Augmented reality application development faces the challenge of balancing key information visualization with users’ cognitive constraints.  VR Rehab Inc. is developing an application (FAR/SV) blending augmented reality (AR) with synthetic vision for use in complex, time-sensitive military operations.  The FAR/SV approach provides several advantages over AR alone.  This paper will describe R&D investigations into effective visualizations and interactions.  The researchers investigated the effects of controller type and AR transparency level on performance.    Participants completed a simulated reconnaissance task on a laptop by viewing feed from a simulated drone flying a preset path, using the sensor camera to search the area, and placing an icon to identify the location of simulated humans in the environment.  They were also asked to recall AR building labels.  These two tasks present a competition between the synthetic vision and the semi-transparent AR overlay.  AR development efforts have failed to establish an optimal transparency level for overlaid graphics, while previous research has noted that foreground task performance diminishes at as little as 5% transparency This study compared transparency levels of 0%, 25%, 50%, and 75%.  Little attention has been paid to controller type, with current AR applications varying from speech to eye tracking to touch-screens.  Controller choice will vary depending on the platform used (e.g. iPad or HMD), and whether the user’s hands are free.  However, some controllers require less cognitive and physical effort than others.  This study compared keyboard/mouse, speech recognition/keyboard, and Xbox One controllers.  Participants’ speed and accuracy in placing markers, and usability ratings of FAR/SV, were collected to compare these conditions.  This research will fill a critical gap in guiding controller choice and AR graphic transparency level in AR applications.  Future research will investigate the effects of varying synthetic vision and AR graphic transparency independently while streaming live sensor feed in the background.