Vault 7: Projects

This publication series is about specific projects related to the Vault 7 main publication.
SECRET STRAP 2 UK EYES ONLY
Whilst running, the Live Listener generates text output to provide the user with feedback as to the level of
success of the data transfer.
The first text line displays the ‘mode’ in which the Live Listener is operating ie. ‘replay mode’, ‘replay usb
files’, ‘live mode, storage on…’ etc.
If listening to the Wifi port, or reading packets from previously saved files, the application reports the
running total of the packets decoded, together with the actual packet sequence number (from data
embedded within the packet). The difference between these two numbers gives an indication of the
number of dropped packets. At the start of each Speex frame being received, the application reports the
size of the new frame being constructed, ‘new frame, size 4248’, followed by the packet sequence no.s
used to create that frame. Once the whole frame has been constructed the amount of data stored is
reported, ie. ‘buffering 4248’.
Due to the nature of the Speex decoder there is a certain amount of internal buffering within the Live
Listener which is outside the control of the Live Listener application, thus it is quite normal for 30 to 50
packets of data to be received before the audio is heard. This will equate to approx. 10 – 15 seconds of
latency.
Dropped Packets
The transport protocol used for the audio data transfer over Wifi is UDP. Unlike the TCP/IP protocol
packets are not guaranteed to reach the destination, this can result in some frames of Speex data being
incomplete. This is further compounded by the fact that it takes approx. 2 to 4 packets of data to make
up a complete Speex audio frame (depending upon the Speex quality configuration setting), so a single
dropped packet will result in a whole frame of data being discarded. The Live Listener monitors the
incoming data and can identify missing headers or headers arrived before their expected time, in these
cases the application will attempt to retrain against the new data and build a complete Speex frame.
Each Speex frame will contain about 0.25 seconds of real time audio.
8.3.4 Troubleshooting
Silence at the Listening end could be caused by the following:
a. The TV is not in range
b. There is nothing being recorded by the implant due to silence
c. The implant is incorrectly configured
d. The Live Listen tool is not running
PAGE 23 OF 31
SECRET STRAP 2 UK EYES ONLY