Devlog: fiddling with text-to-speech AI tools

The following blog post carries the personal thoughts of Elder Prince, founder of Blossomsoft Games.

Hey, folks!

I’ve been tinkering around with a bunch of AI tools lately, trying to figure out which ones could be used in the long run.

First and foremost, I’m really excited about the idea of making my blog more dynamic by incorporating a narrator or even multiple narrators to accompany the text. This would truly enhance the level of engagement and create a fresh, immersive experience. I’m under the impression that in today’s rapidly evolving digital landscape, readers are gradually becoming less prominent. So better try to adapt.

I’m also considering adding voice-overs to the Mandela Edition as part of a special celebration for the game’s late 10th anniversary. It’s not set in stone, but I’m exploring various technologies to make it happen.

In a previous post, I shared an audio file featuring Downey made with Amazon Polly. You can find the link in the related posts section below. I’d love to hear your thoughts on it. Does the voice sound too robotic, or does it come across as more human-like to you? Let me know your impressions.

Now, here’s the thing about text-to-speech tools: they can be quite pricey. Some of them charge a significant amount of money for only a few words.

I gave BeyondWords a shot since it integrates easily with WordPress. However, the free plan doesn’t offer much flexibility. You can use it for just one blog post, and that’s it. If you want more, you’ll need to shell out over $90 per month (yeah, per month) for a total of 1 million text characters, roughly equivalent to 100 blog posts. I am concerned about the company’s ability to thrive if they are unable to offer affordable plans that can accommodate regular bloggers. After all, regular bloggers have the potential to spread the word effectively, so it’s crucial for the company to provide them with affordable plans. $20 per month would make more sense.

Right now, I’m exploring other alternatives, but it’s taking a bit of time.

One of the options I’ve come across is pretty impressive and open source called ‘Bark’ created by Suno. It even allows you to control the emotional state of the voice. However, it might not be a complete tool just yet.

At worst, if I can’t find satisfaction with these technologies, I can always hire some real voice actors within a reasonable budget to breathe life into the game characters. While this approach may not offer the same level of flexibility for re-editing, their amazing voices can create a captivating experience.

Nothing is set in stone but I’m exploring many avenues to make the next game release more up-to-date with new trends.

That’s all for now, folks! Keep an eye out for more updates.