Steve_Graham@ub.cc.umich.edu (04/09/90)
Mark Warren said he believed we basically throw out phase (time) info when we
listen to speakers and get our stereo localization cues from amplitude
differences only. This is not quite true. Headphone listening probably does
make more use of timing info, but I have made several recordings which relied
primarily on phase difference stereo, and they produce (at least for my ears,
and those of an audiophile friend) quite good stereo over speakers. And even
over speakers, if only amplitude differences were recorded, I find depth
lacking. That is why I dislike (intensely--no pun intended) purist
recordings of classical music made with just a single pair of co-incident
cardioid mikes. Ideally both phase (time) and amplitude differences should
be recorded, as perhaps with the ORTF mike setup which makes use of a pair of
cardioid mikes angled at 110 degrees and spaced apart 117 cm, which is
approx. the distance between ears. The stereo system that used the Calrec
Soundfiled mike is called Ambisonic. That system also excludes timing
information. This is a choice the system's designers made quite
deliberately, because they felt the complexity would be unmanageable
otherwise. I have only heard a four-speaker demonstration of it once, but I
wasn't impressed. However conditions were far from ideal. I haven't liked
any of the ambisonic releases I've heard via two speakers, as I find them
lacking in depth (predictably).
Michael Tarr wonders about the incredibly negative tone of Stereophile's
reviews of Carver products. So-called "Sonic Holography" is spectacular,
though I'm not sure if it's very natural. I don't know about most of his
other products, but the tuner, which was claimed to make bad stereo reception
a thing of the past, is a horrible thing! When the signal gets (what would
normally be) noisy, the tuner does some wierd stuff that sounds to my ears
like the *worst* kind of "electronically reprocessed" stereo. Yuck!mwarren@mips2.cr.bull.com (Mark Warren) (04/10/90)
In article <5013382@ub.cc.umich.edu> Steve_Graham@ub.cc.umich.edu writes: >Mark Warren said he believed we basically throw out phase (time) info when we >listen to speakers and get our stereo localization cues from amplitude >differences only. ... YIKES! DID I SAY THAT??? I must have been asleep, and unfortunately I expire this news too quickly to go back and see exactly what I did say. What I meant to say was that in the context of an automobile, the size/shape of the environment causes so many frequency/amplitutde problems that careful attention to phasing is likely to reap only limited benifits to the overall sound quality. I absolutely do NOT mean that we ignore phase/time info. -- == Mark Warren Bull HN Information Systems Inc. == == (508) 671-3171 (FAX 671-3020) 300 Concord Road MS820A == == mwarren@granite.cr.bull.com Billerica, MA 01821 ==
tarr-michael@YALE.EDU (michael tarr) (04/10/90)
Without giving a psychology lecture there are essentially three systems by which humans perceive the location of sound (imaging): Timing or phase differences (best for < 1500 hz) Intensity or amplitude differences (best for > 1500 hz) Onset and offset differences in transients (good all over) Plus particularly around 8000 hz we can use the diffraction patterns of our pinnae to localize. I have seen alot of misconceptions about what is known and unknown about human audition. For a nice readable summary get Brian Moore's "Introduction to hearing" (academic press I think). It is about $10 in paperback. By the way for all the hyper hi frequency peopl -- try getting an audiologist to measure how high you can hear -- most people over 25 can't hear any higher than 15000 hz at anything near normal listening levels. As for Carver -- I still think Stereophile is harder on his products than others. And they definitely have overly friendly relationships with some manufacturers at the expense of others. tarr@cs.yale.edu