Subject: Re: Acoustic Flow Field Generation From: Eckard Blumschein <Eckard.Blumschein(at)E-TECHNIK.UNI-MAGDEBURG.DE> Date: Tue, 4 Sep 2001 10:53:33 +0200
Michael S. Gordon wrote a paper: Echolocating distance by moving and stationary listeners _____ Kevin Austin wrote: No loudspeaker is "flat", and: The speaker enclosure has added colorations _____ Peter Lennox wrote: Loudspeakers ... imitate ... fairly well ... frequency response, dynamic range etc., but none of their design criteria have anything to do with spatial perception. _____ Mikael.Fernstrom wrote: ... very different directivity. Some mics&headphones work better for music, others better for everyday sounds. Blind people ... could determine the size and quality of rooms they were in and objects in front of them quite successful. _____ Brad Libbey wrote: I have not yet found any research showing that localization of multiple echoes is a viable means of removing reverberation for complex signals. ... why in reverberant settings do recordings of loudspeakers reduce intelligibiltiy? _____ In addition, I would like to remind of Josef Manger who has been claiming that design and parameters ignore important aspects of auditory perception. In theory, he is wrong since the ear has proven phasedeaf. However, I am not sure if this dogma is actually applicable. My knowledge of the state of the art is rougly limited to how headwize articles describe Dolby surround, etc. I vaguely recall: With increasing frequency something strange happens in the shadow of the head. The sound wave travels around the head on its surface, and this way it may reach the shadowed ear. Isn't this a whispering gallery mode? Anyway, the illusion of spatial hearing includes a lot of preconditions. Just a few examples: I was told, if polarity is wrong, this does not just mean that the imagined source is located on the wrong side but it also sounds strange. What carries the contradicting information? Doesn't this phenomenon relate to the questions how our two ears distinguish between top and bottom, front and back. I remember of a record of bird song being presented in Oldenburg 2000 in order to demonstrate the reflection of a directed movable sound beam from the ceiling of a room. Everybody was able to precisely locate position and elevation of the mirror source on the ceiling. A record of human voice was "not available". I guess, we hear the precise position by means of a tonotopically diversified field of temporal contrasts in the SOC. In this case I defer to an article concerning vision which has been in discussion on Jont's list. Manger attributes the unwelcome localization of a single loudspeakers out of a stereo arrangement to perception of precedence. This effect may become audible or more audible if the onsets are exaggerated by artifacts. His Manger transducer as well as a few even more expensive electrostatic loudspeakers show a correct impulse response. As a result, one says they don't have a typical sound at all. Any coloration by reverberation, by imperfect components of sound systems, by nonlinear distortion, by antialiasing filters, by directivity, or whatever is much more impressive and sometimes attractive as well as disgusting at a time, and I agree with those who argue that large rooms have so many modes that the listener perceives the coloring effect of the room rather than many details of its shape. In other words, energy acoustics applies rather than ray acoustics. Clearly, in order to check whether a sculpture isn't a picture, one may go around. Why does reverberation reduce intelligibility? I tend to blame misleading distraction by coloration for that, just a slight pattern shift. Doesn't the room also create coloration? Of course, however normally we are well trained to recognize, interprete, and subtract these colorations. This experience may get lost, e.g. if divers are submerged in helium. I will ask Josef Manger for intelligibility reached with his transducer as to get an idea to what extent avoidance of the 'reverberant' artifacts affecting the step response could contribute to the percent correct value in case of playing back records from an environment with spatial reverberation. With all due defer to all, Eckard blumschein(at)et.uni-magdeburg.de http://iesk.et.uni-magdeburg.de/~blumsche/AuditoryPerception.html Please find there also collected responses to 'alias due to lamps' in M92 to M95.