I’ve been receiving emails about Amazon’s Virtual Voice program. They appear to have removed the request that we DO NOT discuss the beta program, so I figure I’m okay to blog about my thoughts on it. The program also appears to be out of the beta phase. Here’s an email I received on the subject about a month ago:
Out of curiosity, I did look into it. It reminds me a lot of text-to-speech software. That’s a form of assistive technology I encounter sometimes in my other job (my non-writer one). It helps with accessibility for people with various disabilities. Admittedly, it can sound a bit robotic. I tried listening to some audio samples online from other authors who used the virtual voice program. They all sounded robotic to me. I suspect that they’re using “AI” as a buzz word for a pre-existing software, but that’s not something I’m really qualified to determine.
I have to admit that I thought about it. I’ve been thinking of making audio versions of my books for years. I haven’t because of the cost (I can’t financially justify it). The Virtual Voice program is designed to appeal to writers like me. I took some time to read about other author’s experiences with it, including [this kdp forum thread] and [this reddit thread]. The internet certainly has a variety of opinions on the subject.
Personally, I don’t like how robotic it sounds. I would also have to put some work into editing it in order to make it sound better. Additionally, I don’t like the ethical concerns in charging money for an audiobook made via a program that works similarly to text-to-speech software. That software is an important accessibility feature and I’d rather it remain available as an accessibility feature.
I’ve thought on-and-off for years about potentially recording the audiobooks myself. I’ve generally dismissed that idea due to the fact that I don’t really have experience in recording audiobooks. I also haven’t historically had the time or space needed. I actually have a little bit of experience with audio editing. I used to edit songs together into mashups in Audacity. There’s probably some recordings I made way back when still on the internet somewhere. I know how to run noise reduction. However, there’s only so much I can do when I lack a good environment to record in.
It may be worth noting that I have some theater experience from when I was growing up. I also do a lot of public speaking in my non-writer job, so I know how to modify my voice. I could probably learn how to record the audiobooks. I suspect that it would sound better than the virtual voice options would. At the very least, it would sound less robotic.
In writing this blog post, I’ve thought again about what it would take to record the audiobooks myself. I moved not that long ago to a new apartment. I don’t have a good space to use in the apartment (it’s a studio). However, the area I live in gets quiet in the evenings. The exception to that is the cicadas, but I can edit them out with noise reduction. I can also try to set up blankets and pillows like the [NPR setup] (scroll down if you click the link).
Is recoding the audiobooks myself among my better ideas? I don’t know. But it might be worth taking the time and trying.
I’ve also been thinking about ACX. That may also be worth looking into again.
In the meantime, I personally have chosen not to do the Virtual Voice program.
(By the way, my favorite mashup I ever made was Sixx A.M. “Rise” combined with Drowning Pool “Bodies.” The laptop I made it on broke, so I don’t know if I have a copy anymore.)
