Speech to text - Diarization Batch API does not work

Question

Hi,

I am using STT API 3.0 (endpoint : https://southcentralus.api.cognitive.microsoft.com/speechtotext/v3.0/transcriptions)

I am using the API Batch Transcription API since I am working with audio files.
I am then retrieving the JSON results and more specifically the property "display" from "combinedRecognizedPhrases".

I am using audio files which contain interviews.
I set the property diarizationEnabled to true to get the distinction between speakers but nothing seems to work and I do not see anything which allows me to understand who is speaking.

Does it work with WAV file with 2 channels?
Do I need to do something specific ?