Designing for the Future

The Warehouse at Dartmouth College is neither a concert hall for classical music nor a studio strictly for recording. It is a spatial sound lab used for both listening and creating. It’s a workshop and gathering space for sonic art, a playground for music creators and listeners of the present and future. Acoustic standards and best practices exist for concert halls and recording studios where the uses are fairly well defined, but how are we to design spaces for a type of music that doesn’t yet exist, and types of interaction that will evolve over time?

The Dartmouth Warehouse project began with cultivating an understanding of the diffusion inherent to the irregularly shaped space.

Designing Traditional Concert and Recording Spaces

There is a wealth of literature around building traditional concert halls, including acoustic measurements and best targets for these spaces. For example, audiences enjoy hearing a certain amount of reverberation in these spaces, and sound energy needs to be reflected and diffused so that it reaches all seats. 

Traditional concert hall spaces are still being designed, but music performance and listening practices have changed significantly in the 20th and 21st centuries. The nature of music itself has changed, as performances regularly include combinations of traditional acoustic instruments, electronic and computer instruments, a variety of types of microphones and synthesizers, and loudspeaker configurations that precisely produce recorded and newly generated sounds. 

Spaces for listening to and recording amplified music are often filled with large amounts of absorption. Whereas a concert hall attempts to assert a unique influence on acoustic sound, these spaces are designed to disappear acoustically. Think of the difference between Carnegie Hall and a modern IMAX theater. In the latter, your awareness is entirely focused on how well the audio system performs. Reverberation is added in the studio, and spaces are tuned via loudspeaker design and placement rather than by shape and surface material alone.

Umbel’s initial steps at the Dartmouth Warehouse were not to dictate what the space needed and how to get there, but to listen.

Listening to the Musician

The Warehouse opened in 2024 as the physical hub for Dartmouth’s new MFA program in Sonic Practice. Umbel played the role of room acoustician in the design of this space. We approached this project as a collaboration with Micah Silver of POLYTOPE, MFA program director Ash Fure, and the architects T+E+A+M and stock-a-studio. We were fortunate to work closely with Ash Fure, whose vision was to create a multi-use sonic environment for Dartmouth students. Ash wanted a versatile space that would work for a wide variety of sonic experimentation.

Listening to the Room

The Warehouse was carved out of Four Currier Place, a three-story mixed-use commercial office building built in 2009. Umbel met with Ash at the site before renovation began. In addition to taking traditional reverberation time measurements,  we spent time listening to the room as it was, in a stripped-down state. 

The space was a complex shape, with non-parallel walls and an exposed ceiling with ductwork and pipes.

Our recommendations focused on keeping the room’s geometric complexities intact and minimizing added acoustic products.

The entire team was surprised at how much we liked the sound of the large space in this condition - it was somewhat lively, but not extremely reverberant, with no discernible flutter echo. The essential characteristics of the sound that we were hoping for were already there. Our recommendations, therefore, included some areas for improvement, but largely focused on keeping the room’s geometric complexities intact and minimizing added acoustic products.

Calibrating Acoustic Simulations

Rather than starting with the assumption that we were working with a reflective box that needed to be controlled and adding absorption to reach a specific reverberation time, this project began with cultivating an understanding of the diffusion inherent to the irregular space. Acoustic modeling with traditional methods (Sabine calculations and Pachyderm simulations) predicted reverberation times for the space that were greater than the actual measured reverberation times.  This discrepancy led us to a better understanding of the simulation model’s limitations, and we were then able to predict how the space would perform with greater accuracy by applying a correction to the modeled results. This brought them more in line with measured reality.

In a space so far removed from the theoretical rectangle that the Sabine equation presupposes, the difference between calculated and measured results is quite significant.

The Sabine equation does not take into account the shaping of the space - only room volume, surface area, and absorption coefficients of the materials in the room. In a space so far removed from the theoretical rectangle that the Sabine equation presupposes, the difference between calculated and measured results is quite significant. Furthermore, absorption coefficients for most of the materials in the space are found by looking for the closest matches available in published data, so they should always be considered approximate.

The Pachyderm simulation uses the same absorption data as the Sabine equation, but it accounts for the shape of the space. The more these simulations can model the actual details of the space, the more accurate they become, limited then by computing power and time. Modeling the overall shape, with its non-parallel walls, brought simulated results closer to the measured results. Modeling all of the space’s details would push the simulation beyond computing capacity, so the simulated results of the overall space were still an overestimation compared to the measured reverberation time.

Listening to the Designers

Umbel’s advice for this space was to use as much of the room’s natural advantages as possible, and to minimize added acoustic products. Diffusion is especially important for eliminating discernible room effects, and this room already included diffusion in the form of the busy ceiling: exposed HVAC work and cable runs, irregular steel beams that run perpendicular to the corrugated ceiling surface, and hanging light fixtures. As we saw when we calibrated our simulations, this complexity reduces reverberation even though the materials are predominantly hard and reflective. 

The architects were happy to leave the ceiling exposed, as the unfinished look complements the programming of the space while adding utility. For example, exposed unistrut channels allow for easy repositioning of lighting and speakers. Modular furniture also allows for a variety of configurations while adding complexity to the space’s geometry.

Having listened to the space, we were confident that the geometric advantages of nonparallel walls and ceiling complexity would outweigh the slightly reduced absorption quantities.

We calculated our recommended quantity of applied acoustic absorption based on meeting previously discussed reverberation time goals for the space. That recommendation is indicated by the solid curve on the plot above. The design team wanted to push the envelope even further and proposed a further reduced quantity of absorption. The reduced quantity is indicated by the dashed curve in the plot above.

Under normal circumstances, we would be wary about changing recommendations. However, having spent considerable time listening to and modeling the space, we were confident that the geometric advantages of nonparallel walls and ceiling complexity would outweigh the slightly reduced absorption quantities. The finished space includes a relatively small amount of applied acoustic absorption scattered over the walls and ceiling. Reverberation is embraced more in this project than is typical for spaces where audio is of paramount importance.

By Elizabeth Teret

Photos used by permission of Brooke Holm