Gary Eickmeier wrote:
BEAR wrote:

Gary Eickmeier wrote:


Possibly good, but I would have to see the speaker designs. Easiest
way to visualize the result of driver positioning is to make an image
model drawing. This is just a plan view of the speakers and their
reflected images in the walls. You can then sort of see the strengths
of the various real and virtual speakers contributions to the model,
based on their radiation patterns.



Ummm... that was like 1973...

In 1975-6 I did this: http://www.bearlabs.com/NEXT/CVSR.html

Got rid of the reflecting variable... ;_)


Bear,

That would be very bad sound, placing any speaker all the way into the
corner. Yes, you got "rid" of the reflecting variable, but that is just
the problem. Image Model Theory says that the reproduction is a MODEL of
the real thing, not a portal into another acoustic. By placing the
speakers out into the room, you are modeling the deep, wide, huge
original soundstage and sound patterns. By placing speakers in the
corners, ALL of the recorded sound is made to come from just two ****
holes in the snow. The space that is in the recording is compressed and
flattened. Can you not hear those effects???

Gary Eickmeier


Gary,

You seem to be having some difficulty with this! That picture was shot
in 1976. I repeat 1976. THREE decades ago. Archival, not modern. The
speaker was NOT in the corner, fwiw, the room was smallish and the
speaker was turned and moved to make a 'more pleasing photograph.' Not
that it matters.

Furthermore, these are not reflecting speakers. So, your comments about
the placement in the room - other than room mode effects and room
surface reflections - makes no sense as far as the imaging. But since in
the real world there are interactions with the room surfaces, I would
agree that a larger room, and with speakers away from surfaces is
usually superior in terms of listening perception.

Are you still using Bose 901s today? Just curious.

I don't know what this "Image Model Theory" is. But, since stereo is a
rather poor sample (best case two point samples) of the original
acoustic environment it's difficult for me to understand what exactly
one would be precisely 'modeling' per se...

I'd prefer to not get into a contentious discussion about this topic
( "**** holes in the snow" - are those point sources?) at this time.

_-_-bear