Room
Figure 2.3 shows the connections between the systems inside both the AEC and control room and the connections between the two rooms, the schematic also shows how the systems are linked and over what protocols. This diagram is
used to show how the different systems communicate and interact when they are discussed in this thesis.
Figure 2.3:Connections between the systems of the AEC and the adjacent control room. Also labeled are the types of connection between each system and in the case of devices connected to the local area network (LAN), the type of protocol used.
2.3
Motion Tracking
The new localisation method requires the measurement of localisation responses to auditory stimuli. A method of doing this is to use motion tracking technol- ogy.The motion tracking system used is the Liberty Latus tracking system [65]. This system was used in the study because it was already part of the AEC and the SOFE system.
The system works using electromagnetic motion tracking markers which are placed on the object or person wishing to be tracked. The markers track in both polar (azimuth (Az), elevation (El) and roll (Ro)) and Cartesian (X,Y,Z) co- ordinates. This type of tracking is referred to as six degrees of freedom (6DOF). These degrees are shown in Figure 2.4 in reference to a head and torso. This thesis will mainly look at the rotation of the head along the azimuth to a sound source.
Motion tracker receivers
The motion tracking system consists of eight tracking receivers which are placed at defined locations within the AEC. These are in turn connected to the video PC via the motion tracking unit inside the control room. The receivers are placed on both the ceiling and floor of the AEC to allow for maximum tracking cov- erage around the participant. The layout of the receivers placed on the ceiling and floor are shown in Figure 2.5 and Figure 2.6 respectively.
Figure 2.5:Schematic and picture of the four motion trackers placed on the ceiling of the AEC. Markers placed on floor and ceiling to increase coverage.
Figure 2.6:Schematic and picture of the four motion trackers placed on the floor and around the seat of the AEC. The markers, along with those placed on the ceiling provide dense coverage around the participant.
The motion tracker is specified to track with an accuracy better than 1◦and with a quoted spatial resolution of under 1cm (manufactures specification ([65]) and also verified by a member of the MRC Institute of Hearing Research).
Motion Tracker Markers
Along with the receivers, the motion tracker has four motion tracking markers which are used to track the response of the participant to the auditory stim- uli. Each marker weights approximately 80grams and has the dimensions, 7.4x4x2.2cm (length, width and height).
During the development of the method it was unclear exactly what sort of re- sponses would be made to the sounds presented so all four markers were used and placed on the participants head, body and arms. Placement of the markers on the participant is shown in Figure 2.7.
Figure 2.8:Picture showing the motion gear which the child wears over their own clothes. The motion trackers are attached to the motion tracking gear using industrial Velcro.
The small size and weight of the makers made it possible to fit them easily to the participants (placed upon special garments) without obstruction to their move- ment. The motion tracking markers are held in place using Velcro attached to the top of the crown and the motion tracking marker respectively. This al- lows for the motion trackers to be removed for calibration and battery changing whilst at the same time providing a reliable method of attachment to the motion tracking gear. The head markers are placed forward facing onto straps on the crown. They are attached to the top of the jacket, between the subjects shoul- der blades. The hand markers are attached by placing them into small pockets sewn onto the wrist bands. Issues did arise from certain children who did not enjoy wearing the garments to which the motion trackers were attached. This is discussed in depth later in the thesis.
Control of the Motion Tracking System
The tracking system is controlled via an independent program (developed by the author in Python [66]. The program was developed in Python as it was faster than MATLAB at reading the motion trackers serial port. The program is controlled via open sound control (OSC) messages. OSC is a protocol com- monly used for the control of music instruments, however, the language allows for the control of almost any device over a network and was used because it was already used by the visual environment as developed by Seeber et al. [64]. OSC works by sending a key word along with a message. The keyword tells the receiving device what to expect and the message contains the data which is being passed. The OSC messages are sent from MATLAB (on the audio PC) to the motion tracking controller (video PC) over UDP to a specific IP and port. The Python script interprets the OSC message and in turn controls the part of the program called by the OSC message. For example, to collect head data, the OSC message′/startcollecting′as well as an integer telling the Python program how long to collect motion tracking data for. The Python program connects to the motion tracker via a serial port. As well as controlling the motion tracker, the motion tracker controller also creates folders and files on the video PC to contain the motion tracking data on a session basis. This information is again
Calibration of the Motion Trackers
When the motion tracking system is first switched on, the motion tracking markers are not seen by the computer and they must first be initialised. Once initialised the motion trackers are calibrated. The launch and calibration of the markers is done via the motion tracking controller and controlled from the MATLAB experimental script. The calibration involves placing each marker in a set location (where the participants head is during the experiment) and then aligning them to the midline speaker until they read zero degrees. This is done because sometimes the marker will be 180◦ out. To overcome this, the marker was powered down and powered back up again. This procedure was introduced to reduce the issues regarding the motion tracking changing their polarities during a testing session or during a trial. Although correct launching and calibration of the markers reduced this happening, the issue still remained on some occasions and the data had to be post processed to eliminate these unwanted effects.