Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

The pronunciation of Mandarin Chinese using ESpeak NJ in NVDA is not normal #1028

Closed
cary-rowen opened this issue Oct 20, 2021 · 10 comments
Closed

Comments

@cary-rowen
Copy link

When reading a Chinese character aloud, in addition to reporting the Chinese character itself, the tones of the Chinese character will also be read out in English numbers. This is abnormal.
This looks similar to this Issue
I have created the same issue in NVDA's Repo:
nvaccess/nvda#12952

Steps to reproduce:

  1. Use NVDA 2021.2 or the latest Alpha version to read the following text, The Chinese meaning of 'this text is(Today's weather is good.':
  • 今天的天气很好
  1. Use NVDA2021.1 to read the Same sample text:
  • "今天的天气很好".

Actual behavior:

In step (1)
NVDA reports ("今One 天 One 的 Five 天 One 气 Four 很 Three 好 Three").
The following is the incorrect recording demonstration (from NVDA-2021.2):
https://drive.google.com/file/d/1wBNLbUzQpPMx-5zLfXYC-XbEg0d6_1fo/view?usp=sharing

Expected behavior:

NVDA reports sample text (" - 今天的天气很好") in Mandarin, just like NVDA-2021.1.
The following is the correct recording demonstration (from NVDA-2021.1):
https://drive.google.com/file/d/1EdaSVpDq8kLXCaAEqZmFcaHRpMwfiJOU/view?usp=sharing

@hozosch
Copy link
Contributor

hozosch commented Oct 20, 2021

This is because you use the voice which reads latin as English, and I think the Chinese characters are converted to Pinjin in ESpeak for it to process, if I'm not mistaken. If it reads anything as English, this is probably why.

@valdisvi
Copy link
Member

valdisvi commented Oct 20, 2021

This is behavior change introduced in commit 4825905.
The old functionality of assuming Latin characters are pinyin can be achieved with new language cmn-Latn-pinyin
and is also described in CHANGELOG.md.

@cary-rowen
Copy link
Author

cary-rowen commented Oct 21, 2021

I don't know the code details of the ESpeak NJ, I am just a user.
I created the same issue in NVDA REPO.What do nvaccess needs to do.
nvaccess/nvda#12952

@hozosch
Copy link
Contributor

hozosch commented Oct 21, 2021

There are two Chinese mandarin voices in NVDA, "Latin as English " and "latin as pinjin". With the latter, the processing works properly. If you still want to have latin read as English but have correct processing of Chinese Characters, then I don't know, but this looks more like an ESpeak issue rather than an NVDA issue, and once this does get fixed, NVAccess needs to update ESpeak.

@cary-rowen
Copy link
Author

Is there the latest news about this issue? Can we fix this error?

@jaacoppi
Copy link
Collaborator

jaacoppi commented Dec 8, 2021

I can't repdoruce this, but it's probably connected to #1044.

@jaacoppi
Copy link
Collaborator

Sorry, my bad. I can reproduce this. The discussion is in the other thread.

@cary-rowen
Copy link
Author

At present, Chinese users are still plagued by this problem. Is there any progress on this issue?

@jaacoppi
Copy link
Collaborator

jaacoppi commented Feb 9, 2022 via email

@cary-rowen
Copy link
Author

Hi,

Please default ESpeak NG's voice role to "Chinese (Mandarin, latin as Pinyin)" for Chinese to fix this issue.
I've created the issue in the NVDA Repo, but it seems like it should be reported to ESpeak.
Please see #13572

Thanks

jaacoppi added a commit that referenced this issue Aug 31, 2022
cmn (Mandarin chinese) has been broken since 4825905.

This fix makes mandarin behave more like Cantonese. Instead of
translating characters, we search for dictionary matches.

The functionality of normal vs Chao tones should be investigated more.
Looks like latin characters as pinyin still uses Chao tones whereas
the characters in cmn_list and cmn_listx do not.

See #1044 for discussion. See also #1028 and #1163.
@jaacoppi jaacoppi closed this as completed Sep 8, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants