Azure AI Speech
An Azure service that integrates speech processing into apps and services.
1,057 questions
InitialSilenceTimeout error using SDK but works fine in short audio REST API
![](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(121.6, 69%, 21%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ECC%3C/text%3E%3C/svg%3E)
Caesar Cavales
25
Reputation points
Hello,
I have been having issues using the SDK with an audio returning with a InitialSilenceTimeout error. I have the setting below(as I saw it being discussed in a different thread)
speechConfigHandler.setProperty(PropertyId.SpeechServiceConnection_InitialSilenceTimeoutMs, '45000');
but I'm still getting the same error detail. I used the same audio in our REST API and it was transcribing correctly.
Please advise.
Thanks!
{count} votes
-
romungi-MSFT 38,481 Reputation points Microsoft Employee2023-09-20T10:09:11.49+00:00 Caesar Cavales I think the maximum value that can be set is only 15 seconds. Please see this note from the documentation.
The end of a single utterance is determined by listening for silence at the end or until a maximum of 15 seconds of audio is processed.
Also, the property is to be set on the
speechConfiginstead of its handler as mentioned below:speechConfig.SetProperty(PropertyId.SpeechServiceConnection_InitialSilenceTimeoutMs, "10000"); -
Caesar Cavales 25 Reputation points
2023-09-20T15:39:13.28+00:00 @romungi-MSFT Gotcha. I actually found that value from a comment you had in this link.
Sorry, the name might be confusing. the "handler" is just how the variable is named in
const speechConfigHandler = SpeechConfig.fromEndpoint(url, subscriptionKey);Then I did this after the line above
speechConfigHandler.setProperty(PropertyId.SpeechServiceConnection_InitialSilenceTimeoutMs, '10000');I tried changing the value to "5000", "10000". And it still failed on "InitialSilenceTimeout" error. This is the audio file I used. The same audio I used in REST API and it was able to transcribe correctly
-
romungi-MSFT 38,481 Reputation points Microsoft Employee
2023-09-21T12:14:50.77+00:00 @Caesar Cavales The following should work. I have noticed you are using single quotes for the value. Not sure, how this is being handled by the SDK.
speechConfig.SetProperty(PropertyId.SpeechServiceConnection_InitialSilenceTimeoutMs, "10000");If it still fails, I suggest raising an issue on the SDK repo with your code sample.
-
Caesar Cavales 25 Reputation points
2023-09-21T12:38:53.8066667+00:00 @romungi-MSFT I made it double quotes to test and still resulted with the same error message. I will be raising an issue in the SDK. Please confirm if this is the correct link. https://github.com/Azure/azure-sdk-for-js Thanks!
-
romungi-MSFT 38,481 Reputation points Microsoft Employee
2023-09-21T12:48:49.0233333+00:00 @Caesar Cavales If you are using JS SDK then the repo is https://github.com/microsoft/cognitive-services-speech-sdk-js/tree/master
Also, check this sample from tests in the repo.
-
Caesar Cavales 25 Reputation points
2023-09-21T12:54:23.2366667+00:00 @romungi-MSFT it doesn't seem like I can create an issue. The "New issue" button is disabled for me.
-
romungi-MSFT 38,481 Reputation points Microsoft Employee
2023-09-21T13:51:55.9666667+00:00 @Caesar Cavales It should be enabled, it is a public repo. Could you try using any other user id?
-
Caesar Cavales 25 Reputation points
2023-09-26T12:30:02.5966667+00:00 Hi @romungi-MSFT , I have reached out to support for this issue. Thanks!
Sign in to comment