Exploring Sound Within Structure
Acoustics Around the Dartmouth Campus
JB Cholnoky & Cyrus Tingley
Math 5 • Professor Barnett
Spring 2007
Project Purpose & Goals
• Given the often-overlooked presence of
sound and acoustics at work all around us,
the purpose of this project was to apply the
theories and methods learned in Math 5 to
our everyday surroundings here on the
Dartmouth campus.
• We hoped to discover interesting and relevant
acoustic anomalies in campus buildings,
explored with sound recording and analysis.
Project Sites, Methods, &
Materials
• We chose three different venues as our
subject structures: Leverone Fieldhouse, the
steam tunnels beneath the Green, and an
empty fuel storage room attached to the
Dartmouth power plant.
• The basic procedure in each location
included the production of a loud sound
signal, which in conjunction with any resulting
reverb, resonance, or echoes was recorded
on a laptop using Audacity.
Project Sites, Methods, &
Materials
• To produce sound, we struck together two
pong paddles (made of laminated plywood
roughly .75 cm thick) - this was sufficiently
loud and quite easily repeatable. Each time,
the edge of one paddle was hit sharply
against the flat center of the other.
• A highly sensitive Logitech digital microphone
was used to record sound directly to Audacity.
Steam Tunnels
Steam Tunnels
Steam Tunnels
• Courtesy of FO&M, we explored a 300-foot straightaway in the
tunnels located between the SE and NE corners of the Green.
• Tunnel dimensions were as seen on the diagram - however, the
presence of two large stream pipes running along one wall is not
appropriately detailed - we are quite sure this effected the
results we got.
• Nevertheless, we did discover some interesting acoustics;
surprising given our expectation of one, there was no echo
apparent in the tunnels - instead we found an example of nearperfect reverb.
• Because the tunnel turned at 90-degree angles at either end of
the 300-foot subject section, it behaved much like a closed room
and sound resonated accordingly.
Steam Tunnels
•
•
•
•
Calculating reverb time (without accounting for pipes in the way):
T60 = 0.161 (V/S); V = 544.4m3
Peak frequency = 557 Hz [from Audacity]
S = (AC of concrete)(SA) = (.02)(904.5m2) = 18.09
– Absorption Coefficients for Concrete: Click Here
• T60 = (0.161)(544.4/18.09) = 4.85 seconds
•
•
By zooming in on our Audacity recording, we found that the sound
began at 4.85 sec, ending at 6.26 sec (see Audacity), implying a reverb
time of only 1.41 seconds.
However, while this does not match the calculated value, the difference
can be attributed to perhaps human error and/or lack of microphone
sensitivity; but most obviously, the calculated T60 value is for an empty
tunnel hallway and thus would be longer than the reverb time in reality,
which is effectively much shorter because of the pipe’s interference in
space, insulating covers, etc.
Power Plant Fuel Room
Power Plant Fuel Room
• We stumbled on this
site basically by
accident while touring
the plant with FO&M - a
large, relatively empty
concrete room once
used to store fuel.
• Dimensions are as seen
on the diagram.
• This room had
incredible acoustic
qualities - though not
exactly what we
expected.
Power Plant Fuel Room
• Surprisingly, this room did not produce the echoes we
expected - considering the sound our voices made
when we walked in, the room having a booming,
amplifying effect on them.
• Most notable about this room was its tremendously
long reverb time.
• Listen to several of our Audacity recordings from this
room - for both higher- and lower-volume noises, the
sounds hangs around for quite some time - without
echoing though!
Power Plant Fuel Room
•
•
•
•
Peak frequency of signal = 464 Hz
Calculating reverb time: T60 = 0.161 (V/S)
V = (30.48m)(6.096m2) = 1133m3
S = (AC of concrete)(SA) = (0.04)(780.4m2) = 31.2
– [Absorption Coefficient for Rough-Finish Concrete at 500Hz found
online]
• T60 = (0.161)(1133/31.2) = 5.85 sec
• By zooming in our recording in Audacity, we found that the
sound appeared to have a timespan of 12.613 sec -> 18.49 sec,
or 5.877 sec.
• At this site, we were thus overwhelmingly more successful in
accurately measuring and calculating the room’s reverb time.
Leverone Field House
Leverone Field House
• Leverone was the first site we recorded at and while
we initially believed there to be nothing special at
work, our results from Leverone were in fact the most
remarkable.
• Dimensions/layout of the building can be seen on the
following diagram.
• What we discovered at one location in the building
was a perfectly clear flutter echo.
Leverone Field House
• We recorded at 3 locations - #2
and #3 provided the best
results. The flutter echo was
heard best when sound was
produced at #3 and recorded at
#2 - we believe the echo was in
fact sound bouncing off the roof
of the building - our
experimentation with
recording/production locations
supported this.
• We believe the large amount of
nets/flags/screens/material
hanging from the ceiling and at
the far end of the building stifled
any echoes we may have heard
traveling end-to-end in
Leverone Field House
• Check out our Audacity recording! It was only by slowing the
sound down 70-90% that we were able to both see and hear the
echo.
• A “flutter echo”, as we learned in class, is the sound of repeated
clicks gradually fading away - continuous ‘bouncing’ of sound
within a structure.
• The time before you hear the echo: T = 2x/c
• According to Audacity, the initial sound arrived at 3.99 seconds,
and zooming in showed echoes arriving at 4.096 sec, 4.198 sec,
and so on - thus the time delay was approximately .104 seconds
or so.
• We suspected that the echo was produced by the sound we
made at Site #3, bouncing off the ceiling.
• At that point, the height of Leverone is roughly 24.5 m.
• Plugging .104 sec into the equation referenced above:
– .104 = 2x/340 m/s
Leverone Field House
• Thus a calculated reflective distance of 17.68 m does
not correspond to any known/relevant distance within
Leverone.
• We suspect that the echo was in some manner
bouncing off the roof above our heads at Location #3
- 24.5 m was a measurement estimated by JB as
FO&M was unable to furnish us with complete
architectural plans.
• In conclusion, it would be nice to definitively know the
path and distances associated with the flutter echo
we heard - though any number of factors could’ve
interered with the connection we sought to make.