Create AI Voiceovers and Dubs with 11Labs (Full Transcript)

[00:00:00] Speaker 1: Creating professional voiceovers no longer requires expensive microphones or a recording studio. That voice wasn't recorded by a person. It was generated entirely with AI using 11 Labs. I'm Elizabeth, and today I'll show you how to use 11 Labs to create realistic AI voiceovers, clone your own voice, dub videos into other languages, and build long narrations. The first thing you'll want to do is head to 11labs.io and create an account. The link to sign up is in the video description and pinned comment below. If you're just getting started, the free plan is a great place to begin, and it's what we'll be using today. It gives you enough usage to experiment with many of the platform's most popular features before deciding whether you need one of the paid plans. Once you sign in, you'll arrive at the dashboard. Now, if this seems overwhelming, don't worry, because in a few minutes everything here will make sense. Along the left-hand side, you'll see the different tools available. Depending on when you're watching this video, you may see slightly different options, since 11 Labs is constantly adding new features, but the core workflow stays the same. We'll get started today by selecting Text to Speech. Everything here is designed around a simple idea. You type text into the box, choose a voice, then 11 Labs generates an incredibly realistic narration. So let's see it in action. I'll start by pasting in a short script, but note that you can include up to 5,000 characters. And that's all I need to do before I generate my speech.

[00:01:44] Speaker 2: Nestled along the rugged coastline, the small fishing village has welcomed travelers for generations. Colorful boats sway gently.

[00:01:53] Speaker 1: That's pretty impressive, but probably not the right tone of voice for this narration. So let's dig into the voices in more detail. One of the things that immediately impressed me about 11 Labs is just how many different voices it offers. And you can browse voices based off of language, accent, gender, or even the type of content that you're creating. And each voice offers a preview so you can hear it in more detail.

[00:02:24] Speaker 3: I totally understand how you feel. Would it work if you tried talking with them again?

[00:02:29] Speaker 4: The clock ticked steadily, marking the time until a revelation that would change everything.

[00:02:36] Speaker 1: You can also search for specific types of voices.

[00:02:40] Speaker 5: And a fine morrow to you there, Captain Marshall tis I. Finding a voice that matches your project usually only takes a few minutes.

[00:02:51] Speaker 1: After narrowing down my options, I've selected Ricky Johnson, an older man with a southern accent.

[00:02:58] Speaker 6: Nestled along the rugged coastline, the small fishing village has welcomed travelers for generations. Colorful boats sway gently.

[00:03:05] Speaker 1: If you've used older text-to-speech tools before, you probably noticed that they often sounded robotic or unnatural. 11 Labs does an excellent job adding realistic pacing and emotion, making it much harder to distinguish from a real person. But let's make this sound even better. Below the voice section, you'll notice several settings that control how the AI speaks. The first is speed, exactly what it sounds like. Move the cursor to the left and the speech will get slower. Move it to the right and the narration will speed up. Next, you'll see options like stability, similarity, style exaggeration, and depending on the AI model you're using, expressiveness. These control how consistently the AI delivers your script. For example, higher stability creates a more predictable voice. Lower stability allows for more variation and emotion. There's no single best setting here. If you're narrating a corporate training video, you might want a steady, consistent delivery. If you're telling a story or recording an audiobook like I am, adding more variation often sounds much more natural. I usually recommend experimenting with these settings until you find a style that you like.

[00:04:32] Speaker 6: Nestled along the rugged coastline, the small fishing village has welcomed travelers for generations. Colorful boats sway gently in the harbor as locals prepare for another day.

[00:04:41] Speaker 1: And one tip that makes an even bigger difference than any of these settings is your writing. AI voices perform best when your script sounds like something a real person would actually say. For example, use commas where you would naturally pause. And don't be afraid to add punctuation, like question marks or exclamation points. Sometimes simply changing the punctuation can completely change how their narration sounds. Now that we've created our first AI voice, let's take things a step further with voice cloning, which is the ability to create an AI version of your own voice and is included in the starter plan. To get started, click on Voices. I'm going to use Instant Voice Clone, which is the fastest way to create your own AI voice. The process is surprisingly simple. All you need to do is upload a short recording of yourself speaking. Once you've uploaded your recording, you'll name your voice, agree that you have permission to use the recording, and then click on Save. After a few moments, your new AI voice is ready to use. Now let's head back over to text-to-speech. This time, instead of selecting one of the built-in voices, I'll choose my newly created model. I'm going to use the same script as before so that we can compare the results. I'll select this, and then click on Generate. Nestled along the rugged coastline, the small fishing village has welcomed travelers for generations. It's pretty remarkable how closely that resembles my voice. Remember, no AI voice clone is going to be perfect, but this one is pretty close. And you can also make adjustments to your cloned voice model similar to what we did before, including the speed, stability, similarity, as well as the style exaggeration. Now that you know how to generate speech and even clone your own voice, let's look at how to use 11Lab Studio for larger projects. Click on Studio in the left-hand navigation pane. Instead of creating individual clips, Studio allows you to build an entire project in one place. To get started, click on New Blank Project. Or you can upload a script. Click on Studio in the left-hand navigation pane. Instead of creating individual clips, we're going to use Studio to build an entire project in one place. You can either upload a document, like a script, or you can write or paste by clicking on New Blank Project. Today, we're going to do an audio-only project. I've pasted in a longer script. You'll notice that 11Labs automatically separated the text into different sections, which makes editing a lot easier. For example, if one line doesn't sound quite right, you don't have to regenerate the entire project. You can simply regenerate that one section until you're happy with the result. But my favorite feature of the Studio space is adding different voices for different characters in my script. For example, I am starting off my script with a narrator. So here, I'm going to highlight which lines I want the narrator to say, and I'll select the voice I want to use. You can use your clone voice that you previously created, explore the default voices that 11Labs offers, or explore the library similar to how we did before. I'll select Bella as a narrator, and you'll see that Bella's icon has been added next to those lines to help me keep track. Then I'll be able to go through my script, line by line, and pick the different voice character for each one of those narrations. And to help keep it easier for you, each one of your used voices will be at the top of the screen.

[00:09:03] Speaker 7: Do you think Grandma made her famous cinnamon rolls?

[00:09:07] Speaker 8: Knowing Grandma, I'd say there's about a 99% chance.

[00:09:11] Speaker 7: Only 99?

[00:09:14] Speaker 8: Well, there's always the chance Grandpa ate them first.

[00:09:18] Speaker 1: And once everything sounds the way that you want, you can simply export it, and 11Labs combines everything into one finished audio file that's ready to use in your video editor, presentation, podcast, or wherever you need it. As you can see, 11Labs offers a ton of different tools to play with. While we're not going to walk through all of them, I want to quickly show you a few more that are available. And the first one of those is VoiceChanger. Instead of typing text, VoiceChanger lets you upload an existing recording and transform it into another voice. Here's the original audio recording.

[00:10:01] Speaker 9: Good morning, and thank you for taking the time to meet with us today. Today I'd like to show you how we're helping organizations solve real business challenges.

[00:10:09] Speaker 1: And now, here is the voice clone that I just created. Good morning, and thank you for taking the time to meet with us today. Today I'd like to show you how we're helping organizations solve real business challenges. Save valuable time. I think this one is pretty incredible. And of course, you don't have to go with your clone voice. You can explore all of the voices in the 11Labs library. The next feature is called Dubbing, and this is great because it can dramatically expand your audience. Let's say you've created a YouTube video in English. Instead of recording the entire video again in another language, 11Labs can translate and dub your video while preserving much of the original pacing and emotion. I actually think David sounds great speaking in French. As 11Labs continues to evolve, they're regularly adding new AI power tools. So it's worth checking back from time to time to see what's new. Thanks for watching. Let me know in the comments what you're using 11Labs for.

[00:11:17] Speaker 3: See you in the next video.