[Gary Eickmeier]

wrote:
On Saturday, June 25, 2016 at 4:38:20 AM UTC-7, Gary Eickmeier wrote:
THE PROBLEM WITH STEREO
at least one set of golden ears [Gordon Holt] said that he in a very
real sense "heard stereo" for the first time when he experimented
with extracting the l-r and r-l signals and sending them to rear
speakers with a small time delay and moderate treble roll-off, and
sending the l+r signal to another speaker in the center between the
conventional stereo pair. I have heard this improvement for myself
and heartily concur with Holt on this. the only difference is that I
find the sound field to be more cohesive ["wraparound"] without the
time delay. this also has the effect of greatly enlarging the "sweet
spot" or at least making the usual lateral collapse of sound to the
nearest speaker much less obnoxious, especially if the front stereo
pair are toed-in.

Well, obviously amen to that. But what I am saying is that stereo as most of
us apply the term today is still a half-baked concept. We have the summing
localization from left to right down pat, but that is not all there is to
it. See, first we had mono and we never expected it to make the music sound
"real" as if the musicians were right there in our room or we were right
there at the concert. Then stereo recordings and sources came around and we
just took two mono speakers and arrayed them left to right and assumed that
that was all there was to it. The implicit assumption of how it works seemed
to be you relay the sound from each microphone to your ears as accurately as
possible with no disortion or interference from the room acoustics and
voila, realism. Recording engineers go to extra lengths to construct
reflection free zones near the speakers so that they get only the pure
direct sound to their ears as the first arrival, and that is stereo.

Well, OK, first, it is true that we will never achieve the absolute realism
of transporting us to the concert hall, because of the fundamental recording
problem, that we have to run the sound through two rooms before we hear it.
In other words, we must hear our room or acoustical situation superimposed
onto the recorded sound. And no, listening anechoically doesn't work because
the sound fields do not sound natural and the imaging will not externalize
(In Head Localization, or IHL).

But we can get closer if we understand basic acoustics and that stereo on
speakers is a field-type system, not a binaural system. Once your ears are
free to hear all sounds in front of them, with no crosstalk cancellation,
then it is a field type system and we must physically reconstruct the sound
fields within our listening rooms to mimic the spatial "shape" of typical
live sound. It has not (dare I say never?) been considered before that the
reconstruction of all fields needs to be a basic part of stereo theory. What
we need to do is position the speakers in a certain way and then cast a
certain amount of direct sound toward the listening end of the room, and a
slightly greater amount of reflected sound toward the front and side walls
of the room in the same way (spatial "shape") as it happens live. You now
have a reconstruction of the live sound that is capable of decoding the
direct and reflected sounds that were recorded by means of time of arrival.
You will get the spaciousness and depth that sound more like live because of
a simple image shift toward the reflecting surfaces, giving great depth and
a soundstage from wall to wall, rather than speaker to speaker. The speakers
will disappear as sources of the sound, and you will perceive a very live
sounding Auditory Scene, or AS, in front of you. Add some surround sound to
complete the reconstruction and you have the best that can be done with
legacy stereo recordings.

The bigger the room the better because larger rooms are more the size of the
real thing, and I use specular reflectivity on the front and near side walls
to get the sharpest focus to the early reflected sound. No sound killing and
no diffusion up front. Most of these techniques are just the opposite of
what we have been assuming all these years and would never occur to audio
engineers to try.

Gary Eickmeier