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
These multilingual voices don't fully support certain SSML elements, such as break, emphasis, silence, and sub.
29
+
30
+
> [!IMPORTANT]
31
+
> The `en-US-JennyMultilingualV2Neural` voice is provided temporarily in public preview soley for evaluation purposes. It will be removed in the future.
32
+
>
33
+
> In order to speak in a language other than English, the current implementation of the `en-US-JennyMultilingualNeural` voice requires that you set the `<lang xml:lang>` element. We anticipate that during Q4 calendar year 2023, the `en-US-JennyMultilingualNeural` voice will be updated to speak in the language of the input text without the `<lang xml:lang>` element. This will be in parity with the `en-US-JennyMultilingualV2Neural` voice.
34
+
35
+
Introducing new features in public preview for below voices:
36
+
- Added Latin input for Serbian (Serbia) `sr-RS` voices: `sr-latn-RS-SophieNeural` and `sr-latn-RS-NicholasNeural`.
37
+
- Added English pronunciation support for Albanian (Albania) `sq-AL` voices: `sq-AL-AnilaNeural` and `sq-AL-IlirNeural`.
Copy file name to clipboardExpand all lines: articles/ai-services/speech-service/speech-synthesis-markup-structure.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -95,7 +95,7 @@ Attribute values must be enclosed by double or single quotation marks. For examp
95
95
96
96
## Speak root element
97
97
98
-
The `speak` element is the root element that's required for all SSML documents. The `speak` element contains information such as version, language, and the markup vocabulary definition.
98
+
The `speak` element contains information such as version, language, and the markup vocabulary definition. The `speak` element is the root element that's required for all SSML documents. You must specify the default language within the `speak` element, whether or not the language is adjusted elsewhere such as within the [`lang`](speech-synthesis-markup-voice.md#adjust-speaking-languages) element.
Copy file name to clipboardExpand all lines: articles/ai-services/speech-service/speech-synthesis-markup-voice.md
+17-6Lines changed: 17 additions & 6 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -226,7 +226,7 @@ This example uses a custom voice named "my-custom-voice". The custom voice speak
226
226
227
227
By default, all neural voices are fluent in their own language and English without using the `<lang xml:lang>` element. For example, if the input text in English is "I'm excited to try text to speech" and you use the `es-ES-ElviraNeural` voice, the text is spoken in English with a Spanish accent. With most neural voices, setting a specific speaking language with `<lang xml:lang>` element at the sentence or word level is currently not supported.
228
228
229
-
You can adjust the speaking language for the `en-US-JennyMultilingualNeural` neural voice at the sentence level and word level by using the `<lang xml:lang>` element. The `en-US-JennyMultilingualNeural` neural voice is multilingual in 14 languages (For example: English, Spanish, and Chinese). The supported languages are provided in a table following the `<lang>` syntax and attribute definitions.
229
+
The `<lang xml:lang>` element is primarily intended for multilingual neural voices. You can adjust the speaking language for the multilingual neural voice at the sentence level and word level. The supported languages for multilingual voices are [provided in a table](#multilingual-voices-with-the-lang-element) following the `<lang>` syntax and attribute definitions.
230
230
231
231
Usage of the `lang` element's attributes are described in the following table.
232
232
@@ -237,19 +237,30 @@ Usage of the `lang` element's attributes are described in the following table.
237
237
> [!NOTE]
238
238
> The `<lang xml:lang>` element is incompatible with the `prosody` and `break` elements. You can't adjust pause and prosody like pitch, contour, rate, or volume in this element.
239
239
240
+
### Multilingual voices with the lang element
241
+
240
242
Use this table to determine which speaking languages are supported for each neural voice. If the voice doesn't speak the language of the input text, the Speech service won't output synthesized audio.
<sup>1</sup> In order to speak in a language other than English, the current implementation of the `en-US-JennyMultilingualNeural` voice requires that you set the `<lang xml:lang>` element. We anticipate that during Q4 calendar year 2023, the `en-US-JennyMultilingualNeural` voice will be updated to speak in the language of the input text without the `<lang xml:lang>` element. This will be in parity with the `en-US-JennyMultilingualV2Neural` voice.
251
+
252
+
<sup>2</sup> The `en-US-JennyMultilingualV2Neural` voice is provided temporarily in public preview soley for evaluation purposes. It will be removed in the future.
253
+
254
+
> [!NOTE]
255
+
> Multilingual voices don't fully support certain SSML elements, such as break, emphasis, silence, and sub.
245
256
246
257
### Lang examples
247
258
248
259
The supported values for attributes of the `lang` element were [described previously](#adjust-speaking-languages).
249
260
250
-
The primary language for `en-US-JennyMultilingualNeural` is `en-US`. You must specify `en-US` as the default language within the `speak` element, whether or not the language is adjusted elsewhere.
261
+
You must specify `en-US` as the default language within the `speak` element, whether or not the language is adjusted elsewhere. In this example, the primary language for `en-US-JennyMultilingualNeural` is `en-US`.
251
262
252
-
This SSML snippet shows how to use the `lang` element (and `xml:lang` attribute) to speak `de-DE` with the `en-US-JennyMultilingualNeural` neural voice.
263
+
This SSML snippet shows how to use `<langxml:lang>` to speak `de-DE` with the `en-US-JennyMultilingualNeural` neural voice.
![prmerger-automator[bot]](https://avatars.githubusercontent.com/u/22479449?v=4&size=40){kind=avatar}
0 commit comments