Create NotebookLM Videos With YOUR Face (No Filming - AI)

Creating Video Podcasts with Notebook LM

Introduction to Notebook LM

• Notebook LM is highlighted as a favorite AI tool for creating audio podcasts based on user-provided information.

• The limitation of Notebook LM is that it only generates audio podcasts, not video ones.

• The speaker introduces the goal of the video: demonstrating how to create video podcasts using Notebook LM.

Researching Podcast Topics

• The speaker navigates to notebookm.google and mentions having already created a podcast topic focused on "no coding tools versus vibe coding tools."

• A comprehensive overview of sources (18 in total) related to no code vs. low code platforms is generated by the tool.

Generating Podcast Content

• After gathering research, users can select different podcast formats such as deep dive, brief, critique, or debate; the speaker chooses a debate format.

• The focus of the podcast is set on whether low code or no code is better for individuals with minimal development experience.

• Once settings are configured, the content generation process begins by clicking 'generate.'

Previewing Generated Audio

• A preview of the generated audio reveals two distinct speakers debating their perspectives on app development approaches.

Preparing Audio Tracks for Video Creation

• The next step involves downloading the audio and splitting it into separate tracks for each speaker using a tool called Speaker Split.

• It’s emphasized that while one speaker talks, there should be silence in the other track to facilitate video editing later.

Processing Audio Tracks

• After processing through Speaker Split, two distinct tracks are created where each speaker's dialogue does not overlap.

Dubbing with 11 Labs

• To personalize one character's voice in the podcast, 11 Labs will be used to clone the user's voice; this requires recording short clips for voice cloning.

Voice Cloning and Dubbing Process with 11 Labs

Introduction to Voice Cloning

• The speaker introduces the voice cloning feature, noting that while it may not sound exactly like them, there is potential for improvement with more training.

• The speaker mentions the emotionality and tonality of their voice are not fully captured but can be enhanced by using the V3 model, which was temporarily unavailable.

Dubbing Process Overview

• After cloning the voice, the next step involves dubbing where a script is uploaded to replace an original voice with the cloned one.

• The project is named "VC versus VB" and set to translate from English to English, focusing on replacing a female speaker's audio.

Creating a Dubbing Project

• The speaker emphasizes clicking on "create dubbing project" to allow for edits after initial processing since first attempts may not be perfect.

• They confirm having enough credits for the process and initiate uploading, indicating it might take some time.

AI Mastermind Promotion

• A brief promotion of an AI mastermind program is presented, highlighting its resources such as courses on AI basics and automation templates.

• The speaker encourages viewers interested in AI to check out this resource as it offers daily updates and tools.

Project Processing Details

• Once uploaded, the project enters a translating phase where English audio will be altered without changing languages.

• To change voices within the project, users must select options from a menu that allows customization of how their voice sounds in playback.

Finalizing Dubbed Audio

• Users are instructed to generate audio after selecting their preferred settings; if errors occur during generation, they should wait for resolution.

• Upon successful generation, users can listen to their dubbed track before exporting it in various formats like MP3.

Transitioning to Video Creation

• After completing dubbing with 11 Labs, users are directed towards creating videos using Hen or similar platforms.

• Emphasis is placed on creating an avatar within Hen by uploading video samples similar to what was done in 11 Labs.

Creating a Digital Avatar

Steps to Create Your Digital Twin

• Begin by uploading a video of yourself, allowing 5-10 minutes for processing. This will generate avatars that resemble you.

• After avatar creation, navigate to the projects section and select "new video" followed by "create in AI studio." Adjust settings such as orientation and size as needed.

• Instead of typing a script, upload an audio file from 11 Labs. This avoids manual adjustments for pauses that may be required if using text-to-speech.

Generating Video Content

• Upload the audio file into Hey Gen; it will take time to process but will allow you to hear your voice synced with the avatar.

• Once uploaded, click on "generate" to create a video featuring your digital twin speaking with the provided audio. Expect processing times of about 5-10 minutes.

Creating Additional Avatars

• While waiting for your first video to render, you can start creating another project for a different speaker (Speaker B).

• Replace the avatar by selecting from public avatars, ensuring gender matching with Speaker B's characteristics. Choose an appropriate background setting.

Finalizing Audio and Video

• Upload Speaker B's audio file without needing transcription since it matches their voice directly.

• Generate this second video after naming it appropriately (e.g., "no code versus vibe coding") and submit it for rendering.

Editing Videos Together

• After both videos are rendered, download them and open your preferred video editing software (e.g., Cap Cut).

• Import both files into the editor; arrange them side-by-side or sequentially for comparison.

Reviewing Output Quality

• Assess each video's quality; ensure they convincingly represent each speaker’s likeness and voice.

• Basic editing techniques can enhance presentation—consider applying masks or adjusting layouts for clarity in viewer engagement.

Exploring No Coding vs. Vibe Coding

The Central Question of Trade-offs

• The debate focuses on the differences between no coding and vibe coding, emphasizing a trade-off in production methods.

• The speaker highlights the seamless transition between speakers without editing, showcasing an efficient workflow in video production.

• Suggestions for improvement include selecting better avatars and adjusting zoom levels to enhance visual appeal during presentations.

• Emphasis is placed on personalizing editing choices to align with individual preferences and project requirements.

• The discussion underscores the importance of experimenting with editing tools to achieve desired outcomes in content creation.