The first time a client asked me to narrate a 60-page eBook, I almost said no. Renting studio time, or hiring a voice actor, would have eaten the entire budget — and the deadline was two weeks out.
Instead I built a workflow around free, browser-based text-to-speech. Here’s what I learned, including the parts nobody tells you.
The free starting point
For short projects, a free text-to-speech tool in the browser is genuinely enough. Paste the script, pick a voice, export. No account, no watermark, no “upgrade to download.” For a five-minute explainer or a sample, this is all you need.
Where the free tier ends (honestly)
Two things break down as projects grow:
- Length. Long manuscripts need chunking, consistent pacing, and a way to stitch files back together without audible seams.
- Confidentiality. Some clients don’t want their unreleased manuscript passing through a random web service. Fair.
That’s the point where I switch to an offline tool. VoiceForge runs the whole thing on my machine — the manuscript never leaves it. For NDA work, that’s not a nice-to-have, it’s the reason I get the job.
My actual process
- Clean the script: remove stage directions, expand numbers into words, break it into natural paragraphs.
- Generate a 30-second sample and send it for approval before producing anything.
- Produce in chunks, then stitch and normalize the audio levels.
- Deliver in both MP3 and WAV, because someone always asks for the other one.
The lesson
Clients don’t pay for studio time. They pay for a finished file that sounds right and arrives on time. Free tools get you 80% there; the last 20% is where you choose between “good enough” and “professional,” and that choice is usually about privacy and scale, not quality.
Start with the free browser tool, and only reach for the offline version when a project actually demands it.
Leave a Reply