feat(inputAudio): Add support for noise reduction settings

[HVbajoria]

Purpose

• Added support for noise cancellation via input_audio_noise_reduction in the session configuration payload.
• Implemented RealtimeAudioInputAudioNoiseReductionSettings with a type field that accepts:
  • "near_field" – optimized for close-talking microphones (e.g., headsets)
  • "far_field" – optimized for distant microphones (e.g., laptop, room mics)
• Integrated the setting within the SessionUpdateMessage structure to enable real-time noise reduction in transcription scenarios.
• Complements existing support for custom transcription models, prompt guidance, and language specification.

Does this introduce a breaking change?

[ ] Yes
[x] No

Pull Request Type

What kind of change does this Pull Request introduce?

[ ] Bugfix
[x] Feature
[ ] Code style update (formatting, local variables)
[ ] Refactoring (no functional changes, no api changes)
[x] Documentation content changes
[ ] Other... Please describe:

How to Test

Get the code

git clone https://github.com/Azure-Samples/aoai-realtime-audio-sdk
cd aoai-realtime-audio-sdk
git checkout Add_Noise_Reduction
npm install

Update test or dev config

In your client-side test (client_test) or runtime session code, use this updated ServerMessageType:

let configMessage: SessionUpdateMessage = {
  type: "session.update",
  session: {
    voice: "verse",
    instructions: instruction,
    input_audio_format: "pcm16",
    input_audio_transcription: {
      model: "whisper-1"
    },
    turn_detection: {
      threshold: 0.9,
      prefix_padding_ms: 500,
      silence_duration_ms: 1400,
      type: "server_vad",
      interrupt_response: true,
    },
    input_audio_noise_reduction: {
      type: "near_field"
    }
  }
};

Test the behavior

• Observe improved transcription accuracy in noisy environments
• Try both near_field and far_field settings in different mic setups
• Validate the presence of the input_audio_noise_reduction key in the outbound WebSocket payload

What to Check

• input_audio_noise_reduction is included and structured correctly in the ServerMessageType
• The type field only accepts valid values (near_field, far_field)
• Backward compatibility: sessions work as expected when noise reduction is omitted
• Noise reduction configuration is applied before the audio stream begins
• No impact to other fields like turn_detection or input_audio_transcription

Note:

Similarly do it for Python

Other Information

• This addition follows the latest Microsoft Azure OpenAI real-time audio spec:
  [Realtime Audio Reference – Noise Reduction](https://learn.microsoft.com/en-us/azure/ai-services/openai/realtime-audio-reference#inputaudionoisereduction)
• Developers can now deliver a higher-quality transcription experience in environments with background noise.

[HVbajoria]

Hi @glecaros , @jpalvarezl, @trrwilson,

I have referred this document: https://learn.microsoft.com/en-us/azure/ai-services/openai/realtime-audio-reference

Under: RealtimeAudioInputAudioNoiseReductionSettings

Could you please check once?