Voice banking is a technology that allows an individual to record his/her own voice, and either donate the files to someone else who is non-speaking (eg: a sibling) or for use by the same individual when their voice is no longer clearly understood. This can happen with progressive conditions such as Motor Neurone Disease.
Charlie Bird is an Irish news journalist who was diagnosed with Motor Neurone Disease in 2021.
Charlie and his wife, Claire worked with Dublin based software company Marino Software to explore the possibility of creating a digital version of his voice. Charlie was fortunate to have a lot of good quality recordings for Marino Software to work with. They worked on finding the pieces of audio that best represented Charlie’s voice and cleaned them of all extra sounds (intakes of breath and so on). The cleaned recordings were then fed into a variety of different machine learning algorithms to generate an artificial Charlie Bird voice. Charlie and Claire were stunned with the results.
This voice banking process allows individuals, who are at a risk of losing their voice due to an illness such as ALS or Motor Neurone Disease, to preserve and create a digital version of their voice.
Charlie did a recent interview on The Ray D’Arcy Show on the 24th June 2022 and here you can hear Chairlie’s voice synthesizer about 18 minutes into the program
With recent advancements in Artificial Intelligence (AI) and deep learning technologies, it’s much quicker and easier to create a personalised digital voice. See our further blog posts on A step by step guide with The Voice Keeper and Voice Banking – ModelTalker
Voice Banking Services
ModelTalker – https://www.modeltalker.org
VocalID – https://vocalid.ai/
CereVoice – https://www.cereproc.com/en/products/cerevoiceme
My-own-voice (from Acapela) – https://www.acapela-group.com/voices/voice-banking/
The Voice Keeper – (UPDATE 08/05/20 – Voice Banking has become a lot easier, read the post from the people at The Voice Keeper here.)
The good: With recent advancements in Artificial Intelligence (AI) and deep learning technologies, it’s much quicker and easier to create a personalised digital voice.
The not so good: You need to record your voice before the quality deteriorates.
The verdict: This voice banking process needs to be started shortly after diagnosis.