The Simulated Open Field Environment (SOFE), a loudspeaker setup in an anechoic chamber to render sound sources along with their simulated, spatialized reflections, has been used for more than two decades in free-field hearing research. In 2004, the concept was revised to incorporate room-acoustic simulation software that computes sound reflections in arbitrarily-shaped rooms and auralizes them via many loudspeakers—the principle of various systems used today (Hafter and Seeber, ICA 2004). For a complete redesign of the system, an anechoic chamber has been purpose-built at TUM and I will talk about its specifications. The anechoic chamber hosts the real-time SOFE (rtSOFE), a setup with 61 loudspeakers to create a spatial sound field in a 5 m x 5 m area along with 360° of visual 3D projection. New room-acoustic simulation software for interactive computation of reflections computes room impulse responses in sub-millisecond intervals and updates a convolution system capable of convolving seconds-long impulse responses for many independent loudspeaker channels with very short latency. I will present the general concept and capabilities of the new rtSOFE, give details about its implementation and first experimental results. The rtSOFE in the new anechoic chamber at TUM forms a cutting edge research facility for interactive psychoacoustic and audio-visual research in virtual acoustic space.