Remember When It was Just Text-to-Speech?

Posted

in

, ,

by

June is Audiobook Appreciation Month. With that being the case, I can think of no better topic to discuss: non-human narrated audiobooks.

Hate them? Love them? Indifferent? I can’t help but think of it all as sophisticated text-to-speech (TTS) technology. Indeed, a tool that has served a critical role in my book publishing process since 2011. I started using NaturalReader software to record my finished stories and correct errors, usually around the fourth and final draft way back then. Indeed, I continue to use NaturalReader, along with a host of other tools, including Word itself, to read my stories back to me during the final edits phase.

To know me is to know I once listened to eBooks read by my Kindle devices. Now, I rely on Alexa to read eBooks to me. But do I love non-human narrated audiobooks? Hardly. However, it would be terribly disingenuous of me to say I hate them.

In view of my longstanding and previously disclosed history with text-to-speech, when presented with the chance to produce audiobooks in collaboration with Apple when it rolled out its digital voice narration capability in early 2023, I was eager to give it a try, trusting as I did that the voice actors said to be involved in the voice generations were fairly compensated. Whether they were or not is not for me to say. What I will say is that I hated it. It was touted as a six-month minimum commitment, so minimal risk, right? I could not wait until I could end the contracts. Yes, the audiobooks sold a few copies worldwide, but I found Apple’s assigned prices outrageous for a product that was merely tolerable and over which I had no control. In some cases, prices were as high as $9.99 for novellas.

In that same spirit of embracing text-to-speech technology, I was also eager to take part in Google Play’s rollout of digitally narrated audiobooks. I found far more success there and some enjoyment as well. The process was more transparent. I had more control over the finished product, and I could set my own prices. Indeed, I own and often listen to several of them. They are no longer available for purchase, however, despite their past contribution to my bottom line.

This brings me to Amazon KDP’s rollout of virtual voice narrations. Who among us did not know that it was just a matter of time? I am a wide author with limited visibility on Amazon these days, even for my eBooks. As Amazon’s pay-to-play business model is my abhorrence, I would have no reason to suppose the visibility of a virtually narrated audiobook would be any different. I am only writing about KDP because it’s a confirmation of what we have long known was coming and now that it is here, it looks like it’s here to stay.

In other areas of text-to-speech technology, I have been playing around with various tools offered by ElevenLabs, Speechelo, and the likes. From what I have discerned, copious amounts of work are required to produce full-length audiobooks effectively. For short voice-overs, not so much.

Something I have done is experiment with using some of ElevenLabs’ synthesized voices with my own natural voice to see how it plays out. Here’s a ‘dual narration’ snippet that I put together a while back. Happy listening!

6 responses to “Remember When It was Just Text-to-Speech?”

  1. Lois Avatar
    Lois

    The TTS sounds quite good. What “tuning” capabilities were available? A few years ago when I was rendering academic material (primarily math-related) to TTS, I used some (fairly crude, it seemed to me) tools to adjust pronunciation when necessary for clarity. The voices were pretty robotic, but we were at least able to fix the worst issues. It was time-consuming, though, since you had to review the output and then tweak until it was tolerable.

    1. P. O. Dixon Avatar
      P. O. Dixon

      Indeed, TTS has come a long way. The software used for the colonel’s voice is pretty powerful, but in this case, I was just playing around–this is a first pass. Polished outcomes are far too time-consuming. Thanks, Lois!

  2. Susan Kaye Avatar
    Susan Kaye

    Good for you for striking out into, and talking about, this fraught territory. AI is here to stay and the only way to deal with it is to learn to use it to our advantage. While we still can.

    BTB, “George” sounds a bit like Stephen Fry. I have to laugh a little at the idea of Fry as Darcy’s cousin. (It’s Odd Association Thursday in my neck of the woods!)

    Thanks for sharing this.

    1. P. O. Dixon Avatar
      P. O. Dixon

      You’re welcome. Fraught territory, indeed. What a good point about learning the technology while we still can. Thanks, Susan!

  3. cindie snyder Avatar
    cindie snyder

    Nice! TTS sounds great and you don’t sound bad either!lol I can’t get audio books on my Kindle and I prefer books in hand any way but it would be nice to listen too, I just may look up your YouTube things sometime!

    1. P. O. Dixon Avatar
      P. O. Dixon

      Thanks! I try. I’m narrating my fourth audiobook, bearing in mind Lady Catherine’s sage advice: “…no excellence in music is to be acquired without constant practice.” Indeed, the same applies to narrating. 🙂 I hope you will give my YouTube channel a try.

Leave a Reply

Discover more from Always Austen

Subscribe now to keep reading and get access to the full archive.

Continue reading