In hearing research, it has been shown that the realism of the test conditions has an influence on speech intelligibility thresholds, the subject's behavior, learning abilities and listening effort. This talk presents two multimodal experiment setups that aim to bridge the differences between the laboratory and reality, focusing on the development and perceptual evaluation of hearing devices. The first laboratory uses 360° video and higher order ambisonics stimuli, a head-mounted display, and a VR perceptual evaluation interface to asses the effect of hearing devices on noise intrusiveness, speech quality, and naturalness. The second laboratory is a large loudspeaker array inside of which subjects movements and behavior can be tracked and videos can be projected on walls. This talk will introduce several challenges encountered in the audio reproduction in these two laboratories and will also present a broad overview of solutions used and developed to ensure the validity of the experimental conditions.