Yesterday, we had the pleasure of hosting a webinar in collaboration with the incredible team at Bridging Voice, where we dove deep into the exciting new integration of ElevenLabs with Predictable. For those of you who couldn’t join live, this post will walk you through some of the key takeaways, including what ElevenLabs is, the variety of voice options it offers, and how you can get your own custom voice. Let's dive in!
What is ElevenLabs?
ElevenLabs is a cutting-edge AI-powered text-to-speech (TTS) platform known for its ability to generate highly realistic and emotionally rich voices. Unlike many traditional TTS solutions, which can sound mechanical or monotone, ElevenLabs uses advanced deep learning techniques to create voices that convey a natural rhythm, intonation, and expressiveness—making them perfect for assistive communication technology.
ElevenLabs’ voice synthesis stands out for its flexibility and lifelike audio quality, allowing users to communicate in ways that sound more authentic and personal. This makes it an ideal partner for applications like Predictable, which empowers people with speech difficulties to communicate using technology.
ElevenLabs Voice Sample
During the webinar, we shared a sample of the impressive voice quality generated by ElevenLabs. If you missed the live demo, press below to hear a sample of one of our team's ElevenLabs voice vs a real recording of her voice. This is a PVC voice created from 30 minutes of audio recording.
(The ElevenLabs voice is Voice 2)
What Are the Different Types of Voices Offered by ElevenLabs?
One of the most exciting features of ElevenLabs is the variety of voices you can choose from. Here’s a breakdown of the voice types:
Pre-built Voices: These are high-quality, pre-generated voices available right out of the box. They come in a range of accents, genders, and age ranges, allowing users to find something that closely matches their preference or needs.
Professional Voice Cloning: creates a highly accurate, custom voice from extensive audio samples for long-term, professional use. This requires a minimum of 30 minutes of recording.
Instant Voice Cloning: offers quick voice replication from minimal audio input, providing a natural-sounding but less precise voice for short-term or casual use. This requires a minimum of 1 minute of recording
How to Set Up the ElevenLabs Integration with Predictable
If you’re using Predictable and want to take advantage of ElevenLabs’ voices, setting up the integration is simple! Here’s a quick guide, it should take you less than 1 minute! :
How to Get Your Own Voice
The highest quality pro voices from ElevenLabs are a paid service. If you are a person with ALS/MND and would like to create an ElevenLabs voice, this is available for free through Bridging Voice and The Scott Morgan Foundation. Please contact Bridging Voice if you are in the US and The Scott Morgan Foundation if you are outside the US.
Comments