Subject: The Central Stereo Image From: "Richard J. Fabbri" <fabbri(at)NETAXIS.COM> Date: Tue, 22 Sep 1998 14:37:39 -0400Dear List, WHY is there is a Central Stereo Image ? And, why is that Central image perceived as BEHIND (not on) a line drawn between the loudspeakers? The first step is to recognize that Stereo loudspeaker playback is simply a way to make reflections and, that stereo can thus be described using the Precedence Effect. Let's discuss these Left and Right "reflections". If we playback different signals (say, two different voices) on Left and Right then, we will not perceive a Central Image. We will perceive each voice as projecting from the location of a speaker ... ... the Stereo "Soundstage" is dead - no "stereo images". ... the voices are each LOCALIZED at a speaker. Perception changes drastically if we play the SAME signal (say, ONE voice) from each speaker (Left and Right) simultaneously. We perceive a Central Image of that single voice when we stand between the speakers, i.e., at a "normal" audition location. The experiment gets more interesting if we turn towards the Left speaker and step one foot CLOSER to the Left speaker ... within practical limits, you have NOT changed your distance from the Right speaker. Effectively, you have caused the signal from the Right speaker to arrive one (1) millisecond later than the signal from the Left speaker. This is similar to driving a US car while listening to a Monaural radio station (the news). In both cases, the Right speaker now seems to have been switched-off ! The Precedence Effect explains the silencing of the Right speaker ... ... the Left and Right signals are the SAME but, the Right speaker signal now arrives 1 msec late and is perceived as a REFLECTION of the Left speaker. The Right signal is silenced and fused with the Left speaker. UK drivers experience this same problem but, with their Left speaker ... though apparently not with Parliament where a Left speaker is not silenced. Our perception of stereo was prewired via the Precedence Effect millions of years before there was stereo. We should NOT be surprised by this explanation of Stereo ... we SHOULD be surprised that something so simple has been ignored for so long. All new explanations need to be tested by application to see if NEW things can be explained. What can the Precedence Effect tell us about the Central Stereo Image ? This new picture visualizes the Left and Right speakers as reflections. So, the logical question is: "Reflections of what?" Let's assume that, when we place ourself in the "normal", centered-stereo audition position, these reflections STILL occur ... after all, if we move Left or Right 1 msec the Central Stereo Image disappears ... ... actually, due to a "reflection", we then perceive only one speaker. Again, going back several million years, what natural situation could have simulated stereo playback over loudspeakers? i.e., what natural situation can deliver the SAME sound from two (2) different directions and, AT THE SAME INSTANT ? This is a fair description of stereo playback of a Monaural signal over loudspeakers, no ? The answer is simple ... ... a Central Sound Source whose DIRECT path is acoustically blocked ;-} The Precedence Effect assumes two (2) identical signals arriving from DIFFERENT directions AND at the same instant, MUST be reflections of ... ... a missing source. Also, this "missing" source would, of necessity, be located BEHIND a line drawn between the speakers ... as is the Central Stereo image! We already have the case of the missing fundamental. We now have the case of the missing source. This is WHY there is a Central Stereo Image. Psychoacoustics is full of such examples. We are not linear systems: we are nonlinear and interactive with both our environment and our internal model of it. As an "exercise left to the student" - anyone want to explain why stereo "balance" is able to change our perception such that, even though the Left and Right signals (still) arrive simultaneously, we perceive a Stereo image that moves Left or Right as we change (volume) "balance" ??? - Richard Fabbri Email to AUDITORY should now be sent to AUDITORY(at)lists.mcgill.ca LISTSERV commands should be sent to listserv(at)lists.mcgill.ca Information is available on the WEB at http://www.mcgill.ca/cc/listserv