MRAstleMRAstle
MRAstle
Feb 13, 2023

Q: Can your AIs read IPA?

I write a lot of content with obscure words. It would be most helpful if your AIs can read the International Phonetic Alphabet (IPA). That way, I can tell them how to pronounce a word.

If this feature hasn't yet been implemented, do you have plans for it?

Thanks mates.

Founder Team
Michael_Pipio

Michael_Pipio

May 15, 2024

A: @MRAstle Ah, I'm glad you asked. So, we already (kind of) support this. If you're not quite getting the pronunciation you need from the TTS voice, you can spell it out phonetically to try and manufacture the intended pronunciation. Here's an example: https://youtu.be/8LhYooC_jkA

We do however, have it on our internal roadmap (which I can add to the public-facing one) to put together a more deliberate, user-friendly way to implement the IPA, explicitly, within your script in the speech editor.

Share
Helpful?
Log in to join the conversation

Verified purchaser

Im not sure if they can yet, but they have a roadmap you can submit and vote on requests from the community.
https://mypipio.com/pipio-roadmap/

That is a great one to add to it.

Verified purchaser

Thanks. It will be much more useful for me once this is implemented.

Verified purchaser

Thanks. I checked the link. The problem with merely spelling "phonetically" is that different accents interpret what that means differently.

IPA can help to overcome that. In particular, the option to use strict pronunciations (i.e. those shown between square brackets as opposed to between slashes) enables much greater clarity. But it should not be the only option as users may not want strict pronunciations all the time.

Another problem I've found with IPA on other sites is that bots often ignore stress markers. You may want to consider that when implementing it. While syllable breaks typically aren't noted in IPA, it can be useful to allow a dot (i.e. a period mark) to indicate a syllable break as well, which is common practice in many circles already.

Keep up the good work mates.