Speech Cognitive services Authentication error (401) even with correct subscription key

Question

Hi
I meet a problem for building a speech to text with Azure. I followed this tutorial

https://learn.microsoft.com/en-us/azure/cognitive-services/speech-service/get-started-speech-to-text?tabs=script%2Cbrowser%2Cwindowsinstall&pivots=programming-language-python

And nothing is possible with my speech config. I set it as following :

speech_config = speechsdk.SpeechConfig(subscription="{<!-- -->{MY SUBSCRIPTION KEY}}", region="westus")  
speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config)  

I tried to re generate key many times, there is nothing to do I always get the error :

Speech Recognition canceled: CancellationReason.Error    
Error details: WebSocket upgrade failed: Authentication error (401). Please check subscription information and region name.    

I read everywhere that I certainly have a mistake in my key, or that my subscription is over, or I have reached the quota limit. But that's not the problem because I try with this method and get a success :

I followed advices of microsoft doc and tried this :

curl -v -X POST "https://YOUR_REGION.api.cognitive.microsoft.com/sts/v1.0/issueToken" -H "Ocp-Apim-Subscription-Key: YOUR_SUBSCRIPTION_KEY" -H "Content-type: application/x-www-form-urlencoded" -H "Content-Length: 0"  

And it worked, I got a token with success
Then I tried :

curl -v -X POST "https://YOUR_REGION.stt.speech.microsoft.com/speech/recognition/interactive/cognitiveservices/v1?language=en-US" -H "Authorization: Bearer YOUR_ACCESS_TOKEN" -H "Transfer-Encoding: chunked" -H "Content-type: audio/wav; codec=audio/pcm; samplerate=16000" --data-binary @YOUR_AUDIO_FILE  

And I got in text my audio_file
So that's not a problem of subscription, my key is valid.
So if anyone could help me to resolve this "Authentication Error Please check subscription" problem it would be amazing because it's making me mad since 3 days

Thank's for reading

Answer 1

Hello everyone,

This works for me now. Please check my code sample and let me know if you still have issue.

# Copyright (c) Microsoft. All rights reserved.  
# Licensed under the MIT license. See LICENSE.md file in the project root for full license information.  
  
# <code>  
import  azure.cognitiveservices.speech as speechsdk  
  
# Creates an instance of a speech config with specified subscription key and service region.  
# Replace with your own subscription key and service region (e.g., "westus").  
speech_config = speechsdk.SpeechConfig(subscription="1fac40698f86************", region="westus")  
  
# Creates a recognizer with the given settings  
speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config)  
speech_config.set_property(speechsdk.PropertyId.Speech_LogFilename, "LogfilePathAndName")  
  
print("Say something...please")  
  
  
# Starts speech recognition, and returns after a single utterance is recognized. 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.  The task returns the recognition text as result.   
# Note: Since recognize_once() returns only a single utterance, it is suitable only for single  
# shot recognition like command or query.   
# For long-running multi-utterance recognition, use start_continuous_recognition() instead.  
result = speech_recognizer.recognize_once()  
  
# Checks result.  
if result.reason == speechsdk.ResultReason.RecognizedSpeech:  
    print("Recognized: {}".format(result.text))  
elif result.reason == speechsdk.ResultReason.NoMatch:  
    print("No speech could be recognized: {}".format(result.no_match_details))  
elif result.reason == speechsdk.ResultReason.Canceled:  
    cancellation_details = result.cancellation_details  
    print("Speech Recognition canceled: {}".format(cancellation_details.reason))  
    if cancellation_details.reason == speechsdk.CancellationReason.Error:  
        print("Error details: {}".format(cancellation_details.error_details))  
# </code>  

Regards,
Yutong

Answer 2

Same happening to me for cognitive-services-speech-sdk Node.js quickstart tutorial -
https://github.com/Azure-Samples/cognitive-services-speech-sdk/tree/master/quickstart/javascript/node/from-file
I also checked token by:

curl -v -X POST "https://westeurope.api.cognitive.microsoft.com/sts/v1.0/issueToken" -H "Ocp-Apim-Subscription-Key: <my key>" -H "Content-type: application/x-www-form-urlencoded" -H "Content-Length: 0"

which properly returned 200 with new token, however the recognizeOnceAsync returns Unauthenticated 401 code:

SpeechRecognitionResult {
  privResultId: '193B48A21A994C568302EAEB95C674EE',
  privReason: 1,
  privText: undefined,
  privDuration: undefined,
  privOffset: undefined,
  privLanguage: undefined,
  privLanguageDetectionConfidence: undefined,
  privErrorDetails: 'Unable to contact server. StatusCode: 1006, undefined Reason:  Unexpected server response: 401',
  privJson: undefined,
  privProperties: PropertyCollection {
    privKeys: [ 'CancellationErrorCode' ],
    privValues: [ 'ConnectionFailure' ]
  },
  privSpeakerId: undefined
}

I am using westeurope region & key 2 listed in Azure RESOURCE MANAGEMENT > Keys and Endpoint, on MacOS Bug Sur
I checked audio file properties: 16000 bitrate & mono

@edit - using sdk.SpeechConfig.fromAuthorizationToken() instead of fromSubscription() with hard-coded bearer token retrieved from above curl works

Answer 3

By doing trial and error, I found that I got the error when I use the Free Tier (F0), but as soon as I switch to a Paid Tier (S0), then it works. It has noting to do with subscription key nor region setting.

Answer 4

Hi,

In your example, I see you have used { } around the key, what is the reason for that ?

Please do the call without those.
speech_config = speechsdk.SpeechConfig(subscription="copy_paste_your_key_here", region="westus")

And that should work ok.

Regards,
Jarno

Answer 5

@Jarno Ensio Hakulinen Could you please share your workable sample? I am using below sample but same error happens.

import azure.cognitiveservices.speech as speechsdk  
  
# Creates an instance of a speech config with specified subscription key and service region.  
# Replace with your own subscription key and service region (e.g., "westus").  
speech_config = speechsdk.SpeechConfig(subscription="a25efc907e11498cb***********", region="westus")  
  
# Creates a recognizer with the given settings  
speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config)  
  
print("Say something...")  
result = speech_recognizer.recognize_once()  
  
# Checks result.  
if result.reason == speechsdk.ResultReason.RecognizedSpeech:  
    print("Recognized: {}".format(result.text))  
elif result.reason == speechsdk.ResultReason.NoMatch:  
    print("No speech could be recognized: {}".format(result.no_match_details))  
elif result.reason == speechsdk.ResultReason.Canceled:  
    cancellation_details = result.cancellation_details  
    print("Speech Recognition canceled: {}".format(cancellation_details.reason))  
    if cancellation_details.reason == speechsdk.CancellationReason.Error:  
        print("Error details: {}".format(cancellation_details.error_details))  
# </code>