🎬 THE DIRECTOR – AI Image, Script & Video Generation Tool for Creators
v2.0 – April 16, 2025
Built for influencers, storytellers, educators, and visual content pros.

Turn ideas into cinematic short scenes — with images, scripts, voiceovers, and sound effects — all in one seamless AI pipeline.

🎥 Walkthrough video: https://youtu.be/L7SYD_pbraA

🌟 What’s New in v2.0

• ✅ MMAAudio Support – auto-generates matching sound/music for each scene

• 🚀 Faster Rendering – scenes now complete in under 4 minutes on Ultra Pro

• 🧠 Start from Text or Image – describe your character or provide a visual

• 🎞️ Improved Scene Generation – better detail, consistency, and flow

• 🛠️ Modified Interface – cleaner, more intuitive node layout

• 🧩 Seamless integration with ComfyUI

🔊 Complete 3-scene video with audio in under 17 minutes

📘 Version History

v1.0 – March 27, 2025
• Initial Release – basic scene generation and rendering engine

🧭 STEP-BY-STEP WORKFLOW GUIDE

🔹 Step 1 – Get a Gemini API Key
• Go to: https://aistudio.google.com/apikey
• Log in with your Google account
• Click "Create API Key"
• Copy it into the purple GEMINI API KEY node in ComfyUI

🔹 Step 2 – Choose a Reference Image
• Upload a visual reference OR
• Set Use Reference Image = False to start from a text description
• (Optional) Enter a project name for easy folder organization

🔹 Step 3 – Enter a Story Prompt
• Be as detailed or brief as you like
• If you skipped the reference image, describe your character clearly

🔹 Step 4 – Select Mode + Audio
• Choose: Portrait or Landscape
• Toggle MMAudio ON if you want auto-generated sound

🔹 Step 5 – Click QUEUE
• Generation begins — each step takes ~32 seconds
• Increasing the number of steps improves quality (but takes longer)

🔹 Step 6 – Select Your Scenes
• You’ll see image batches (4 at a time)
• Pick your favorites and drag them into order (1–24)
• Hit Cancel + Retry if the results aren’t what you want

🔹 Step 7 – Generate Video & Stitching
• Images are rendered into scenes
• Selected scenes are stitched into a full video (with audio, if enabled)

🔹 Step 8 – Retrieve Your Final Video
• Check the output folder
• Look for the .mp4 file (with "audio" in the name if you chose MMAudio)

💡 Pro Tips:
• For best results: allow ~4.5 mins per scene on Ultra Pro
• Want higher detail? Increase steps from 8 → 16
• Imperfect outputs? Cancel & re-run! AI isn't flawless — but it is fast

👥 Creators & Credits

AJO6268 aka KurtCPhotoEd
Clark Glenn Davis aka Verevolf
SoundTech: manu_le_surikhate_gamer

🧰 INSTALLATION GUIDE – GETTING STARTED WITH THE DIRECTOR

✅ Requirements

• Latest ComfyUI (Portable or Custom Build)

• AjoNodes

• MMAudio

• Wan2.1 Native model

• A Google account to access Gemini API

🔧 Step-by-Step Installation

AjoNodes contains all the custom logic that powers The Director’s workflow.

📦 GitHub: https://github.com/AJO-reading/ComfyUI-AjoNodes

🛠️ To install:

bash

CopyEdit

cd ComfyUI/custom_nodes git clone https://github.com/AJO-reading/ComfyUI-AjoNodes

Restart ComfyUI after installation.

This is the core model used for visual generation.

🌐 Wan2.1 ComfyUI Workflow - Complete Guide | ComfyUI Wiki

Adds voice/music/sfx to your generated scenes.

📦 GitHub: kijai/ComfyUI-MMAudio

Make sure to download the models and put them in ComfyUI/models/mmaudio

Depending on the theme, you may want character-specific LoRAs or style models. Place those in:

• models/loras/

• models/embeddings/ (if using textual inversion)

• Open ComfyUI
• Load the provided TheDirectorV2.json workflow
• Paste your Gemini API Key in the designated node
• You’re ready to go!

🎉 That’s it — you’re ready to start generating movies like it’s Hollywood, minus the budget.