Properly handle Azure speech service voice var (#2107)

Author: pamelafox

.azdo/pipelines/azure-dev.yml

@@ -101,6 +101,7 @@ steps:
       AZURE_SPEECH_SERVICE_RESOURCE_GROUP: $(AZURE_SPEECH_SERVICE_RESOURCE_GROUP)
       AZURE_SPEECH_SERVICE_LOCATION: $(AZURE_SPEECH_SERVICE_LOCATION)
       AZURE_SPEECH_SERVICE_SKU: $(AZURE_SPEECH_SERVICE_SKU)
+      AZURE_SPEECH_SERVICE_VOICE: $(AZURE_SPEECH_SERVICE_VOICE)
       AZURE_KEY_VAULT_NAME: $(AZURE_KEY_VAULT_NAME)
       AZURE_USE_AUTHENTICATION: $(AZURE_USE_AUTHENTICATION)
       AZURE_ENFORCE_ACCESS_CONTROL: $(AZURE_ENFORCE_ACCESS_CONTROL)

.github/workflows/azure-dev.yml

@@ -87,6 +87,7 @@ jobs:
       AZURE_SPEECH_SERVICE_RESOURCE_GROUP: ${{ vars.AZURE_SPEECH_RESOURCE_GROUP }}
       AZURE_SPEECH_SERVICE_LOCATION: ${{ vars.AZURE_SPEECH_SERVICE_LOCATION }}
       AZURE_SPEECH_SERVICE_SKU: ${{ vars.AZURE_SPEECH_SERVICE_SKU }}
+      AZURE_SPEECH_SERVICE_VOICE: ${{ vars.AZURE_SPEECH_SERVICE_VOICE }}
       AZURE_KEY_VAULT_NAME: ${{ vars.AZURE_KEY_VAULT_NAME }}
       AZURE_USE_AUTHENTICATION: ${{ vars.AZURE_USE_AUTHENTICATION }}
       AZURE_ENFORCE_ACCESS_CONTROL: ${{ vars.AZURE_ENFORCE_ACCESS_CONTROL }}

app/backend/app.py

@@ -444,7 +444,7 @@ async def setup_clients():
 
     AZURE_SPEECH_SERVICE_ID = os.getenv("AZURE_SPEECH_SERVICE_ID")
     AZURE_SPEECH_SERVICE_LOCATION = os.getenv("AZURE_SPEECH_SERVICE_LOCATION")
-    AZURE_SPEECH_VOICE = os.getenv("AZURE_SPEECH_VOICE", "en-US-AndrewMultilingualNeural")
+    AZURE_SPEECH_SERVICE_VOICE = os.getenv("AZURE_SPEECH_SERVICE_VOICE") or "en-US-AndrewMultilingualNeural"
 
     USE_GPT4V = os.getenv("USE_GPT4V", "").lower() == "true"
     USE_USER_UPLOAD = os.getenv("USE_USER_UPLOAD", "").lower() == "true"
@@ -567,7 +567,7 @@ async def setup_clients():
             raise ValueError("Azure speech resource not configured correctly, missing AZURE_SPEECH_SERVICE_LOCATION")
         current_app.config[CONFIG_SPEECH_SERVICE_ID] = AZURE_SPEECH_SERVICE_ID
         current_app.config[CONFIG_SPEECH_SERVICE_LOCATION] = AZURE_SPEECH_SERVICE_LOCATION
-        current_app.config[CONFIG_SPEECH_SERVICE_VOICE] = AZURE_SPEECH_VOICE
+        current_app.config[CONFIG_SPEECH_SERVICE_VOICE] = AZURE_SPEECH_SERVICE_VOICE
         # Wait until token is needed to fetch for the first time
         current_app.config[CONFIG_SPEECH_SERVICE_TOKEN] = None
         current_app.config[CONFIG_CREDENTIAL] = azure_credential

azure.yaml

@@ -100,6 +100,7 @@ pipeline:
       - AZURE_SPEECH_SERVICE_RESOURCE_GROUP
       - AZURE_SPEECH_SERVICE_LOCATION
       - AZURE_SPEECH_SERVICE_SKU
+      - AZURE_SPEECH_SERVICE_VOICE
       - AZURE_KEY_VAULT_NAME
       - AZURE_USE_AUTHENTICATION
       - AZURE_ENFORCE_ACCESS_CONTROL

infra/main.bicep

@@ -57,6 +57,7 @@ param speechServiceResourceGroupName string = ''
 param speechServiceLocation string = ''
 param speechServiceName string = ''
 param speechServiceSkuName string // Set in main.parameters.json
+param speechServiceVoice string = ''
 param useGPT4V bool = false
 
 // https://learn.microsoft.com/azure/ai-services/openai/concepts/models?tabs=python-secure%2Cstandard%2Cstandard-chat-completions#standard-deployment-model-availability
@@ -326,6 +327,7 @@ var appEnvVariables = {
     : ''
   AZURE_SPEECH_SERVICE_ID: useSpeechOutputAzure ? speech.outputs.resourceId : ''
   AZURE_SPEECH_SERVICE_LOCATION: useSpeechOutputAzure ? speech.outputs.location : ''
+  AZURE_SPEECH_SERVICE_VOICE: useSpeechOutputAzure ? speechServiceVoice : ''
   ENABLE_LANGUAGE_PICKER: enableLanguagePicker
   USE_SPEECH_INPUT_BROWSER: useSpeechInputBrowser
   USE_SPEECH_OUTPUT_BROWSER: useSpeechOutputBrowser

infra/main.parameters.json

@@ -200,6 +200,9 @@
     "speechServiceLocation": {
       "value": "${AZURE_SPEECH_SERVICE_LOCATION}"
     },
+    "speechServiceVoice": {
+      "value": "${AZURE_SPEECH_SERVICE_VOICE}"
+    },
     "useAuthentication": {
       "value": "${AZURE_USE_AUTHENTICATION=false}"
     },