Home / Video / Educational Film
Updated Jun 25, 2026 · Affirmology_EduFilm_TreatmentAndBible_v1.md
Prepared: 2026-06-25
What this is: the definitive plan for the graphically driven educational segment of the investor film. It fuses your Nano Banana deck aesthetic with researched best practices from the top educational and brand filmmakers. It supersedes the visual direction in Affirmology_EducationalVideo_ProductionPlan_v1.md (that doc's cosmos look is replaced by your emerald-and-gold deck system; its scene logic still informs this).
Lock the emerald-and-gold cosmic system from the deck. Do not redesign it. The film is that exact world in motion.
The single translation from deck to film: density. A slide can hold ten facts because the viewer controls the clock. A film frame cannot, because the film controls the clock. So every dense slide becomes a sequence:
The deck is the storyboard skeleton. The film reveals one idea per frame, in lockstep with the voice. This is the most important rule in the whole bible.
These are the highest-leverage, most-cited rules from the craft research. Build to all of them.
Story and structure - Open on the misconception, not the pitch. Derek Muller's PhD research: clean explanation that confirms what people already think changes nothing; surfacing the wrong belief first, letting them commit, then breaking it roughly doubles retention of the point. Our hook is the viewer's own frustration: "Why don't affirmations work?" - Why, then How, then What (Sinek). Lead with the belief and the change, then the mechanism, then the product and the numbers. The limbic brain decides on the why before the neocortex reasons the what. - The but/therefore spine (Parker and Stone). Every beat connects to the next with BUT or THEREFORE, never "and then." This is what makes the opportunity feel inevitable instead of like a list. - Open loops. Plant a question in the first 10 seconds and do not pay it off until the end. Each section ends on a small curiosity gap the next section opens. - Close on the uplift that resolves the opening tension, then the wordmark and one line in silence. Never end on the problem.
The text layer (carries the dense info) - One idea per frame. One hero line, optionally one quiet supporting line. If a frame crowds, split it into two beats. - Cap on-screen text at a few words for hero beats. The number or the phrase is the hero; the voice points to it rather than reading it. - Reading and hold times are sacred (subtitling science). The reveal animation does not count as reading time. The read clock starts only when the text is fully sharp. Hold a 1 to 3 word hero line about 2 seconds, a 6 to 10 word line about 3.5 seconds, a two-line statement about 5.5 seconds. Never begin the exit before the line has been read. Target a generous 12 to 15 characters per second. - One number at a time for data. Never a table. Big number in, let it land, one line of why it matters, transition out.
Motion - Never linear easing. The premium reveal: blur 10px to 0, translateY +12px to 0, opacity 0 to 1, scale 0.94 to 1, about 500ms, on cubic-bezier(0.16, 1, 0.3, 1). Word stagger 60ms. Use blur when a transition feels rough. - Transitions: cut or a fast 0.3 to 0.5s cross-dissolve on the music beat. Reserve one or two match-morphs (carry a shared element forward) for the biggest pivots. Never decorative wipes or spins. - Micro-motion always. No static frame. Every still gets a slow push-in or parallax drift so it is alive.
Voice, music, sound - Calm, slow, low, intimate voice against propulsive music and alive visuals. The contrast is the engine. The voice withholds energy; the music and motion supply it. - Drop the music to near-silence one beat before the single biggest line, and land a low boom on it. The drop and the held pause are what make the line hit. - Protect voice clarity above everything. Duck music under the voice.
Visual and b-roll - Literal for facts, abstract for feeling. Show the real thing when naming a concrete thing; run beautiful atmospheric footage under the conceptual and emotional lines. - One grade over everything. A single LUT unifies every AI-generated frame into one film. Inconsistent grading is the number-one tell of amateur AI video. - Compose negative space for text. Generate frames with empty sky or fog where the words will live. - Let it breathe. Default shot 4 to 7 seconds, hold the most beautiful frames 8 to 12 seconds. Premium is restraint, not density.
Why, then How, then What, built on your existing slides plus the new market data.
Each beat: ON-SCREEN TEXT (hero) | VOICE (male Atlas, sparse) | VISUAL (Nano Banana frame or b-roll) | hold. Times target a ~2:50 cut. Verbatim brand lines are pulled from Affirmology_CoreThesis_BrandMission_v2.md and the InvestorBrief.
1.1 TEXT: "Why don't affirmations work?" | VOICE: none, let it sit. | VISUAL: dark emerald cosmos, a faint gold waveform pulsing, slow push-in. | hold 2.5s.
1.2 TEXT: "48,000,000 people try every month." | VOICE: "Tens of millions of people try." | VISUAL: the waveform multiplies into a field of faint figures. | hold 2.5s.
1.3 TEXT: "83% say it doesn't stick." | VOICE: "Most say nothing changes." | VISUAL: the figures dim. | hold 2.5s.
1.4 TEXT: "Say 'I am wealthy' - and a quiet voice answers: 'no you're not.'" | VOICE: (the line, slow). | VISUAL: a gold affirmation pulse travels, hits a wall, scatters (your broken-bridge motif). Music dips. | hold 3.5s. This is the misconception rupture.
2.1 TEXT: "There are two roads. Each fails without the other." | VOICE: "There are two roads." | VISUAL: the "Insight" slide, the two cliffs and the unbuilt bridge, revealed as two fields. | hold 3.5s.
2.2 TEXT, left: "The inner tools. Affirmations, tapping, meditation. Delivery without truth." | VOICE: "One has the delivery." | VISUAL: left cliff lights, headphone and heart icons drift in. | hold 3.5s.
2.3 TEXT, right: "The cosmic blueprint. Astrology, Human Design, Gene Keys. Truth without delivery." | VOICE: "The other has the truth." | VISUAL: right cliff lights, chart and DNA icons. | hold 3.5s.
2.4 TEXT: "Two massive audiences. One unbuilt bridge." | VOICE: "And no one has joined them." | VISUAL: the gap between the cliffs glows, the bridge not yet there. Open loop planted. | hold 3s.
3.1 TEXT: "This isn't woo. It's neuroscience." | VOICE: "None of this is magic." | VISUAL: your "Science" slide world, three glowing emblems not yet lit. | hold 2.5s.
3.2 TEXT: "Your brain rejects what it doesn't already believe." | VOICE: "The brain is a prediction machine." | VISUAL: the brain emblem lights, a thought-pulse meets resistance. | hold 3.5s.
3.3 TEXT: "A question it cannot reject. A breath that signals safety." | VOICE: "So we change how it arrives." | VISUAL: afformation and heart-coherence emblems light in sequence. | hold 3.5s. (One match-morph from the brain into the heart here.)
4.1 TEXT: "Your chart becomes audio that rewires you." | VOICE: "We write the audio as your chart." | VISUAL: your "Solution" slide, five islands dark. | hold 3s.
4.2 Pipeline reveal, one island lighting per phrase, the gold thread drawing between them: "Birth data" then "A verified chart engine" then "AI agents read every system" then "Written as your script" then "Voiced, set to music." | VOICE: sparse, names only the first and last. | VISUAL: the islands light boom, boom, boom across the frame. | ~7s total, fast and satisfying.
4.3 TEXT: "Audio engineered to reach the subconscious." | VOICE: "Delivered the only way the brain will accept it." | VISUAL: the final island blooms, a figure receives sound as light. | hold 3s.
One number at a time over the cosmos. Music builds.
6.1 TEXT: "Until now." | VISUAL: the bridge finally completes across the gap, gold light crossing. | hold 2.5s. Pays off the open loop.
6.2 TEXT: "AFFIRMOLOGY. Your Subconscious Operating System." | VOICE: "There is a place in this for you." | VISUAL: the wordmark with the breathing gold star, settle to stillness. | hold 3s, then hand into the trailer.
Total on-screen text is sparse and readable; the voice says even less. The viewer reads the argument while feeling the conviction.
You already proved the hardest part: Claude drafts the frame content, Nano Banana makes it beautiful. We extend that into motion.
Stage 1 - Stills (you already nail this). Generate each beat's frame in Nano Banana in the deck style. To keep the whole film one consistent world: pick one of your existing deck frames as the style reference, carry it into every generation, and reuse a verbatim style tag ("deep emerald cosmic, gold filament linework, ornate gold corner frame, cream text space, cinematic"). Nano Banana has no seed, so consistency comes from the reference image plus the repeated tag. Compose deliberate negative space where the kinetic text will sit. Generate the frames "clean" (without the dense baked-in text) so our sharp text layer goes on top.
Stage 2 - Animate the stills into motion. Match the tool to the shot: - Higgsfield Cinema Studio for controlled, repeatable camera moves on the ornate framed slides (slow push-in, parallax, orbit). This is the fastest path to reliable cinematic motion. - Runway Gen-4 or Google Veo for the atmospheric beauty plates (nebula drift, light, particles). - Kling for any fluid or light-driven motion (the gold thread drawing, the bridge completing). Slow, motivated moves only. Every still gets at least a gentle push or parallax so nothing is static.
Stage 3 - Kinetic text layer (the half of the film that carries the info). Build the text in Remotion (code-driven, on-brand, reusable, and it hits the exact cheat-sheet timings and easings) or CapCut for speed. The text is always a separate sharp layer over the footage, never baked into the Nano Banana frame, so it stays crisp and editable. Use the brand fonts: Cormorant Garamond display, Inter light, JetBrains Mono for labels.
Stage 4 - Voice. The male Atlas voice, sparse, slow, low. Render it the way the Lily trailer VO was rendered. [Confirm the Atlas voice ID.]
Stage 5 - Music. One cinematic, propulsive, building bed from Suno. Cut visuals on its beats. Drop it to near-silence before the "one unbuilt bridge" line.
Stage 6 - Grade and assemble. One LUT over every clip to unify the look. Assemble in CapCut or Resolve: footage on the bottom, sharp text over it, voice and music under. Export 16:9, then reframe a 9:16 cut.
Tell me to start on step 1 (the script) and I will write it next. That is the piece everything else hangs on.