Recording - October 2013 - 51

Peter D’Antonio tests the Harmonix-K diffusor in RPG’s new reverberation chamber.

So choosing a set of room ratios based on spreading out the
axial modes evenly and saying “That’s as good as we’re going
to get,” that’s not the whole story?
It’s such a small part of the whole story... if I read another article about
room modes, I think I’m going to explode, because these calculations are
simplified to the point where they’re almost meaningless in the real world!
For one thing, they assume there’s no particle velocity at the boundaries. For that to be true, your walls, floor and ceiling would have to
be perfectly reflective. You’re assuming they don’t flex a tiny bit and
absorb energy when sound hits them—this may be okay for concrete
or stone, but not for drywall! They also assume a perfectly rectangular
room, and even a small change from that shape changes everything.
And most important of all, they’re only valid when all of the room
modes are being excited and all of the room modes are being listened
to. In real-world terms, that means these calculations are done with the
speaker in one corner and the listener in the opposite corner—and that
is your worst possible situation for doing that calculation, because neither your speaker nor your ear is ever going to be in the corner!
So unless your “room” happens to be a speaker and mic built
into the corners of a concrete box, the math simply doesn’t
translate into the real world.
Yes! When we first recognized this, our original programs had
the room dimensions, but they also had the coordinates for positioning the speaker and the listener. That’s the only way to do it. As
you vary the room size, it’s correlated with the two positions. You
have to work with them all simultaneously to get meaningful results.
The minute you put the speaker away from the corner and the

minute you move the microphone or listener away from the corner, then you have no handle (unless you do it systematically) on
what modes are excited and what modes are being heard. One
of the main points of understanding is: the speaker position determines which room modes are excited, and the listener position
determines which modes are heard.
We developed software to accomplish this whole process all at once,
but what we discovered was that people had a tendency to just build
their rooms with no attention to what kind of speakers they were getting,
where they were going to put them, or where they were going to sit.
And so it became apparent that we should offer the Room Sizer first,
which is the worst-case scenario where all of the modes are excited and
all of the modes are audible, and then follow up with Room Optimizer
to handle speaker and listener placement.
Using Room Optimizer we predicted the position of the loudspeakers in Bob Ludwig’s studio up in Maine when it was new; we actually put in a concrete plinth, so we had to have it exactly right. He
moved the speakers around himself and later told us that where we
told him to put them is where he’d have personally chosen to put them.
That was a nice reinforcement! [laughs]
As you move your speaker out of the corner, now you have to balance making the modal response good with minimizing the speaker boundary interference, because the speaker boundary interference introduces a notch, and you can’t equalize a notch. Once you
get that notch, you’re screwed. Audiophiles have been criticized for
years for moving a speaker a silly millimeter here or there. What
they’re hearing is this notch, which is typically in the lower frequencies, and it’s as important as the modal response.
RECORDING October 2013

Recording - October 2013

Table of Contents for the Digital Edition of Recording - October 2013