Change resources used in Live AI tests

[rhurey]

Drop the placeholder KV resource.
Create a Foundry Resource in East US (One region VL is enabled)
Change VL Live tests from using saved wav files to instead using the Azure Speech SDK to synthesize text.

[Copilot]

Pull Request Overview

This PR refactors the Azure AI Voice Live test infrastructure by removing placeholder Key Vault resources and enabling Voice Live testing in East US with dynamic audio generation. The changes streamline the test setup by eliminating unused infrastructure and replacing static audio files with runtime-synthesized speech.

Key Changes:

• Removed unused Key Vault resources from test infrastructure
• Changed Voice Live tests to use Azure Speech SDK for dynamic audio generation instead of static WAV files
• Added East US location support for test resources

Reviewed Changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
sdk/ai/tests.yml	Added East US location parameter for test deployment
sdk/ai/test-resources.json	Removed Key Vault resources and outputs; renamed 'name' parameter to 'baseName'
sdk/ai/Azure.AI.VoiceLive/tests/VoiceLiveTestBase.cs	Replaced audio file loading with Speech SDK-based audio generation; removed Setup method
sdk/ai/Azure.AI.VoiceLive/tests/LiveTests/BasicConversationTests.cs	Updated test calls to use text-to-speech instead of WAV file names
sdk/ai/Azure.AI.VoiceLive/tests/ContentLoggingTests.cs	Removed base.Setup() call and override keyword
sdk/ai/Azure.AI.VoiceLive/tests/ContentLoggingPerformanceTests.cs	Removed unused imports and added Ignore attribute to flaky test
sdk/ai/Azure.AI.VoiceLive/tests/Azure.AI.VoiceLive.Tests.csproj	Added Microsoft.CognitiveServices.Speech package dependency

[Copilot]

This using statement for 'System.Security.Authentication.ExtendedProtection' appears to be unused. It should be removed if there are no references to types from this namespace in the file.

Suggested change

using System.Security.Authentication.ExtendedProtection;

[Copilot]

Corrected spelling of 'synthsizer' to 'synthesizer'.

Suggested change

	using (var synthsizer = new SpeechSynthesizer(sc, ac))
	{
	var result = await synthsizer.SpeakTextAsync(text).ConfigureAwait(false);
	using (var synthesizer = new SpeechSynthesizer(sc, ac))
	{
	var result = await synthesizer.SpeakTextAsync(text).ConfigureAwait(false);

[Copilot]

The temporary file created on line 93 is never deleted, creating a resource leak. The file should be deleted after the audio synthesis is complete, preferably in a finally block or by wrapping the path creation in a try-finally pattern to ensure cleanup even if an exception occurs.

Suggested change

	var of = SpeechSynthesisOutputFormat.Riff24Khz16BitMonoPcm;
	var sc = SpeechConfig.FromEndpoint(new Uri(TestEnvironment.Endpoint), new AzureKeyCredential(TestEnvironment.ApiKey));
	sc.SetSpeechSynthesisOutputFormat(of);
	using (var ac = AudioConfig.FromWavFileOutput(path))
	using (var synthsizer = new SpeechSynthesizer(sc, ac))
	{
	var result = await synthsizer.SpeakTextAsync(text).ConfigureAwait(false);
	if (result.Reason != ResultReason.SynthesizingAudioCompleted)
	{
	throw new Exception($"Error {result.Reason} was not synthesis completed");
	}
	return result.AudioData;
	}
	try
	{
	var of = SpeechSynthesisOutputFormat.Riff24Khz16BitMonoPcm;
	var sc = SpeechConfig.FromEndpoint(new Uri(TestEnvironment.Endpoint), new AzureKeyCredential(TestEnvironment.ApiKey));
	sc.SetSpeechSynthesisOutputFormat(of);
	using (var ac = AudioConfig.FromWavFileOutput(path))
	using (var synthsizer = new SpeechSynthesizer(sc, ac))
	{
	var result = await synthsizer.SpeakTextAsync(text).ConfigureAwait(false);
	if (result.Reason != ResultReason.SynthesizingAudioCompleted)
	{
	throw new Exception($"Error {result.Reason} was not synthesis completed");
	}
	return result.AudioData;
	}
	}
	finally
	{
	try
	{
	if (File.Exists(path))
	{
	File.Delete(path);
	}
	}
	catch
	{
	// Ignore exceptions during cleanup
	}
	}

[Copilot]

The AudioConfig writes to a file on line 97, but line 106 returns result.AudioData from the synthesizer result. This makes the file output configuration and temporary file path on lines 93-97 unnecessary. Either remove the file-based AudioConfig and use a memory-based configuration, or read from the generated file instead of using result.AudioData.

[jsquire]

Would you please be so kind as to explain an override is being used here and the plan for removing it for the longer-term?