---
title: "Turn NotebookLM into Videos with YOUR Face (NO FILMING!)"
source_url: https://youtu.be/JPP82B7-U0c
captured_at: 2025-11-11
author: "Jonathan's Jam"
contributor: "ชิดชนก"
platform: youtube
---
# Turn NotebookLM into Videos with YOUR Face (NO FILMING!)

youtube video by **Jonathan's Jam**, 29m19s, uploaded 2025-10-06. URL: https://youtu.be/JPP82B7-U0c
Tags: AI podcasting, NotebookLM, AI podcast workflow, AI podcast tutorial, HeyGen, ElevenLabs, AI voice cloning, AI avatars, digital twin, AI to video, podcast automation, Speaker Split, CapCut, Audacity, agentic AI, AI content creation, AI studio setup, AI podcast video, AI video editing, synthetic media, AI storytelling, AI voiceover, podcast to video, AI podcast tools, AI production pipeline

## Shared in LINE Note+
- 2025-11-11 00:42 by ชิดชนก
- 2026-03-14 08:33 by ชิดชนก

## Description
In this tutorial, I show how to turn NotebookLM’s AI-generated podcasts into realistic talking head videos featuring your own face and voice, using the following cutting edge AI tools: SpeakerSplit audio separation, HeyGen avatars, ElevenLabs voice cloning, and CapCut video editing — all synced perfectly for natural dialogue.

🌟 Want early access to my upcoming project plus exclusive bonus content?
Help shape what I create next by filling out this quick survey:


=============

*LINKS MENTIONED IN THE VIDEO:*

*SpeakerSplit* for automatic separation of speakers:

Get 10% off with my discount code. *HUBJAM10*

*HeyGen* industry leading personal avatar creation:


*ElevenLabs* industry leading AI voice software:


*CapCut* free video editor:


*Audacity* open source audio processing software:


📝 *Show Notes* – all links + tools from the video


=============

*GIVEAWAY RULES*

No purchase necessary.
Open worldwide, ages 18+.

To enter: subscribe to this channel and comment SPEAKERSPLIT on this video.

Giveaway starts Monday 6th October and ends Wednesday 16th October.

Three winners will be randomly chosen from the comments and announced on Thursday October 17th.

Prizes: There will

## Transcript (captions:auto:en-orig)
Hello and welcome back. Notebook LM can generate human sounding conversations about any topic that are so realistic people are turning them into podcasts. It's like a podcast in the box which can churn out multiple daily episodes. The opportunity this opens up for creating engaging content is insane except for one thing. There's no video. This means no human presenters and so distribution and engagement are limited. So in this video, I'm going to show you how to take a notebook LM audio and turn it into a talking head video that looks and sounds just like you, matching the original Notebook LM speech word for word and with the exact same delivery. Let me show you an example from this Notebook LM podcast. Today we're diving into u something pretty radical. It's that pipeline made famous by Jonathan's Jam on YouTube. >> Oh yes, the one turning standard notebook LM audio into these like hyperrealistic talking ahead videos. >> Exactly. Super synchronous. >> I made this video with my face and voice as well as an accompanying host. Today we're diving into uh something pretty radical. It's that pipeline made famous by Jonathan's Jam on YouTube. Ah yes, the one turning standard notebook LM audio into these like hyperrealistic talking ahead videos. >> Exactly. Super synchronized. Looks amazing. So our mission today, let's you know unpack the actual step-by-step process, >> right? How you can put your face, maybe even your clone voice onto that AI generated content. >> Okay, let's start at the beginning. By the end of this video, you'll know how to put your own face on a Notebook LM audio overview as well as your own voice. You'll also learn how to use an avatar to do the same. Perfect if you don't want to be on camera. Shy. >> And as a bonus, I'll even show you how to change the gender of the original Notebook LM hosts. So, let's get this jam started by generating our Notebook LM podcast. So, first we come to Notebook LM. You'll find the link in the description. And we click try Notebook LM. Then create new. And we will provide the URLs of the sources we'll be using. And we'll be generating our podcast on the subject of AI automation versus AI agents. Once the sources have been processed, then we can generate the audio overview. Now, just to note, if you click these three buttons, you can set the length of the output. Default is normally about 20 minutes, but for the purposes of this tutorial, I'll set it to shorter. And now we just hit generate. And after 5 to 10 minutes, our podcast is ready. Let's have a listen. >> Welcome to our deep dive. Today, we're looking at a topic that honestly feels like it's everywhere. >> Absolutely. We've gathered a bunch of sources focusing on automation AI and uh something more specific called Agentic AI, >> right? Agentic. >> As always, that's topnotch. Now, if we click the three buttons here, we can download our audio file. Now, moving on, we can get started on the first method, adding our own face or the face of our choice to the Notebook LM podcast episode. And the first thing we want to do is to separate out the speakers in the notebook LM audio. The reason we want to do this is that we are going to generate each speaker's talking head video separately. So we will need each speaker's audio on a separate track. And it's vital that we preserve the silences, not just the speech, so that the two videos are synchronized. Luckily, there is a fantastic tool whose sole purpose is to separate speakers from a single audio track. The tool is called Speaker Split, and it's really easy to use. So, we come over to the Speaker Split website. You'll find the link in the description, and all we do is click to upload, select the number of speakers, and hit process audio. As simple as that. After a few minutes, Speaker Split delivers the two separated speaker tracks, speaker A and speaker B, as well as the full transcript it has detected. Now, we download the tracks so we can start using them to make our podcast video. Just to show you how they came out, I'll load them into Audacity, open-source audio processing software. link in the description where you can see that they've been more neatly separated than Marie Condo's sock draw seem to draw a really important line. >> They really do. And understanding that line, that spectrum is key. >> So from your reading, >> a really experienced personal driver. >> Oh, okay. So not just the map, but someone driving the car. >> Precisely. It's goal driven. >> Now, just before we go any further, Speaker Split have kindly offered to give away 60 credits each to three lucky viewers. The competition is free to join and open to everyone over the age of 18. To participate, subscribe to my channel and write speakerlit in the comments below and I will draw three names at random. The competition runs for 10 days from the date of publishing of this video. On the 11th day, the winners will be announced in a community post and I will also contact them in the comments. So keep a lookout in the comments to find out if you're one of the lucky three and good luck to everyone who participates. Okay, so now we can move on to the next step which is to create our digital twin avatar or avatars in Hen. So for that we come over to the Hen website. The link is in the description. And once you've created an account you choose create your avatar. Just before we go any further, let me explain what you can and can't do on the free plan. On the free plan, you can create one digital twin avatar of yourself. And you can even give it up to 300 different looks, i.e. the same person, but in different environments or camera angles. And you can create up to three videos per month. But they are limited to 3 minutes long. If you want to make podcast videos, which are obviously going to be longer than 3 minutes, then you'll need the creator plan, which gives you an unlimited number of videos up to 30 minutes duration. More than enough for the average notebook LM podcast. Now, let's get on with the Hen process. So, once you've clicked create avatar, Hen will want some video of you so it can generate your digital twin avatar. Agen gives you some good advice here on how to prepare and light your video. So, I suggest you take a look at that before recording yourself. Once past the advice section, we get to the upload step where you can either upload your pre-recorded video, record yourself on your webcam, or your phone. I have my video prepared, so I'm going to choose to upload footage. Then, we need to verify our identity via the webcam. You may need to give your browser permission to use the webcam and mic for recording. Record yourself speaking the text shown on screen. Once your consent has been validated, hit the next button, submit your footage, and Hen will review the quality and warn you of any possible issues. Here it says, I could have more pauses between sentences, but I'm going to submit anyway. And after a few minutes, that is ready. and it looks awesome. Hey Jonathan, your digital twin is ready. Try creating videos with it. Also, click the feedback button to share what you think. Hope you enjoy. Now for the fun part. To make my avatar speak the audio of the notebook LM mail host. And to do that, we click the create with AI Studio button and choose the landscape option. Then the H gen editor opens up with my avatar loaded onto the timeline. All we have to do is upload our separated speaker file, speaker A. Make sure voice from recording is selected and hit generate. Give it a title, then hit submit. And that's all there is to it. And after a few minutes, that is ready. So let's have a look at the result. >> Welcome to our deep dive. Today we're looking at a topic that honestly feels like it's everywhere. The future of work, business, and technology. As you can see, the crucial silences are maintained during which the other host will be speaking, >> right? Egentic AI and our mission today is >> that's great. So now we have one video ready. We need the other one for the female host. And for this one, I'm going to use an avatar from the Hen library. But obviously if you want the other host to also have the face of …