You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Espeak is dead and espeak-ng is its successor that we already use. It would be
interesting to see a proper comparison from the other tools that you mention
against espeak-ng from somebody who can read IPA (I cannot). The huge advantage
with espeak-ng is that it is used by thousands of blind people around the world
for their daily work, so that it has a large user base reporting pronunciation
errors.
As Espeak is unmaintained for a while and Espeak-ng is its successor, we decided
to use this for the IPA conversion. Michael's script does all the XML and
Espeak-processing work, so using your first option would only add another
abstraction layer without any benefit. From our perspective, 1) and 2) are
equivalent.
I am not sure about this. Its wording suggests that people with a more
lexicographic background are involved. However, I bet that Espeak supports more
languages and given that it's pronunciation is very accurate and its used for
everyday-use in many [screen
readers](https://en.wikipedia.org/wiki/Screen_reader), I am convinced that it is
the superior tool. More contributors, more users, more corrections.
I would vote to close this issue, if you do not have any objections.
Thanks for looking at the issue, Sebastian -- I only meant to share a couple of links that came up in a mailing list conversation as recommendations. No problem about closing this ticket.
I've just noticed this thread at Corpora-L: "IPA conversion tool" https://mailman.uib.no/public/corpora/2019-April/030032.html
People mention:
So I thought I'd make a note of that here, just in case.
The text was updated successfully, but these errors were encountered: