[Gary Eickmeier]

KH wrote:

I say that it contains NO directional information. Obviously it
contains spatial clues in the form of delayed and attenuated
information from the reverberant field. These effects clearly can be
interpreted as a sense of spaciousness. Spaciousness is an attribute
unrelated to direction, and directional information is what you need
for your model to work the way you seem to think that it works.

This is a most extraordinary statement. Spaciousness is unrelated to
direction? Was that a typo?

Let me relate an allegory that tries to show the difference between the
spatial and the temporal.

A novice goes to Best Buy and purchases a surround sound home theater in a
box. It has the 5.1 speakers in it, but he loses the instruction sheet. So
he places all of the speakers on top of the TV and the subwoofer underneath
on the floor. His left and right and center speakers are placed OK, on the
left and right side of the big TV, but his surround speakers have been
placed on top of them. He calls you up and complains that he is not
satisfied, just not hearing all of those fancy spatial effects that are
supposed to be in the movie. So you go over and behold what he has done, and
instruct him that he has gotten some of it right, with the left, center, and
right speakers, but he has not got the spatial aspect correct. His perfectly
accurate speakers are playing all of the sounds contained in the recording,
he can hear the temporal effects of the reverberation and reverb time of
the hall and discrete effects, but these spatial effects must come from
different incident angles than the direct sound in order to work.

In acoustics, many sources note that in order for the eary reflections to
work they must come from a different set of incident angles than the direct
sound, or else most of it will be masked. So you cannot simply "play" the
stereo recording and have all of the recorded ambience come fron the same
point sources as the direct sound, or it will not be heard as ambience -
just smear, really.

I have found that it is difficult for most hobbyists (at first) to
distinguish in their minds between the spatial and the temporal. The
temporal from the live venue is contained in the recording, but the spatial
effects must come from a different set of incident angles than the direct
sound. Some people incorporate side speakers with time delay for this
reason. I use reflection in a way that mimics a typical live sound field.


Throughout history most things have had models or theories generated
to explain how things worked, not the other way around.

Yes, the image model explains how it works to model the spatial aspects of
the live sound.

That said, are you really implying that no one designs speakers
against physical models of how acoustics and reproduction work?

YES! Most of them learned audio in the mono era, when the charge was just
"let's make sound." When stereo came along, they just assumed that the same
speakers could be used, just that you need two of them.


As discussed earlier, there clearly are spatial clues available in the
recording, in the form of the delayed and attenuated reverberant
field. OK so far?

Ok, so now suppose we each sit down in front of our systems. Let's
also suppose that you limit your system to the front two speakers, and
further, that you eliminate the rear-firing drivers. Now, assuming we
play the exact same (good) recording, we will be hearing basically the
same sounds (for discussion, let's stipulate similar forward radiation
patterns, and similar room interactions). Still with me?

Now, in my room, I hear the direct sound, and the room effects, and I
get a sense of spaciousness from the reverb in the recording, and the
stereo effect, and to a small degree from the room interactions. I
have a well defined soundstage, with a proper localization of
instruments and vocals.

Not so fast. I hope from the discussion above you might be able to see that
you don't just "hear" the reverb in the recording, like in the mono days. In
order to be heard properly as reverberation (early reflections, actually),
it must come from different incident angles than the direct sound.

Now, in *your* room, one of two situations obtain; you hear
*basically* what I hear, or you hear something significantly
different. From how you frequently describe box speakers, you'll
apparently hear no
soundstage, no spaciousness, a "window into another room", or a "hole
in the middle", and/or a flat presentation that is lifeless.

Alright, let's look at these two possible scenarios. In the first, we
hear *basically* the same thing. We both hear the spacial cues in the
recording in the direct sound from the speakers. We are, at this
point, both hearing ALL of the spacial information available on the
recording - we're hearing the entire unprocessed signal, within the
limits of our equipment. Nothing is hiding, nothing awaiting
"decoding", we have the whole tamale. Now, you then take all of this
information and direct it rearward to create a second, wholly
synthesized, delayed soundfield comprising all of the information -
including the spacial cues - of the recording. Every reflection
comprising this synthesized field will contain the entire signal,
delayed (to a much lesser degree, and in different ratios, than in
the venue), attenuated (including the already delayed and attenuated
information in the recording), and coming from directions different
than the original. To you, this creates realism. To me, this creates a
sense of smearing that is incompatible with my
sense of realism.

Not quite. Take a couple of examples from the real world of audio. The
Wilson WAMM vs the MBL omni. The WAMM might provide your kind of sound,
giant direct firing boxes that aim their output at your head. The MBLs, on
the other hand, are totallly different in the spatial department, being
omnis. Their sound has been described as huge, spacious, and deep, with a
sense of floating a soundstage surrounded by the ambience in the recording.

Two very different designs that sound different, because the radiaiton
pattern is very audible. Neither speaker is more "accurate" than the other,
but the MBL gets the spatial characteristic better arrayed on playback.

In the case of the second scenario, there's no use for discussion, and
any hypothesis, model, or theory you come up with won't work for me;
we simply don't interpret audio signals in the same manner.

I am sure that we do; it's just that you don't as yet understand why
different systems sound the way they do, some better than others, some
worse. I say that if ever and whenever you hear superior spaciousness and
depth in a recording, it is not due to the accuracy in the direct sound, but
rather to the different spatial characteristics of the speakers.

I have started another thread I hope you wioll fnd interesting. It
simplifies down the larger example of typical rooms and sources to a single
instrument.

Gary Eickmeier