Azure Speech Recognizer gives me error about header

Question

I am trying Speech to text with Azure Speech Recognizer service but when ı run this code, ı get error

code is;

import azure.cognitiveservices.speech as speechsdk

subscription_key = "key is here"
service_region = "region is here too"

audio_file = "data file path"

speech_config = speechsdk.SpeechConfig(subscription=subscription_key, region=service_region)

audio_input = speechsdk.AudioConfig(filename=audio_file)

speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_input,language="tr")

result = speech_recognizer.recognize_once()

if result.reason == speechsdk.ResultReason.RecognizedSpeech:
    print("Metin: {}".format(result.text))
elif result.reason == speechsdk.ResultReason.NoMatch:
    print("Eşleşme bulunamadı: {}".format(result.no_match_details.reason))
elif result.reason == speechsdk.ResultReason.Canceled:
    cancellation_details = result.cancellation_details
    print("Tanıma iptal edildi: {}".format(cancellation_details.reason))
    if cancellation_details.reason == speechsdk.CancellationReason.Error:
        print("Hata ayrıntıları: {}".format(cancellation_details.reason_details))

and error is;

---> 16 speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_input,language="tr")

RuntimeError: Exception with error code: 
[CALL STACK BEGIN]

/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1aa875) [0x7fd4c79aa875]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1aaf64) [0x7fd4c79aaf64]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1ac537) [0x7fd4c79ac537]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1acbee) [0x7fd4c79acbee]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1919bc) [0x7fd4c79919bc]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x194b3e) [0x7fd4c7994b3e]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1346c5) [0x7fd4c79346c5]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x19b253) [0x7fd4c799b253]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x19b6c2) [0x7fd4c799b6c2]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x13e447) [0x7fd4c793e447]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x1e65f6) [0x7fd4c79e65f6]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x139b9b) [0x7fd4c7939b9b]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(+0x20dfe2) [0x7fd4c7a0dfe2]
/usr/local/lib/python3.10/dist-packages/azure/cognitiveservices/speech/libMicrosoft.CognitiveServices.Speech.core.so(recognizer_create_speech_recognizer_from_source_lang_config+0x116) [0x7fd4c78bf641]
/lib/x86_64-linux-gnu/libffi.so.8(+0x7e2e) [0x7fd509008e2e]
/lib/x86_64-linux-gnu/libffi.so.8(+0x4493) [0x7fd509005493]
/usr/lib/python3.10/lib-dynload/_ctypes.cpython-310-x86_64-linux-gnu.so(+0xa3e9) [0x7fd50902e3e9]
[CALL STACK END]

Exception with an error code: 0xa (SPXERR_INVALID_HEADER)

ı have azure portal account and ı got relative information from Azure Speech service

Answer 1

Based on the provided information, the error message suggests that there might be an issue with the configuration of the SpeechRecognizer object.

One possible issue might be the language code you are using in your code. To fix this, you can use tr-TR in place of tr as your language code.

speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_input,language="tr-TR")

With above change the code successfully executed.

Please refer to this documentation to get details regarding other supported languages and codes.

Answer 2

SPXERR_INVALID_HEADER is an indication that the audio file in use doesn't have a valid WAVE header. Please double check to make sure that you are pointing to a valid 16KHz, 16bits per sample, mono WAV file in your audio config.

You can find more information about supported formats, as well as how to optionally use compressed audio formats here: https://learn.microsoft.com/azure/ai-services/speech-service/how-to-use-codec-compressed-audio-input-streams?pivots=programming-language-python