I tried a bunch of speech synthesis, with speed and intelligibility in mind.
ESpeakng-ng barely intelligible past ~500 words per minute, and just generally unpleasant to listen to. Maybe my brain just can't acclimatize to it.
Microsoft Zira Mobile (unlock on win11 desktop via regex) sounds much more natural and intelligible at max windows SAPI speech rate, which I estimate is around ~600 and equivalent to most conversation/casual spoken word at 2x speed. I wish windows could increase playback even further, my brain can process 900-1200 words per minute or 3x-4x normal playback speed.
On Android, Google's "United States - 1" sounds a little awkward but also intelligible at 3x-4x speed.
I tried a bunch of speech synthesis, with speed and intelligibility in mind.
ESpeakng-ng barely intelligible past ~500 words per minute, and just generally unpleasant to listen to. Maybe my brain just can't acclimatize to it.
Microsoft Zira Mobile (unlock on win11 desktop via regex) sounds much more natural and intelligible at max windows SAPI speech rate, which I estimate is around ~600 and equivalent to most conversation/casual spoken word at 2x speed. I wish windows could increase playback even further, my brain can process 900-1200 words per minute or 3x-4x normal playback speed.
On Android, Google's "United States - 1" sounds a little awkward but also intelligible at 3x-4x speed.