Tips on Getting Great Voice Clones from ElevenLabs

Today, I finally made a voice clone of my own voice to use on this blog.

I had actually done this process once before using the professional voice cloning option from ElevenLabs. But I didn’t end up using that version—it just didn’t sound quite right to me. 

The mistake I made was my own fault. To create a professional voice clone you need at least 30 minutes of clean audio. I discovered that without being very judicious in my selection of audio, I had uploaded audio that included unnatural pauses or sentence fragments and those aspects of my audio ended up in the voice clone as well. It turns out that to get a good result with professional voice cloning, you need really high-quality voice samples, and a lot of them. 

This time around I used the Instant Voice Cloning feature.

It still requires a good voice sample, but not as many minutes of it. Finding or createing a minimum of two minutes of clean autio is much easier to accomplish. I’d encourage you to start with this option if you’re experimenting with creating your own voice clone.

Here’s my key tip if you’re planning to try it:

Use your own writing as the source material when recording your voice sample.

If you plan to use your voice clone to narrate your own writing, this will make a big difference. That’s because the way you write and the way you speak are usually tied together. Your natural intonation, rhythm, and emphasis come through more easily when you’re reading something you wrote yourself. That makes the training sample more expressive—and that expressiveness carries through when ElevenLabs generates new audio from your future writing.

In short:

1. Choose the Instant Voice Cloning option.
2. Record yourself reading something you wrote.
3. Use that as your training sample.

You’ll likely end up with a voice clone that sounds more natural, more expressive, and more you.