Configuration settings that affect audio routing

Telephony Session service

Static audio forking:

• ts.APSIPGatewaySupportsForking
• ts.RTPBridge

Dynamic audio forking:

• ts.DynamicRTPMode
• ts.ForkingGatewayModel

Audio forking configures Voice Platform to receive audio during a bridged transfer. By default, audio is bridged only between the two parties.

Audio forking is required for:

• Hot word recognition on party A or C
• Recording a party’s audio during the bridge in a whole call recording
• Tone or device detection on bridged transfers
• Establishing secure incoming calls over SIP/TLS

Use static audio forking for outbound calls and systems using only one gateway model. See Static audio forking for more information.

Use dynamic audio forking for systems using multiple gateway models. See Dynamic audio forking for more information.

Voice Browser service

transferaudio attribute on <transfer>

Specifies the URL of an audio file to play while Voice Platform initiates the transfer.

The only impact on audio routing is that Voice Platform sets up a temporary RTP connection with party A to play audio.

farenddialog attribute on <transfer>

Specifies a dialog to conduct with party C before completing the transfer, for example, “Do you accept the charges?” How party C responds determines if Voice Platform connects the two parties. If the return variable is true, Voice Platform connects the call.

The only impact on audio routing is that Voice Platform sets up a temporary RTP connection with party C to conduct the dialog. The default is not used.

farendhotword attribute on <transfer>

Specifies which end of a transfer Voice Platform monitors for hot word. By default, Voice Platform monitors party A (near end). If set to true, Voice Platform monitors party C (far end).

For hot word recognition, Voice Platform, specifically the Telephony Session service, must be configured to support audio forking.