On the Differences in Preferred Headphone Response for Spatial and Stereo Content

When reproducing spatial audio over headphones, ensuring that these have a flat frequency response is important to produce an accurate rendering. However, previous studies suggest that, when reproducing nonspatial content such as stereo music, the headphone response should resemble that of a loudspeaker system in a listening room (e.g., the so-called Harman target). It is not yet clear whether a pair of headphones calibrated in such way would be preferred by listeners for spatial audio reproduction too. This study investigates how listeners' preference regarding headphone frequency response differs in the cases of stereo and spatial audio content reproduction, rendered using individual binaural room impulse responses. Three listening tests that evaluate seven different target headphone responses, two headphones, and two reproduction bandwidths are presented with over 20 listeners per test. Results suggest that a flat headphone response is preferred when listening to spatial audio content, whereas the Harman target was preferred for stereo content. This effect was found to be stronger when user-specific equalization was used and was not significantly affected by the choice of headphone or reproduction bandwidth.