Odiogo, whose text-to-speech (a.k.a. speech synthesis) service I use on this blog of mine, seems to have improved the quality of the service, that is, the quality of the synthesized speech. I tried it just now (after a while - it was reasonably good even earlier), and was surprised to find that it was almost completely intelligible / correct. I tried it out on this blog post about Disqus that I wrote earlier today, and found that, apart from two minor points, the speech synthesis seemed almost flawless. The two points were:
- it ran together words from separate sentences that were not separated by a period or full stop (such as my name at the top of the post followed by the first sentence of the post itself); but that's probably not a flaw anyway.
- it did not pronounce my name correctly, but again, that's probably because it's an Indian name and the "voice" was a Western one, so its built-in vocabulary would not include my name.
Cool service. Check it out if you are interested in text-to-speech. The service is free for personal blogs.
- Vasudev Ram - Dancing Bison Enterprises
Share this post:
No comments:
Post a Comment