Sound rendering with early reflections extracted from a measured spatial room impulse response

Abstract

There are a lot of colleagues that I grateful for the support and companionship. I thank Professor Jaakko Lehtinen for nudging me towards the academy and showing what can be achieved by thinking big; Tapani Pihlajakuja for breathing new life to my PhD; and Nils Meyer-Kahlen for reaching completely new heights of invention. I also thank coauthors Jukka Pätynen for introduction to research work and Raimundo Gonzalez for intriguing journeys inside and outside office hours. I am also grateful to all the colleagues and partners that helped me along the way: Benoit

Direct and Residual Subspace Decomposition of Spatial Room Impulse Responses

“…To increase the interpretability of the results, another approach is added to the comparison that involves a temporal cut-out of the reflections. It is similarly performed in [22] and, using an omnidirectional RIR, in [57]. The approach is in the following also referred to as TempCut and comprises cutting out individual reflections via a 1 ms window that is centered around the reflection.…”

“…An extension of [17] generalizes the method using higherorder SHs and processing in angular sectors [21]. A recently proposed method [22] analyzes DOAs of reflections as in [18] but explicitly cuts out salient reflections from the omnidirectional RIR to resynthesize the early part of an SRIR. The methods render diffuse reverberation either implicitly by a fast modulation of the reproduction direction, or explicitly by using a diffuseness estimate and decorrelating diffuse signal parts to create multi-channel loudspeaker signals.…”