures from public agency collections, often free of copyright restriction and suitable for educational reuse.
Free to use still means credit is required.
"Freely licensed" is not the same as "no obligations." Image attribution means crediting the creator and stating the license of a reused picture, exactly as the license terms require. Skipping attribution on a Creative Commons image (the same license family from Chapter 1) is a license violation, not a shortcut.
Keeping Images Fast to Load
A page full of unoptimized images loads slowly no matter how well-sourced the pictures are. Image compression reduces the file size of pictures so pages load quickly, while keeping quality acceptable for reading, handled at scale by compress-images.py: reducing picture file sizes across an entire project in one pass. Before compression even happens, trim-padding-from-image.py removes surrounding blank space from a picture so it aligns correctly when placed, avoiding a picture that looks fine alone but sits oddly next to text once it's actually embedded.
Trim first, then compress — the order matters.
Compressing an image with excess blank padding still leaves the padding in the file, just smaller. Trim the padding first, then compress what's left — you'll get a tighter final file either way, and a cleaner layout as a bonus.
One More Overlay Mode: Quiz
Chapters 25 and 26 covered explore mode for an interactive overlay; there's a second mode built on the same markers. Overlay quiz mode is an interaction style in which a reader is asked to identify a named feature by selecting the correct marker, turning the same callout infrastructure from Chapter 26 into a self-check instead of a reference.
Hearing the Book: Text-to-Speech
For a reader who'd rather listen than read, text-to-speech narration generates spoken audio of written material, offering an alternative to reading rather than a replacement for it. The specific voice that reads it is shaped by ElevenLabs voice settings: the parameters controlling a synthesized voice's identity, pacing, and expressiveness. Delivered through audio streaming playback — sound that begins playing before the whole file has downloaded — narration doesn't force a reader to wait through a long download before hearing the first word. At the smallest scale, a pronounce button is a small control beside a defined term that plays its spoken pronunciation, helping readers with unfamiliar vocabulary hear a term like "GraphRAG" or "idempotent" spoken correctly.
Key Takeaways
- A PowerPoint lecture deck with speaker notes serves the classroom; the illustrated story and graphic novel formats, built via generate-images.py, introduce ideas through narrative.
- Freely-licensed images, sourced from Wikimedia Commons or government archives, still require proper attribution.
- Image compression and trim-padding-from-image.py, run via compress-images.py, keep a book's pages fast to load.
- Overlay quiz mode turns an interactive overlay's markers into a self-check.
- Text-to-speech narration, tuned by ElevenLabs voice settings and delivered through streaming playback, plus a per-term pronounce button, make a book listenable as well as readable.
You've now met every media format this book's library can generate.
Slides, stories, sourced images, and a spoken voice — that's the complete media toolkit sitting alongside the chapters and MicroSims you've already mastered. Right tool, right task!