Some things to note from the above images:
- The ceiling structure (joists) is exposed. There are three types
of obstructions below the joist plane:
- the two support crossbeams
- two A/C ducts
- a bunch of sprinkler pipes
See the sketch above for details on where these obstructions are.
- Light fixtures are entirely ad hoc -- there are several flourescent
fixtures that are semi-permanently attached to the ceiling, but those
are used only during exhibit installation work. More noteworthy are
the many incandescent clamp lamps used to spotlight the art work --
these are moved
around routinely to adjust to new art installations (approximately once
every 1-2 months). For this reason, a K-13 foam treament would not work --
they need to be able to clamp the lights anywhere, and the exposed joist
grid works well for this.
- There is about 24 inches from the bottom of the joists to the top
of the A/C vents. Most of the air is directed at a downward angle,
although a portion of the air from each vent goes straight out
horizontally. This is important because we would want hanging baffles
to not be hit by HVAC air and thus obstructing flow and swinging in
the breeze.
- The vents are about 7-feet on center down the length of the duct.
Thus baffles could be hung such that they avoid the vents.
- The ceiling structure is entirely wood, above the metal crossbeams.
The joists are unpainted. FYI for K-13 adhesion considerations.
Finally, the above pictures were obviously taken when the gallery was
empty. It is "between installations" right now, with the next
installation running from Dec 1st to Jan 5th. Thus, any major work
would have to be targeted for the next break. Even if the treatment
involves just hanging some baffles, one does not want to be handling
these baffles and a 14-foot ladder around one-of-a-kind artworks.
01-Dec-2001:
I spoke with Jeff Rackley, the guy who does the sound for most/all music
performances at Eyedrum. Besides the reverb problem, he said that there
was a resonance around 250 Hz. Note that this room is not rectangular, although
it is a parallelogram (remember junior high geometry?), so the resonance
may or may not have anything to do with the room dimensions. If it does,
note that the room dimensions result in a theoretical resonance around 10 Hz.
Not 250 Hz.
Anyway, I added links to bass trap information in the acoustic treatments
summary above.
I also noted that in addition to the steel support beams and the A/C ducts,
there is sprinkler plumbing crisscrossing the ceiling. Somehow I didn't
notice that before. I've made a sketch and included it alongside the
architectural drawing above.
Another "obvious" acoustic treatment that should be considered is something
near the music performers -- curtains or foam behind the band, foam above
the band in the ceiling, etc. If the stage is to be movable, then the curtains
could be hung from a movable rig. Or even from hooks in the ceiling, if the
stage isn't going to move around much and they don't mind installing new hooks
every time they pick a new stage location.
03-Dec-2001:
I updated the Sabine reverb spreadsheet with more accurate wall characteristics,
and revised the reverb time numbers posted above.
12-Dec-2001:
Yesterday I finally found a local source for Owens Corning 700 series fiberglass
insulation panels. These are what are used in the typical DIY baffle.
For one 2-foot-by-4-foot panel, the cost is $4.08 for 1-inch thick 703,
and $7.36 for 2-inch think 703. They also have 705, but apparently that
panel is not so good for our purposes (more absorption at low freqs but
more reflectivity at higher frequencies). They also have rock wool (aka
mineral wool) which is even cheaper but bulkier and less rigid.
Having handled the 1-inch-thick 703, I think it's possible we might be able
to build baffles with no wood frame -- that is, the 703 would provide the
rigidity. Rigidity is an important issue because I'm considering canting
the panels on either side of the steel support beams, at about a 30-degree
angle off vertical. Effectively the beam would be hidden in a "V" of
panels on either side.
And if I am doing this V-configuration, then the side facing the beam
(the "inside" of the V) is not that important and could be a thin plywood
backing perhaps. Or would that completely screw the absorption characteristics?
I thought this 703 worked best when mounted to a solid surface.
Of the different 700-series panels available, and FRK- or ASJ-faced, the
best one at absorbing 250 Hz is the 2-inch-thick unfaced (plain) 703.
I updated the acoustics spreadsheet
(same as the one linked above) to include the effect of adding
the baffles. I also added some data showing the "diminishing returns"
from adding more and more baffles -- in my opinion, we should do about
30 panels max. Also, if this gets hung in the V configuration, the calc
would probably
be overcounting the contributing baffle surface area, since it's 'hidden'
in the V and not really as exposed to the sound bouncing around the room.
Right?
Another thing to consider when building these baffles is fire safety --
they need to be made of materials that won't burn easily or produce
toxic fumes when burning, at least not right away.
Finally,
a very useful post on rec.audio.pro about DIY baffles.
13-Dec-2001:
One of the things I was advised on was regarding the stage location.
I thought I needed to absorb as much as possible (e.g. heavy curtains,
foam overhead), but I've been advised that diffusion is actually
more important. You don't want the stage area so dead that the musicians
feel like they can't adequately hear each other. One style of diffusor
that was recommended was a large curved surface, like the side of a can.
You take a thin piece of plywood, 4'x8', and warp it around an 8-foot radius.
that would be about 28 degrees of the full cylinder, so it's really a
shallow warp -- only one inch of warp at the peak. These warped panels
then get placed on the walls and ceiling with the convex side facing the
band. Plus you can stuff polyester batting behind it to absorb sound, but
the primary purpose is to diffuse.
Another thing to note is that there's only so much you can do by adding absorption
overhead. Eventually you have to start thinking about other treatments,
such as wall panels for absorption or diffusion.
I'm starting to solidify a plan now:
- OC-703-based hanging baffles, probably on either side of the beams (no "V").
- Diffusion around the stage. Eventually I'll have a separate page for
that too.
- Possible wall panels for specific problem spots if allowed by gallery.
- Bass traps as needed
Step 1 should be done first, and then Step 2 when we can do as time permits.
Step 3 can be held off until we see how things are after the first two
steps have been completed -- we may not need any bass traps.
An interesting link: The
HyperPhysics
site, courtesy of GSU's physics department.
Click on "Sound & Hearing".
17-Dec-2001:
I met with Wanda Dye, Eyedrum's architectural consultant, at the space to
discuss visual impact and placement. She thought white baffles would be
best, so that will be the color of the prototype. For placement, we were
thinking about putting the baffles in the overhead of the central space
defined by the [forthcoming] temporary walls -- approximately a 20' x 25'
space. Foru or five rows of 3 baffles each would fit well.
But placement of these things can change on a whim -- you just need to
install new hooks and move the baffles.
21-Dec-2001:
There is now a separate page detailing
the design and construction of the baffles.
I have built a prototype of the baffle, and it is now at Eyedrum.
The plan right now is as follows:
- Everyone takes a look at the baffle so they see what it really looks
like. In particular, note the color (it's supposed to be white, but
really it's more of a cream color) -- your choices are white, black
and natural (close match to color of wood joists in ceiling).
Note also the hanging hardware options (chain or aircraft cable);
see the baffle design page (bottom)
for more about that issue. A printed copy of that page is with the
baffle.
- I make a list of exactly what materials we need to build the first
batch of 16 or so baffles. The plywood can get delivered on the same
truck that delivers the materials for the new wall. The fiberglass
panels would need to be picked up from the vendor on the southside
-- it'll fill two cars or one pickup truck. Or maybe they'll deliver
it to us (it's a $250 order).
- One day in January we meet at Eyedrum and build all of these.
We'll need a power stapler -- Hormuz says he can borrow one from work.
I only need two people to help -- more is just chaos.
- After placement has been agreed upon, we install the hooks and hang
them. We can always move them -- the hard part is building them.
I plan to do some recordings of the space to a multi-track tape (my Tascam
4-track deck) so we can do before-and-after comparisons.
Questions for an acoustics expert:
- What kind of bass traps should I use? (see links above for various types)
- Where should bass traps go? (note: room is parallelogram, not rectangle)
- What do you think of the DIY bass traps? (see links above for various types)
Summary of acoustic treatment issues:
- Cost (of course)
- Absorption of reverb
- Absorption of high bass (250 Hz) resonance
- Avoid ceiling obstructions and HVAC duct exhausts
- Can't touch floor except for small stage riser (already carpeted anyway)
- Can't touch walls except maybe behind stage
- Stage placement has not been decided yet -- may be movable
Chris Campbell
/ www.chriscampbell.org
/ c-dot-campbell-at-pobox-dot-com
