1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
02174e2f7670: Add test for Klatt voices

Added: tests/klatt.test
Modified: .gitignore
Modified: Makefile.am

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By valdisvi:

As stress is relative, for sound it is the same. Only difference is that phonetic output is longer/more cluttered.

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By michael-conrad:

Doesn't look correct to me! It has all the syllables marked as stressed instead of non of them being marked for stress.

3 New Commits:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
9adc6e157e21: docs: fix link

Modified: docs/guide.md

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
e6af1e7c593f: docs: make direct links to guides from description to make list of docs shorter

Modified: README.md

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
30d5735db5aa: docs: fix link and make it more descriptive

Modified: docs/guide.md

[espeak-ng:master] New Comment on Issue #337 Why my espeak-ng program doesn't say anything?
By valdisvi:

I have created more detailed guides for usage and error solutions here.

2 New Commits:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
3aa4c10ef8cb: docs: remove entry about binaries, move dependencies into dependencies chapter

Modified: docs/building.md

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
b694576fd1b4: docs: add dedicated user guide for installation and usage

Added: docs/guide.md
Modified: README.md
Modified: docs/index.md

1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
df282f8f3dac: docs: add dedicated user guide for installation and usage

Added: docs/guide.md
Modified: README.md
Modified: docs/index.md

[espeak-ng/espeak-ng] Pull request updated by jaacoppi:

#795 Various code cleanups

Various code cleanups: change globals to function arguments, combine functions and remove unused code.

Most changes are with the voice output (klatt.c, wavegen.c etc). There might not be enough test cases to detect if an error has been made here.

1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
e7cc4d360964: docs: add examples how to use eSpeak NG as a C library

Added: docs/integration.md
Modified: README.md
Modified: docs/index.md

1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
aec8baf093a8: docs: add examples how to use eSpeak NG as a C library

Added: docs/integration.md
Modified: README.md
Modified: docs/index.md

[espeak-ng/espeak-ng] Pull request opened by jaacoppi:

#795 Various code cleanups

Various code cleanups: change globals to function arguments, combine functions and remove unused code.

Most changesa re with the voice output (klatt.c, wavegen.c etc). There might not be enough test cases to detect if an error has been made here.

[espeak-ng/espeak-ng] Pull request updated by jaacoppi:

#795 Various code cleanups

Various code cleanups: change globals to function arguments, combine functions and remove unused code.

Most changes are with the voice output (klatt.c, wavegen.c etc). There might not be enough test cases to detect if an error has been made here.

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By valdisvi:

Hm..., it works well for me:

espeak-ng -v chr -X "wị²dị²jị²yạ²ga."
Replace: ² > xc
Replace: ² > xc
Replace: ² > xc
Replace: ² > xc
Translate 'wịxcdịxcjịxcyạxcga'
  1	w        [w]

 36	ị      [i]

 22	xc       [2]

  1	d        [d]

 36	ị      [i]

 22	xc       [2]

  1	j        [dZ]

 36	ị      [i]

 22	xc       [2]

  1	y        [j]

 36	ạ      [0]

 22	xc       [2]

  1	g        [g]

  1	a        [0:]
 22	a (_     [0~43]

w'i2d'i2dZ'i2j'02g'0~43_

3 New Commits:

[espeak-ng:master] By Tobias Platen <tplaten@...>:
a6dbed81682e: add output hooks for wavegen

Modified: src/include/espeak-ng/espeak_ng.h
Modified: src/libespeak-ng/synthesize.c
Modified: src/libespeak-ng/synthesize.h
Modified: src/libespeak-ng/wavegen.c

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
1b73cebeb9c9: Disable secondary stress for Cherokee

Modified: src/libespeak-ng/tr_languages.c

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
04caefe54e6b: Merge pull request #791

Modified: src/include/espeak-ng/espeak_ng.h
Modified: src/libespeak-ng/synthesize.c
Modified: src/libespeak-ng/synthesize.h
Modified: src/libespeak-ng/wavegen.c

[espeak-ng/espeak-ng] Pull request closed by valdisvi:

#791 add output hooks for wavegen

I've added some hooks to wavegen, which can be used to output aligned speech. This information can be used by prosody transplantation tools similar to MBROLIGN. There is also a function to set a constant f0 in Hz, which may be used by some prosody transplantation tools. If none of the functions is called, the behaviour of espeak-ng is unchanged.

[espeak-ng:master] New Comment on Pull Request #791 add output hooks for wavegen
By valdisvi:

Thanks! This approach seems more general to do integration with other tools than #749.

[espeak-ng/espeak-ng] Pull request closed by jaacoppi:

#380 Move LOPT_DIERESES and LOPT_SONORANT_MIN from tr_languages.c to language files

Contributes to #218.

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By michael-conrad:

Meh... had to revert all the tr_* changes, it started switching to ENGLISH for my tone marks and read out the symbols. ????

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By michael-conrad:

The offending words have a short ị:

~/espeak-ng/bin/espeak-ng -v chr -X "wị²dị²jị²yạ²ga."

[espeak-ng:master] New Issue Created by george-roussos:
#794 espeak-ng: symbol lookup error: espeak-ng: undefined symbol: espeak_ng_SetVoiceByFile

Hi,

I have a compute instance on Google (Ubuntu 18.04) and trying to build this from source, because I want to add custom words and I cannot do this if I install espeak-ng from apt. I build from source following the instructions to the T, but I get the error espeak-ng: symbol lookup error: espeak-ng: undefined symbol: espeak_ng_SetVoiceByFile

It might be something very known, so I am sorry if I am a bother.

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By michael-conrad:

@michael-conrad, you need to disable secondary stress for language. I have done it with this commit, which is compiled on test site. If it is ok (all syllables are equally stressed), I'll push it to upstream project also.

I can't seem to get this working locally.

The patch is applied, but I'm still showing a sentence final stress mark in the -x output.

I also have added the 'case 0' -> abort operation to SetWordStress(...) and set stressRule 0 in the language file.

What am I missing?

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By michael-conrad:

OK, just did a test with the testing website using "U²no¹de⁴na hạ²wi²ya u²no¹de⁴na hạ²wi²ya.".

Though the final word is still getting it's final vowel lengthened, it definitely shows an improvement with not adding stress to the whole word (as evidenced by the 'h' portion of the wave form).

We, (myself + a couple of testers), have been having difficulties with differentiating between "t", "d", "g", "k", and "n". Hopefully having less stress will make these consonants more clear to us.

See screen shot. I cut the audio up into segments for easier compare. The first segment is in the middle of the sentence, the second segment is at the end of the sentence.

Right now the clause, sentence, etc intonation patterns are all set to '0 ... 0', would this be having any sort of impact?

Just as an FYI:

• Word final open vowels in modern Cherokee are short or very short or just dropped are generally nasalized by default and normally have a special end of word pitch/tone (which I can't get espeak to generate per praat analysis).

[espeak-ng:master] New Comment on Issue #785 Full stop only recognized when followed by an uppercase letter?
By jaacoppi:

https://github.com/espeak-ng/espeak-ng/blob/8aff6072e803055b8802fa670f0db9545c5849ad/src/libespeak-ng/readclause.c#L1096

Setting that to true would recognize sentences starting with lower case. I would define a new language option with #define CLAUSE_LOWERCASE_PUNCTUATION in translate.h. Any other ideas?

[espeak-ng:master] New Comment on Issue #785 Full stop only recognized when followed by an uppercase letter?
By jaacoppi:

https://github.com/espeak-ng/espeak-ng/blob/8aff6072e803055b8802fa670f0db9545c5849ad/src/libespeak-ng/readclause.c#L1094-L1096

Setting that to true would recognize sentences starting with lower case. I would define a new language option with #define CLAUSE_LOWERCASE_PUNCTUATION in translate.h. Any other ideas?

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By valdisvi:

@michael-conrad, you need to disable secondary stress for language. I have done it with this commit, which is compiled on test site. If it is ok (all syllables are equally stressed), I'll push it to upstream project also. There are more settings, which you can set only in code, e.g. for Latvian and other languages. These settings are described, but to understand, what description exactly means, you can usually through tests.

1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
8aff6072e803: Remove BOM from files

Modified: dictsource/ja_list
Modified: dictsource/ja_rules

1 New Commit:

[espeak-ng:master] By Valdis Vitolins <valdis.vitolins@...>:
881e3a5101b6: Remove BOM from file

Modified: dictsource/ja_rules

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By jaacoppi:

This diff: `diff --git a/espeak-ng-data/lang/iro/chr b/espeak-ng-data/lang/iro/chr index abe29382..9c94836c 100644 --- a/espeak-ng-data/lang/iro/chr +++ b/espeak-ng-data/lang/iro/chr @@ -17,7 +17,7 @@ words 2 1 phonemes chr

//stress on all syllables to simulate stress on no syllables -stressRule 9 +stressRule 0 stressLength 175 175 175 175 175 175 175 175 //all vowels the same length regardless of stress stressAmp 10 10 10 10 10 10 10 10 //all vowels the same strength regardless of marked stress

diff --git a/src/libespeak-ng/dictionary.c b/src/libespeak-ng/dictionary.c index bd144fa8..4f220c87 100644 --- a/src/libespeak-ng/dictionary.c +++ b/src/libespeak-ng/dictionary.c @@ -1119,6 +1119,9 @@ void SetWordStress(Translator *tr, char *output, unsigned int *dictionary_flags,

    switch (tr->langopts.stress_rule)
    {

• case 0: // no stress except what's in the language rules and dictionary
• return; + case 8: // stress on first syllable, unless it is a light syllable followed by a heavy syllable if ((syllable_weight[1] > 0) || (syllable_weight[2] == 0))`

produces: _espeak-ng -X -v chr+f2 -x "U²no¹de⁴na hạ²wi²ya hạ²wi²ya hạ²wi²ya." _ u:2no~:1de:4n0~43_ h02wi:2j0~43_ h02wi:2j0~43_ h02wi:2j'0~43_

My solution simply skips making any stress changes and returns from the function. The word "hạ²wi²ya" is handled correctly two out of three times. However, there's still a stress at the end of the sentence. The same effect happens with "hạ²wi²ya!" and "hạ²wi²ya?". It's added somewhere else. My guess is S_NO_EOC_LENGTHEN in setlengths.c. Kazakh has it implemented in tr_languages.c.

[espeak-ng:master] New Comment on Issue #793 There needs to be a way to explicitly set stress only via rules to support unstressed languages and other non-European languages.
By jaacoppi:

This diff: `diff --git a/espeak-ng-data/lang/iro/chr b/espeak-ng-data/lang/iro/chr index abe29382..9c94836c 100644 --- a/espeak-ng-data/lang/iro/chr +++ b/espeak-ng-data/lang/iro/chr @@ -17,7 +17,7 @@ words 2 1 phonemes chr

//stress on all syllables to simulate stress on no syllables -stressRule 9 +stressRule 0 stressLength 175 175 175 175 175 175 175 175 //all vowels the same length regardless of stress stressAmp 10 10 10 10 10 10 10 10 //all vowels the same strength regardless of marked stress

diff --git a/src/libespeak-ng/dictionary.c b/src/libespeak-ng/dictionary.c index bd144fa8..4f220c87 100644 --- a/src/libespeak-ng/dictionary.c +++ b/src/libespeak-ng/dictionary.c @@ -1119,6 +1119,9 @@ void SetWordStress(Translator *tr, char *output, unsigned int *dictionary_flags,

    switch (tr->langopts.stress_rule)
    {

• case 0: // no stress except what's in the language rules and dictionary
• return; + case 8: // stress on first syllable, unless it is a light syllable followed by a heavy syllable if ((syllable_weight[1] > 0) || (syllable_weight[2] == 0))`

produces: _espeak-ng -X -v chr+f2 -x "U²no¹de⁴na hạ²wi²ya hạ²wi²ya hạ²wi²ya." _ u:2no~:1de:4n0~43_ h02wi:2j0~43_ h02wi:2j0~43_ h02wi:2j'0~43_

My solution simply skips making any stress changes and returns from the function. The word "hạ²wi²ya" is handled correctly two out of three times. However, there's still a stress at the end of the sentence. The same effect happens with "hạ²wi²ya!" and "hạ²wi²ya?". It's added somewhere else. My guess is S_NO_EOC_LENGTHEN in setlengths.c. Kazakh has it implemented in tr_languages.c.

Edit: S_NO_EOC_LENGTHEN should be moved from tr_languages.c to a keyword in the language definition file lang/iro/chr. See #218. So when making a solution, don't add anything to tr_languages.c.