I have the habit of convert everything I write into audio first before I publish it. It's a good way to make your mistakes pop out.

I use fromtexttospeech[1] to convert to audio. Judging by the voices, it seems like the author is also using the same speech engine.

If I'm not mistaken, these are from Nextup's TextAloud software.

[1]:http://www.fromtexttospeech.com

Edit: Though this beats the 50k character limit