This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(3.2, 63%, 10%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ESH%3C/text%3E%3C/svg%3E)
Hello,
Im using the Speech Studio>Audio Content Creation tool to produce audio files with ssml. I need to get the audio files synchronized in different languages. In an ssml document with 2 sentences, is there a way to set the starting time of the second sentence, taking as a reference the first sentence?
Example:
<speak>
<par>
<media xml:id="test" begin="0.5s">
<speak>This is the first sentence</speak>
</media>
<media xml:id="answer" begin="test.end+2.0s">
<speak>This second sentence starts 2 seconds after the begining of the first sentence.</speak>
</media>
</par>
</speak>
If this is not possible, is there another option, with neural voices?
Many thanks for help
@sam husson The tool has an option to break i.e set a time in ms to wait for the next sentence. I tried this out using the text mode with two different voices and it seems to wait until the end of first sentence before pronouncing the second one.
Thank you for your answer.
Yes this function, in ssml that would be <break time="2s" />, makes it possible to set a time to wait until the next sentence.
However that does not help for synchronizing 2 different languages, since we need to set the same time, in each language version, when the sentence is pronounced. As an example, you need 20% more time in Spanish, to say the same thing than in English (text expansion). So if we use the same break time that would not help.
On this page it says that the prosody element is used to specify changes to pitch, contour, range, rate, duration, and volume for the text-to-speech output. However i tried the duration and it does not seem to work with neural voices.
https://learn.microsoft.com/en-us/azure/cognitive-services/speech-service/speech-synthesis-markup?tabs=csharp
I was wondering if there were another way to time synchronize the same sentence, translated, in different languages, using ssml.
Many thanks!
@sam husson I am curious to understand your use case scenario of synchronizing the two audio files. Are you looking to create separate audio files and play them at once? I think this feature of synchronization is not available now directly on the tool but we can provide feedback to our team based on your use case.
Yes, the duration is only available for standard voices.
duration The period of time that should elapse while the speech synthesis (TTS) service reads the text, in seconds or milliseconds. For example, 2s or 1800ms. Duration supports standard voices only. Optional
yes, exactly, the idea would be to create 2 different audio files, in two different languages. Time synchronization is needed for synchronizing with other contents, such as music or video.
Again, thanks!