Dual Channel Transcription with Split Recording

Nexmo has introduced a new feature called split recording for its Voice API offering, allowing users to record two participants in separate audio channels, making call transcription easier. The feature enables users to work with the audio from a single participant easily. With this enhancement, users can record calls and fetch the recordings once the call is completed, enabling common tasks like call transcription. The new feature uses a specific Nexmo Call Control Object (NCCO) to enable dual-channel recording, which records the audio in both channels and sends the URL for the recording once the call is complete. Users can then transcribe the individual tracks using tools like IBM Watson or FFmpeg. By separating the audio into two mono tracks, users can accurately transcribe the conversation without having to manually merge the recordings. The feature also allows users to rebuild a timeline of the conversation as it happened, providing a more accurate representation of the call. Overall, Nexmo's split recording feature makes call transcription easier and more efficient for developers and businesses.