A major revision when compared with the original system, CEDAR Trinity 2 offers a raft of significant enhancements over its predecessor, many implemented as a direct consequence of requests from customers in the fields of law enforcement, security and counter-terrorism.
Most obvious, perhaps, is a huge expansion of the recording, listening and speech enhancement structure. In the original version, only eight audio streams could be ‘live’ at any given time and a maximum of only two could be viewed.
In CEDAR Trinity 2, the Project is now a wrapper for as many audio Assets as may be required, and up to 24 channels of audio – for example, probes in hotel or domestic rooms, microphones in interview rooms, live streams from automobiles, and so on – can be live at any given time.
Operators can view these channels simultaneously, recording as many as required, switching with a single click to any feed that seems to be displaying activity of interest. These assets are time-stamped with Universal Time Code, now with the correct application of daylight saving time, so that the timing of events can be pinned down with complete accuracy and confidence. Assets also include Events, some generated by the system and some inserted by the operator, which can be used to describe activity of interest or real-time transcription of the audio. Just as significant is the ability to import existing audio files into a project for listening, speech enhancement, and adding events.
The administration level has also been significantly enhanced, and CEDAR Trinity 2 now allows administrators to decide which operators have access to which assets so that multiple teams can work on specific assets on a single system.
To comply with the law in a number of jurisdictions, we have also added the ability to erase audio from assets, but only under strict conditions, and with the system reporting that this has been done. Of course, this feature can also be precluded for use in countries where erasure is not allowed. Likewise, we have added a switch that allows incoming audio to be monitored when no recording is taking place or (for use in jurisdictions where this is not allowed) that mutes the feed unless recording is taking place.
Other enhancements include improvements to the archiving feature, which now allows operators and administrators seamlessly to restore archived audio to its position in the original asset. This allows teams to revisit surveillance material at any time in the future. Archived audio can also be used to create new assets should this be required.
Of course, we couldn’t undertake all of this work without revisiting the speech enhancement within CEDAR Trinity 2, and this now includes a greatly improved noise reduction process based upon our flagship FNR process. This makes it possible for all users – novices and experts alike – to obtain significantly better results than ever before.
Next, exports as well as assets and archives are now cryptographically signed so that users can use CEDAR’s TVerify utility to see whether anyone has tampered with the audio, whether held in assets, in exports of specific portions of the audio (say, for evidential purposes), or in archives. If no evidence of tampering is found, the information contained with the audio can then be used with confidence.
Finally, we have improved the metadata capabilities of CEDAR Trinity 2, so that the user’s ID is added invisibly to events, and the speech enhancement parameters are embedded as metadata in exports. For security and verification purposes, these allow investigators to know with even greater precision who did what, and when.
Training on all aspects of CEDAR Trinity 2 is offered at CEDAR Audio in the UK.
CEDAR Audio Limited, 20 Home End, Fulbourn, Cambridge, CB21 5BS, United Kingdom.
t: +44 1223 881771
f: +44 1223 881778