Writing / Datatalk / TTS.
Windows has a build-in TTS engine which is better than the older SAPI5 engine, but... it's a little hard to get at. And silly enough, any regular program (for example Balabolka) that calls the SAPI5 API routines isn't using the new code in Windows.
But it's actually very easy to use with a little workaround! And the quality is definitely better than, for example, Word's built-in TTS tool.
Requirements
Microsoft Edge, and (probably) Windows 10 or Windows 11.
(Let's hope Microsoft doesn't nerf this feature any time soon...)
Notes
1. It actually uses some Microsoft servers on the internet, so will eat up your bandwidth, which is important for people with limited data bundles. (You can easily test it by disconnecting from the Internet.)
2. I suspect it's actually using the Azure Neural Text-to-Speech functions. Maybe not all the voices nor the best quality, but still these are pretty damn good.
3. Also, they'd probably use anything you submit to profile you and build up their own AI models... I tried to find a license of some sort but couldn't spot anything.
4. This is NOT the same thing as the built-in Word function 'Immersive Reader'.
I wonder how long before they take it down 😒
Writers using Word
1. Save your document as .PDF (or .HTML)
2. Open Edge
3. Drag your document into Edge
4. Hit the [...] button
5. Click on 'Read Aloud'
Of course, the Immersed Reader in Edge doesn't handle all .PDF files well. Those created in Word are obviously fine, but not when coming from other sources.
Writers using Google Docs
1. Save your document as .HTML
2. Unpack the downloaded .ZIP file
3. If you use '--' in your manuscript then launch notepad, and replace all '--' with '—'
4. Open Edge
5. Drag your document into Edge
6. Hit the [...] button
7. Click on 'Read Aloud'
Automating
The process can be more or less automated, but it's tricky business. This is how it would look.
Go to this page and scroll to the bottom to see how I did it.
More
No comments:
Post a Comment