Sweet area size for the envelopment of a recursive and a non-recursive diffuseness rendering approach
We compare the extent of the usable audience area of two algorithms that produce diffusely enveloping, multi-channel surround playback from a single-channel input. The FIR approach designs a set of random group-delay allpass filters to generate a set of minimally correlated playback signals. Canfield-Dafilou presented a frequency-dependent maximum group delay value as a constraint to keep audible artifacts small, in studio environments. To enlarge the audience area in which an enveloping and diffuse listening experience is achieved, we relax this constraint while having to accept an unavoidable impression of spaciousness and reverberation. Consequently, the FIR approach naturally competes with IIR feedback-delay network as alternative approach. We conduct listening experiments to reveal quality and effectiveness of both methods, in particular regarding sweet area size and sound quality.