The system permits manual obfuscation of calls using a DTMF key sequence at the user's handset. When a user enters a specified sequence of digits, the obfuscation begins, and any audio will be immediately and irreversibly overwritten with the tone specified in the section of the file. To stop obfuscation, the  sequence must be entered, whereupon the audio will be written normally to the call's stream.

 The and  key sequences can be defined in the  file located in , using the and attributes. The following  snippet is an example of a configuration to enable manual call obfuscation when a user enters

<audio>
    <obfuscation>
        <dtmfStart>**</dtmfStart>
        <dtmfStop>##</dtmfStop>
        <dtmfTimeout>2000</dtmfTimeout>
        <stopTimeout>5000</stopTimeout> 
        <toneFile>
            <default>{app}/tones/alaw.wav</default>
            <codec id="0">{app}/tones/ulaw.wav</codec>
            <codec id="8">{app}/tones/alaw.wav</codec>
        </toneFile>
    </obfuscation>
</audio>

The  attribute was provided to prevent a situation where extra digits have been erroneously entered, and therefore making it impossible to enter the correct sequence. After the specified DTMF timeout has passed, all the digits previously entered will be cleared. This also ensures that obfuscation does not occur during normal negotiation of DTMF telephone selection menus (inbound or outbound). To set a DTMF entry timeout, specify the value (in milliseconds) using the element.

Additionally, an obfuscation timeout can be specified to automatically stop the obfuscation after a specified period of time, to prevent the entire remaining call from being obfuscated in case the user forgets to manually stop the obfuscation. This timeout period can be specified (in milliseconds) using the element.