Hello,

for the purposes of an external Speech Analytics solution, can we export recordings in HD quality and stereo (separation of the customer's voice from that of the employee)?

have 2 separate streams, in two audio files

and the idea is to have 2 different audio files 1 agent, 1 customer

Thanks

Hello Rihab BEN MALEK, 

At this time it is not possible. I see that you posted an Idea in our enhancement request portal. I have went ahead and voted for it. As I think its an excellent Idea and a good enhancement to our base product. 

Hi Rihab BEN MALEK,

We do allow for dual channel recording to be enabled on the trunk. This will generate one file containing two separated audio streams. When listening to the recording in Genesys Cloud you can control the volume of each audio stream, 1 agent and 1 customer.

Enabling dual channel

If you want to record both sides of a call as separate channels or want to have your recordings transcribed, you must enable the Dual channel setting. However, the Dual channel setting is only available if you select one of the following audio formats:

• PCMU (G.711 u-law)
• PCMA (G.711 A-law)
• L16 (Uncompressed Linear 16-bit PCM)
• Opus

When you enable the Dual Channel setting, the system saves each channel of the recording in a separate stream. The system uses audio channel 0 to save the external participant recording and audio channel 1 to save the internal participant recording.

https://help.mypurecloud.com/articles/enable-recording/

With dual-channel recording enabled, I found that you can use the following API and you get different recording file URLs depending on whether you specify download=true or download=false:

GET /api/v2/conversations/{conversationId}/recordings/{recordingId}

With download=true, you get one recording URL which you can use to download a stereo recording of the call.

With download=false, you get two recording URLs which you can use to download recordings of each 'side' of the call.

Honestly I think it is weird that the download flag changes this behaviour. But if you have a play with it, you might find it does what you are looking for?

NB: Please be aware that the first time you make the API request it initiates the recording export, and you get no response body. So wait a few seconds and then issue the same request again and you should get the response body you are expecting.

Nick.