Those of us of a certain age will remember when the in thing was celebrities selling voice packs for GPS allowing popular characters – both fictional and real – to get you to your destination and back with a familiar cadence.
Even then we knew this was a novelty. It was, of course, the era where entire companies existed to sell ringtones and wallpapers for the increasingly popular mobile phone.
The thing is: was it that much of a novelty? People like and instinctively trust what they already know – any student who just finished the first semester in marketing will tell you that, but things go much deeper than that when we consider accents and how digital speech sounds.
Those of us in South Africa and other countries which aren’t particularly Western are used to bog standard American and British English voices when we turn on anything that is meant to speak to us. Voice assistants were the latest prime example of this, but as we get deeper into the 2020s there’s a new challenger on the block.
That challenger is text-to-speech for internet articles, newsletters and other traditional media offered online. Automatic text-to-speech using some clever software is a new Wild West of taking this legacy media and making more accessible as well as more flexible, but it’s not repeating the mistakes of the past.
SpeechKit is doing things differently this time around. No longer are we dependent on existing libraries of voices that, while speaking a familiar language, sound very alien to us.
SpeechKit uses Natural Language Programming (NLP) modules and Speech Synthesis Markup Language (SSML) to create audio that is accurate while also custom fitted to the local landscape to make the automated voice familiar and understandable.
This isn’t just pie in the sky ideas either as you can already hear it in action. South African platforms such as News24 and Daily Maverick already use SpeechKit to make their articles audible with a South African bent. You can see this in action on either site, or in this exhaustive article which covers how it’s done.
And things are looking even better going into the future. Something that has us very excited about SpeechKit is currently in beta function called Clone. This would allow you to take your own voice and use it as a base to create text-to-speech. We can see so many applications for this but the biggest must be for smaller influencers to create written content that will be read in their voice, maintaining the close social link which makes them popular.
Clone could also be used to replicate someone else that an audience would resonate with. This is great for those of us who are really good at the keyboard, but less so at the microphone.
If you’ve been looking for a flexible platform to add text-to-speech to your endeavours give SpeechKit a free try thanks to a 7-day trial that won’t cost you a thing. For those who want 24/7 support, API integration and more with a focus on enterprise features, SpeechKit will organise a demo for you and your company.