I Cloned Myself with AI… Here’s the NEW Workflow
Character cloning and advanced video avatars have evolved to offer high fidelity, providing indie filmmakers with powerful tools for rapid dialogue coverage and virtual presenters.
What this lesson covers
This tutorial outlines a modern workflow for cloning characters and generating talking head videos using HeyGen. It covers recording training data, identity protection, background removal, and uploading custom audio for lip-syncing.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 01:26 ↗
Recording clean training footage in a staged environment with natural lighting is essential for creating an accurate AI clone.
- 03:00 ↗
HeyGen incorporates identity protection features that require authentication to ensure creators maintain control over their likeness.
- 05:07 ↗
Uploading pre-recorded audio clips instead of relying solely on text-to-speech yields a much more natural and expressive clone performance.
- 06:40 ↗
Tools like voice mirroring help match the audio quality of a newly recorded track to the reference training footage.
Workflow outlined in the video
- Record clean reference video footage of the subject speaking naturally in a well-lit environment.
- Upload the footage into HeyGen and complete the identity authentication steps to create your avatar.
- Generate realistic dialogue by uploading custom audio files rather than relying entirely on text generation.
- Utilize background removal and customization options within HeyGen to place your clone into different virtual environments.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
Explore the tools
How to interpret the numbers
20,101 views captured 23 October 2026. ReelStack recommendation score: 36/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 45% performance: 0.60× views vs 7 other channel videos observed at a similar age. View-evidence factor 95% (views / (views + 1,000))
- 25% freshness: 0/100 with a 14-day half-life
- 20% momentum: 40 views/day vs 13 channel baseline (6 comparable uploads), with the same view-evidence factor
- 10% engagement: 58/100; likes + 4× comments, smoothed with a 500-view neutral prior