Perceptual comparison of two different simulation algorithms applied on the example of an open-air theatre

Open-air ancient theatres represent non-standard spaces from the point of view of room acoustics. The absence of a roof causes a remarkable decrease of the sound reflections density within the theatre space, what makes unsuited the usage of the standard room acoustic parameters, such as T30, EDT, C80 or G. In closed spaces, these parameters are generally obtained from Impulse Responses (IRs) which, besides direct sound, are made up of strong early reflections followed by long reverberation tail; in ancient theatres, only few early reflections are recognized, with following minor reflections from auditorium steps (cavea). This article is based on the hypothesis that scattered reflections play a significant role in the auditory perception, as the number of specular reflections in a semi-open space will be very limited. Thus, the research concerns the comparison of in situ measurements and simulated data from two prediction software algorithms, Odeon v13.1 and CATT-Acoustic v9. This analysis was achieved on the example of the ancient Greek theatre of Syracuse. Six different alternatives were simulated, with different scattering (s = 0.25, 0.55, 0.75) and absorption (αw = 0.05, 0.10) coefficient values applied to the cavea stone. Firstly, a comparative analysis on omnidirectional measured and simulated IRs was carried out in a parametric study. Secondly, a subjective investigation was conducted on auralized speech and music stimuli obtained from convolution of binaural measured and simulated IRs. The survey was presented to expert listeners via headphones and the results were analysed taking into consideration the comparison with objective results.