STUDENTs

Hui Yee Poh

SEMESTER

HS26

Assignment 1

Space A_Level 4 Corridor

Soundscape Sketch:

Space A is a long, narrow corridor on the 4th floor of my student accommodation in Zurich. It is mainly used to access the communal roof terrace, communal bathrooms, and each individual’s room. Occasionally, people also use the corridor to have short, friendly chit-chats with their neighbours. The space is largely untreated, featuring a bare concrete ceiling, painted concrete walls, and metal doors.

Audio of muffled conversations happening down the corridor, followed by a loud, echoey clap.

Acoustically, it is a highly reverberant, almost cavernous space. When I chat with my neighbours, our voices sound echoey and especially loud, lingering along the length of the corridor. I can also hear conversations happening further down the corridor quite clearly, as well as the sound of running water from the showers. This makes the corridor feel very lively, even when there are only a few people using it.

Audio of my voice stepping further away from the recorder.

The sounds that capture my attention the most are the mechanical sounds of my neighbours’ doors and the bathroom door opening and closing. The clicking sound of a doorknob turning is particularly crisp and distinct, even when the source is quite far down the corridor. Conversations are also especially attractive to my attention, despite generally having a lower-frequency character than these sharper mechanical sounds. This could be because there is interesting information being transmitted through speech. Perhaps the further away the conversation is, the more attractive it becomes to me, as I become more curious about what is being said. Interestingly, when listening back to my recordings, I felt that my voice became clearer as I moved further away from the recorder. This may partly be because I was also stepping further away from the wall, reducing some of the immediate reflections around my voice and making the direct sound feel more distinct.

Personally, I do not find the soundscape very pleasant because it creates a sense of lacking privacy. The corridor amplifies very intimate moments of daily life, such as showering, entering or leaving a room, and having personal conversations. At the same time, some sounds make me particularly alert. The click of a door opening or closing immediately signals that a neighbour may be entering the corridor. When I feel social, it alerts me to greet the incoming neighbour; when I am out of social battery, it makes me want to quickly return to my room.

I think the acoustic character of the corridor is largely caused by its hard, reflective surfaces. The concrete walls and ceiling, together with the metal doors, provide very little sound absorption, allowing sounds to reflect repeatedly through the long and narrow space. This may explain why voices and everyday noises feel louder and more reverb-y than I would expect them to be.

Looking back at my soundscape sketch, I also noticed that most of the louder and higher-pitched sounds are clustered around the bathroom, next to where I was standing. This includes the clicking and closing of the bathroom door and sounds associated with people entering and leaving the space. Meanwhile, voices and running water can be heard from further away and seem to travel along the corridor. This spatial pattern matches my experience of the corridor: nearby sounds can feel sharp and immediate, while more distant sounds remain surprisingly clear because of how reflective the space is.

Space B_Studio Space

Soundscape Sketch:

Space B is my studio space under Chair Cifuentes Utrero, located in the Immersive Studio in the HIL Building. My studio has a unique focus on working with local natural materials, and the space has therefore also been converted into a model-making and construction area for activities such as rammed earth and dry stone wall construction.

Audio clip of my voice stepping further away from the recorder.

The space was initially intended as Gramazio Kohler’s research studio, but it has since been repurposed for our studio use. Because of this, it does not seem to have been designed acoustically for the kind of loud and repetitive construction work that now takes place inside it. There is no extensive acoustic treatment to absorb the mechanical noises and heavy thumping sounds produced during ramming, with only parts of the ceiling covered by acoustic panels.

Audio clip of ambient sounds in the studio (ramming, chatting, footsteps…).

Unlike the corridor in Space A, where individual sounds can travel clearly through an otherwise relatively quiet space, the studio usually contains several overlapping sources of sound at once. Conversations happen continuously around the worktables, while footsteps, tools, drilling, hammering, and the ramming of earth occur intermittently across the room. This makes the space feel active and occupied almost all the time.

The sounds that capture my attention most strongly are the mechanical and construction-related sounds. Drilling and hammering have a sharper, higher-pitched character, while the ramming of earth produces a lower and heavier thumping sound. These sounds make me more alert and aware of my surroundings, especially when construction activities are happening nearby, as I need to make sure that I am working safely. In comparison, conversations feel less intrusive and tend to blend into the general background of the studio.

Surprisingly, I find the overall soundscape quite acceptable, despite the frequent use of heavy machinery. When the professor or teaching assistants make announcements to the studio, their voices are usually clear and easily understandable without having to raise their volume significantly. Group discussions around each table also feel relatively well-contained and do not strongly interfere with neighbouring discussions. I think this may partly be attributed to the acoustic panels on the ceiling, which help absorb some of the reflected sound.

At the same time, the soundscape is also influenced by the large open volume of the room and its relatively hard surfaces. Although the ceiling panels provide some absorption, the floor, walls, tables, and other exposed surfaces still allow sound to travel across the studio. This could explain why activities happening in the model-making area remain clearly audible from my working position further away.

Compared to Space A, however, the studio feels less echoey and cavernous. Instead, the large number of simultaneous sound sources makes it feel more diffuse and acoustically busy. Looking back at my soundscape sketch, the louder and higher-pitched mechanical sounds are concentrated around the model-making area, while conversations are distributed more evenly throughout the studio. This spatial pattern closely matches how the room feels to me: the construction zone is acoustically more intense, while the rest of the studio is filled with a softer and more continuous background of voices.

Assignment 2

Space A_Corridor

The corridor is not acoustically treated. The ceiling is bare concrete, while the walls are concrete with some textured paint and the floor is concrete with some laminate. They are materials that do not effectively absorb sound across all frequency ranges. The space is notable with its long, extended parallel surfaces that continues in an L-shape manner. All these hard surfaces are sound reflecting, hence amplifying all the sounds in the space. With such little acoustic treatment and hard materials, it is expected for this space to be especially reverb-y and cavernous.

RT60 = 2.36 secs

There is subsequent echoes visible in the decay as seen by the discrete late peaks.

Floor area: 50m^2

Wall area: 200m^2

Volume: 155m^3

With reference to coefficients suggested by LLMs,

Floor (500Hz/1000Hz) = 50 x 0.03 = 1.50

Walls (500Hz) = 200 x 0.02 = 4.00

Walls (1000Hz) = 200 x 0.03 = 6.00

Ceiling (500 Hz) = 50 x 0.015 = 0.75

Ceiling (1000Hz) = 50 x 0.02 = 1.00

Total Absorption (500Hz) = 6.25

Total Absorption (1000Hz) = 8.50

Predicted reverberation time = 0.161 x (155/6.25) = 3.99s (500Hz)

Predicted reverberation time = 0.161 x (155/8.50) = 2.94s (1000Hz)

500Hz1000Hz
Predicted2.002.00
Calculated2.362.36
Sabine3.992.94
Online tool3.752.75

The major difference of the online tool, compared to the Sabine calculation is that the modeled space is a long rectangular volume, when in reality it is an L-shape volume where more reflection may happen at the turning corner, where some reflection is lost as it travels along the bend. The background noise of footsteps in the calculated recording may have also affected the accuracy of the calculation. Additionally, the east wall has clerestory glass windows, which could have higher sound absorbancy, thus causing the calculated result to be lower than that of the Sabine and online tool.

It is interesting to note that in person, the reverb I heard was not perceived to be as long as when I heard the recording captured by the microphone.

Space B_Studio

The studio has perforated ceiling panels that appear designed to absorb sound, while the painted walls, glass windows and hard floor remain largely reflective. Tables, chairs and models break up the otherwise parallel surfaces and scatter reflections, while occupants and their clothing provide additional absorption. The ceiling treatment is visually integrated with the lighting in a regular grid. Compared with the corridor, the studio is expected to sound less reverberant, although overlapping conversations and construction activities can still make it feel loud and busy.

RT60 = 0.12secs

There are no subsequent echoes in the recording. (Note, there were quite some background noise during this recording)

Floor: 180m^2

Ceiling: 180m^2

Walls: 168m^2

Volume: 540m^3

With reference to coefficients suggested by LLMs,

Floor (500 Hz/1000 Hz) = 180 × 0.03 = 5.40


Solid walls (500 Hz) = 120 × 0.02 = 2.40


Solid walls (1000 Hz) = 120 × 0.03 = 3.60


Glass windows (500 Hz) = 48 × 0.18 = 8.64


Glass windows (1000 Hz) = 48 × 0.12 = 5.76


Acoustic ceiling panels (500 Hz) = 135 × 0.90 = 121.50


Acoustic ceiling panels (1000 Hz) = 135 × 0.65 = 87.75


Remaining hard ceiling (500 Hz) = 45 × 0.015 = 0.675


Remaining hard ceiling (1000 Hz) = 45 × 0.02 = 0.90


Total absorption (500 Hz) = 5.40 + 2.40 + 8.64 + 121.50 + 0.675 = 138.615


Total absorption (1000 Hz) = 5.40 + 3.60 + 5.76 + 87.75 + 0.90 = 103.41


Predicted reverberation time at 500 Hz RT60 = 0.161 × (540 / 138.615) = 0.63 s


Predicted reverberation time at 1000 Hz RT60 = 0.161 × (540 / 103.41) = 0.84 s

500Hz1000Hz
Predicted0.500.50
Calculated0.120.12
Sabine0.630.84
Online tool0.420.43

I think the disparity betwen the calculated and Sabine/online tool is because the studio contained many people, furniture and model making materials (rammed earth that has good sound absorbancy) that were ommited from the simplified calculation. Additionally, some timber was also placed on the walls that may have absorbed more sound.

©

Architectural Acoustics

Department of Architecture logo
ETH Zurich logo