For interior designers, architects, and 3D visualizers, short vertical video formats (Instagram Reels, YouTube Shorts, TikTok) have become the single most profitable channel for landing high-ticket clients. Yet 95% of studio founders and solo designers face the exact same roadblock: filming talking-head videos drains enormous time, creative energy, and project focus.

Setting up studio lighting, clipping lavalier mics, battling camera hesitation, recording endless retakes, and editing subtitle timing after hours creates rapid fatigue. Meanwhile, world-class interior 3D renders gather digital dust in portfolios without generating active inbound inquiries.

In this comprehensive guide, we unpack a completely automated, zero-camera video production pipeline built for interior studios: from free neural voice cloning in Fish Audio to directing bespoke motion graphics with an AI agent.

🏗️ End-to-End Pipeline Architecture

The system functions on an atomic, modular architecture where each engine handles one phase with uncompromising fidelity:

1. SCRIPT & TIMECODES → 2. FISH AUDIO VOICEOVER → 3. GRAFIKA DIRECTOR AGENT → 4. NANO BANANA 2 STORYBOARD → 5. OMNI FLASH ANIMATION
  1. High-Impact 10-Second Script: Spotlights a single design pitfall (kitchen work triangle mistakes, hidden tiling markups, architectural lighting layers, socket placement).
  2. Free Studio Voiceover & Voice Cloning in Fish Audio: Accurately clones your natural vocal tone from a 20-second sample and outputs high-fidelity audio with exact timestamps.
  3. AI Agent Powered by the "Grafika Director" Skill: Analyzes phrase timecodes and automatically directs a camera-free motion graphics storyboard (pure vector design, floorplans, stats, and kinetic cards).
  4. Nano Banana 2: Generates a unified storyboard sheet in your studio's brand palette with strict mobile safe zone compliance.
  5. Gemini Omni Flash: Synthesizes the storyboard into smooth, fluid motion graphics animated over your interior renders while maintaining frame-accurate audio sync.

🎙️ Step 1. Fish Audio: Free Voice Cloning & Timed Voiceover

Traditional voice synthesis tools (like ElevenLabs) enforce costly subscriptions and tight character quotas. Fish Audio delivers cutting-edge zero-shot voice cloning in English and Russian with a robust free tier and realistic vocal dynamics.

How to clone your voice and generate audio:

  1. Create a free account at fish.audio.
  2. Navigate to the Voice Clone workspace.
  3. Upload a 15–30 second clean sample of your voice recorded on your phone in a quiet room.
  4. Input your video script (up to 10 seconds of spoken content).
  5. Adjust vocal pacing and generate the studio voiceover.
  6. Download the audio master and record the sentence timestamps (or export an SRT subtitle file).

Pacing Principle: Never let any single sentence linger past 1.5–2.0 seconds in short-form reels. For an 8–10 second video, 5 to 6 concise semantic beats deliver maximum viewer retention.

🤖 Step 2. AI Agent Skill: "Grafika Director"

Instead of designing motion graphics manually, we deploy the Grafika Director agent skill. It turns timed subtitles into two ready-to-run prompts:

  • A JSON prompt for Nano Banana 2 (to create the reference storyboard sheet).
  • A direct animation prompt for Gemini Omni Flash (to drive continuous layer morphing over your video).

Copy this complete skill definition directly into your AI coding assistant (Antigravity, Claude Code, Cursor, or ChatGPT Custom Instructions):

GENERATION PROMPT
---
name: grafika-director
description: Turns timed captions (subtitles with timestamps) for a short VOICE-OVER-ONLY video (up to 10 seconds) into a ready motion-design storyboard for Nano Banana 2 and an animation prompt for Gemini Omni Flash — NO person on screen, pure voice-over motion graphics. Seven brand styles to choose from (asked in two short rounds — category, then specific style — see Step 0 and the palette table in Step 2b), plus a choice of vertical (9:16) or horizontal (16:9) orientation with matching safe-zone rules (see Step 2a). MUST activate this skill when the user's message starts with the word "animation" and includes captions or subtitles with timecodes. The skill analyzes the meaning of each phrase, invents a logical sequence of compositions and infographics to match the words, and outputs two ready-to-use prompts — one for generating a storyboard reference image, the other for animating that storyboard onto real footage in Omni Flash.
---

# Grafika Director — voice-over motion-design storyboard generator for timed captions

## What this skill does
The user sends timed captions for a short voice-over video (up to 10 seconds), with timestamps. There is NO person on screen in this video — only a voice-over and clean motion graphics on screen. The skill:
1. Asks a short round of setup questions (style category, style, orientation 9:16 vs 16:9, standalone vs series part).
2. Analyzes the text by meaning and timing.
3. Invents a logical storyboard — which composition/infographic/number/diagram appears under which words.
4. Outputs a prompt for Nano Banana 2 — generating a storyboard reference image (one sheet with panels).
5. Outputs a prompt for Gemini Omni Flash — animating the video according to that storyboard.

## Trigger
Activates when a message starts with the word "animation" and contains timed captions (text with timestamps, in any format — SRT, VTT, or a plain list of phrases with times).

## Core Rules:
- Hard anchor: NO people, NO presenter, NO faces, NO silhouettes anywhere in any panel — pure motion graphics only.
- Safe zone vertical (9:16): All content in top 70% of screen, bottom 30% always empty (Instagram/TikTok UI safe zone).
- Safe zone horizontal (16:9): Content in center 80%, left/right 10% empty, bottom 15% empty.
- Max 6 panels per 8-10 seconds. Never more than 6 panels.
- No panel static for >1.5-2 seconds without a visual event change.
- Hard rule: NO panel may be text-only. Every panel must include an icon, number, diagram, timeline, or visual metaphor beyond caption text.
- Text-on-panel rule: All text in ONE zone only (top only or center only) to eliminate duplicate text bugs.
- Always end Omni Flash prompt with: "Keep my original audio exactly as it is."

## Style Palettes:
1. Graphite + Warm Gold: background deep graphite #0E0E10, warm gold #C6A461 accents, cream white text. Soft warm glow, faint gold particle dust, thin gold borders.
2. Emerald Editorial: near-black with green undertone #0D1210, emerald green #3ECF8E accents, soft white text. Editorial, biophilic architecture mood.
3. Ice Blue: dark navy graphite #0A0E14, cool ice blue #5EC8FF accents, cool white text. Clean tech, blueprint minimalism.
4. Yellow-Black: pure jet black #000000, bright lemon yellow #F2E400 accents, clean white text. High contrast, punchy design advice.
5. Coral Punch: neutral black #141414, coral orange #FF6B4A accents, warm white text. Warm, human, cozy interiors.
6. Ultraviolet: deep violet-black #12081C, soft lilac #B47CFF accents, pale lavender white text. Modern art-deco, creative interior studios.
7. Mono Red Alert: near-black #0E0E0E, saturated red #E8402C accents, clean white text. Critical renovation mistakes and budget alerts.

## Composition Types:
- FULLSCREEN KEY WORD: Hook or hero word filling frame, soft glow in accent color.
- ICON CASCADE: Card with icon replaces card with icon under each word.
- NUMBER HERO: Giant number filling frame, caption above, thin progress arc in accent color.
- SPLIT CHOICE: Screen splits into two halves (good/bad, do/don't), active side glowing accent, inactive dimmed grey.
- TIMELINE / PROGRESS: Horizontal timeline with stage dots (step 1 → step 2 → step 3), running progress fill.
- FULLSCREEN CONTENT / DIAGRAM: Architectural layout, diagram, comparison, blueprint infographic.