NemoClaw Knowledge Wiki

AI-Driven Podcast-to-Video Conversion - Zero Editing

3 min read

Clip title: Turn Podcasts into Videos Instantly with AI. (ZERO Editing Required!). Author / channel: Jonathan’s Jam URL: https://youtu.be/Mf9F6j_RWpI

Overview
The video provides a comprehensive tutorial on how to automatically convert AI-generated audio podcasts into video podcasts featuring talking avatars. The primary workflow involves generating an audio conversation with Google’s NotebookLM and then using a new AI tool called JoggAI to seamlessly transform that audio into a video, all without manual editing or audio separation. The creator demonstrates four distinct methods using JoggAI.

Method 1: Converting NotebookLM Audio to Video (Keeping Original Voices)
This is the core, two-step process to create a video podcast while retaining the original NotebookLM voices.

  1. Generate Audio: The creator starts in NotebookLM, creates a new notebook (themed around “Vibe Coding”), inputs a YouTube URL and a Wikipedia page as sources, and generates an “Audio Overview.” He then downloads the resulting audio file.

  2. Create Video in JoggAI: He moves to JoggAI and selects the “Video Podcast” tool. Choosing the “Upload Audio” option, he uploads the NotebookLM audio file. Once processed, he creates a custom “Host Avatar” pair (a male and a female to match the audio), selects a background theme (Mountain Lodge), and generates the video.

  • Result: The final video features the selected avatars lip-syncing perfectly to the original NotebookLM audio. The creator notes that while it handles the conversation well, very rapid, overlapping dialogue (like “mm-hmms”) can sometimes cause slight visual stuttering.

Method 2: Swapping NotebookLM Voices for Custom AI Voices
If a user wants to use different voices to make their podcast sound more unique, JoggAI offers a solution.

  1. Instead of “Upload Audio,” the user selects the “Upload Script” option in JoggAI and uploads the NotebookLM audio file.

  2. JoggAI processes the audio and extracts a written transcript.

  3. The user can then edit the script to fix errors or add/remove content.

  4. Finally, the user selects an avatar pair and actively chooses new, distinct AI voices from JoggAI’s voice library for each avatar before rendering the video.

Method 3: Creating Videos from ANY Podcast Audio
The creator emphasizes that JoggAI is not limited to NotebookLM. You can use audio from any podcast source.

  1. He demonstrates this by taking an audio episode generated by another AI tool called Jellypod.

  2. He uses the “Upload Audio” feature in JoggAI, uploads the Jellypod audio file, selects an avatar pair (without selecting new voices), and renders the video.

  3. The output successfully pairs the avatars with the original Jellypod voices.

Method 4: Generating Podcast Videos directly from Source Materials
JoggAI can also create a video podcast from scratch without needing pre-existing audio.

  1. Using the “Generate Script” option, users can input a Website URL, a YouTube video link, a PDF, or long text.

  2. The creator inputs a YouTube link to one of his own videos, selects the desired video duration (e.g., 3-5 minutes) and turn length, and JoggAI automatically generates a conversational script summarizing the source material.

  3. The user then selects an avatar pair, assigns them AI voices, and renders the video. This method creates an entirely new conversation rather than replicating the source word-for-word.

Conclusion
The creator concludes that JoggAI is a rapidly improving tool that significantly lowers the barrier to entry for creating video podcasts. He also hints at a future video exploring JoggAI’s feature that allows users to upload themselves as a custom avatar.

Graph View