Hey gang, how you doing? It’s Adam Roxby here. Now if you notice something slightly strange about my voice, or should I say my delivery, then that is for a very good reason. For a number of months, I have been tinkering with the idea of cloning my voice, and this is the result. To be clear, all the words I am speaking come from the mind of Adam Roxby, but are instead spoken not by me sitting in front of a microphone, but by copying and pasting the script of this article, and having a clone of my voice reading it out for you.
I’ll firstly begin explaining the why, and then the how. I started getting adverts for a company called Eleven Labs, which is one of the world leaders in this kind of technology, and decided to try it out. Not only because they offered an app on your phone which can read articles, PDFs, webpages, and your own text in pretty convincing text-to-speech voices, but they also have a revenue stream where if other people choose your voice for their projects, then you can get a little bit of money back. As somebody who is thinking about ways to diversify his income, this appealed to me.
I’ve used text-to-speech programs for years, even before my diagnosis of dyslexia when I was studying my nursing diploma, over ten years ago. Back in those days, they were robotic, harsh, and lacking any sort of inclination, and if there were any special characters in there, then they would read them out. To the outside observer, it sounded almost like a foreign language, but to somebody who consumed hours of written content this way, my ear had become strangely accustomed and attuned to listening to this alien language.
Then as technology developed, the voices became more and more authentic, realistic, listenable, and somewhat enjoyable. It’s still a place where we find ourselves now, where not only does the technology read out text in a human-sounding voice, but can also recognize intonation, changes in tonality, where a sentence ending with an exclamation mark will sound different to ones ending with a question mark. They can whisper, they can shout, to such a point where you could easily copy the text of an open-source or copyright-free book, and it sounds as if a professional voice actor has sat in front of a microphone dictating it. Well, almost.
I also use this to read bedtime stories to my children when I can’t be there, although I do have to say I don’t always talk in such a monotone delivery, or at least I hope I don’t. The how was pretty simple because I’ve already lent my voice as part of a voluntary effort for the LibriVox organization, which takes books that are no longer under copyright and volunteers record sections for them to turn them into freely available audiobooks. I already had hours of high-quality voice recordings which I could simply upload to the website, let their technology run their magic, and enjoy the fruits of my labor.
So what am I doing this? Well, obviously I don’t create mountains of content that would mean I would be saving a significant amount of time by simply copying and pasting a script into the editor and having my AI voice read it out. I think if you are that stretched for time, then there are obviously advantages to it. I still prefer the way in which I speak as a human, particularly because my writing isn’t always fantastic, so when I read a mistake I can correct it on the fly without having to make sure that I’ve got a perfectly spelt script, as my previously mentioned dyslexia makes that more challenging.
I can see the advantages for other content creators who are perhaps more self-conscious about the sound of their own voice or perhaps are looking for a voice which they don’t possess, or even perhaps if somebody’s looking for the benefits of a co-host but finds the logistics of that to be difficult. But there is a somewhat selfish reason for me explaining this to you because if you do find the sound of my voice appealing in any way and want to have my voice reading to you for free, then you can download the Eleven Labs reader application and simply search for the name Roxby, where you can have access to my voice reading all sorts of things from newspaper articles to copy and pasted emails or whatever you desire.
You can even use it for your YouTube videos or for your podcasts. I have to say my girlfriend finds it a little bit strange that I listen to my scanned-in copies of the satirical magazine Private Eye read in my own voice, but if you are used to talking to yourself, thinking out loud, and dictating your thoughts, then I find that I’m already attuned and accustomed to the tone of my own voice, and for some reason, the information just seems to stick.
I suppose it’s when they say that if you write something out by hand, then it’s more likely to be remembered than if you just skim it over on a page. Well, if you read something out loud, a similar effect seems to happen. I’m sure one day there’ll be studies as to whether you reading it out loud is as effective as your voice reading it out loud has on your memory and recall. My guess is probably not as effective because you’re missing out the reading part of the process. But still, it’s better than nothing.
Anyway, this is AI Adam saying goodbye for now, and if you enjoy my voice, then there are links in this article on how you can have the benefit of my voice for free, without having to have me sit in front of a microphone.
