So I have done a couple of YouTube videos lately where I have had to do voiceovers and despite spending the last 20+ years talking to people for a living I’m not sure I like the sound of my own voice.
Well with all the new crazy AI stuff kicking about I thought that I would have a look at the one area that I have not looked into yet. AI text-to-speech (AI-TTS) and see if it can do my voice-over for me.
If you don’t already know most modern operating systems already have TTS built into the OS and considering they process on the fly they are not too bad.
I often use the iPhone accessibility “two finger swipe down” on my phone to read the current web page I am on when I’m busy. When you use it at often as I do the quality of the voice generation feels more than sufficient to get through a two-thousand-word News article without any issues.
So how has AI and machine learning, large language models improved TTS?
Well here is a voice-over created in almost real-time speed of my last blog post about Bing AI search:
This was made using eleven labs’ Prime Voice AI.
It’s fair to say there are quite a lot of usecases where a voice of this high-quality could be used.
Sign up for a free account to try it out the service at: elevenlabs.io