The Crane room

Rose Kitz, Ismael Diaz Mateo, & Geri Tabbachino

Background

ROOM OVERVIEW

The Crane Room in Paige Hall is a shoebox shaped room with a high ceiling. Its total volume is 23,912.5 ft^3 and total surface area is 5,782.82 ft^2. Its main features are its decorative ceiling, ornate wood wall décor, stage at the front of the room, and seven large windows. The main materials in this room are the wood paneling of the lower walls, combed drywall of the upper walls, plaster of the ceiling, thin carpet of the main floor, linoleum of the stage, and double-paned glass of the windows. Assumptions were used for specifics of the materials to better fit observations.

MATERIALS:

As touched upon, not many absorptive materials are in play in this room. Wood, glass, and drywall are not very absorptive especially for higher frequencies.

The carpet in the room is also thin, decreasing its absorption as opposed to if the carpet was thick with a lot of surface area to cause friction. Also, the absence of absorptive, cushioned seating typical of concert halls may contribute to a stark difference in how the room sounds with few people in it compared to during a class, or even more absorptive, a concert.

The prediction is that room will be fairly reverberant especially for higher frequencies given the majority of reflective surfaces in play.

Ceiling

Coffered style, treated as 7/8" plaster, lath on studs

Reflective (diffuse) at higher frequencies, absorptive at lowest (125, 250 Hz) frequencies

Upper Walls

Plasterboard on frame, 50mm airspace w mineral wool

Reflective across all frequencies

Stage Steps

Linoleum, on concrete + wood

Reflective across all frequencies, slightly more absorptive at higher frequencies

Stage Floor

Linoleum, vinyl on subfloor

Reflective across all frequencies, slightly more absorptive at mid-high (2 kHz) frequencies

Lower Walls

12mm wood paneling on 25mm battens

Mostly reflective, quite absorptive at low (125, 250 Hz) frequencies

Carpet

Loop pile tufted carpet, 1.4kg/m2, 9.5mm pile height: On hair pad, 3.0 kg/m2

Slightly absorptive across all frequencies, more absorptive at mid frequencies

Additional Materials & Descriptions

Windows -- 1/8" thick double pane glass with 1-1/4" airspace
- - mostly reflective, with some absorption at low (125 Hz) frequency
People
- - absorptive, with musicians with instruments being nearly twice as absorptive as people standing in audience

Sizing Guidelines

200-300 ft^2 per seat is standard for a multipurpose hall
For the Crane Room's volume of 23,912.5^3, the room is ideal for 80-120 seats.
Currently, there are about 60 seats in the Crane Room. Adding another 20-60 seats would drastically change the way sound sounds in the room due to the fact that the room would be much more absorptive with more people. The sound of the room for a lecture might bee quite dry and would maybe require amplification which is otherwise unnecessary at its current state.

DIMENSIONS

Measurements were taken in the room and were sketched and calculated by hand.

Side Walls

Floor

Ceiling

Front & Back Walls

PROGRAM

The Crane Room in Paige Hall is a lecture hall for classes but also doubles as a concert hall at Tufts for mainly smaller, rock bands as shown to the right. Musical performance events like AppleJam occur in the Crane Room.

Classes typically have one lecturer speaking from the stage towards the students at the ~60 moveable plastic desk-chairs on the carpeted floor. Some classes held in the Crane Room are discussion-based, while others rely more on lecturing.

Concerts typically include rock and jazz bands with many different, fairly loud & amplified instruments of all different timbres. The plastic desk-chairs are removed from the space for concerts.

Rock concert in Crane Room (facing stage)

Amplified performance in Crane Room (facing stage)

Typical Crane Room set-up for classroom lectures (from stage)

Room Characterization

METHODS

IMPULSE RESPONSE MEASUREMENTS

#1: Balloon Popped Center Stage with Microphone Centered 4ft Away in Front of Stage

#2: Balloon Popped Center Stage with Microphone Centered in the Back of the Room

CRITICAL LISTENING - listening to how different sound sources react in the room with various receiver locations

MUSICAL OBSERVATION

Ismael played acoustic guitar while Rose sat towards the front of the room and Geri sat towards the back of the room to observe and listen intently.

Each took notes individually & discussed afterwards to compare listening experience in different parts of the room regarding 5 different parameters - Reverberation, Clarity, Loudness, Warth, & Envelopment.

Geri and Rose also sang “Row Row Row Your Boat” while Geri sang higher pitches and Rose sang lower pitched - simultaneously.

SPEAKING OBSERVATIONS

Given the relatively symmetrical shoebox layout of the room and multipurpose nature, a variety of speaking tests were performed. For each test, Rose stood on the stage (at original balloon source location), Ismael sat in middle of room, and Geri stood in back of room.

3 different tests:

Rose talking (from source location on stage, where musicians or a lecturer would stand)
Three people talking through normal conversation
Simultaneously, Geri spoke the alphabet, Rose talked about her run that morning, and Ismael spoke in Spanish

QUANTITATIVE RESULTS

Impulse Response Measurements

Note: T20 was selected over T30 to represent the reverberation time (including late reflections) in the Crane Room, since the range of -5 to -25 dB includes a more linear Schroeder decay and higher signal-to-noise ratio than the range of -5 to -35 dB (see plots below).

See the 'Discussion' section for a comparison of measured T20 to theoretical reverberation time found using the Sabine Equation.

Microphone in Front of Room

Bass Ratio = 0.75251

Microphone in Back of Room

Bass Ratio = 0.82351

QUALITATIVE RESULTS

Reverberation

The higher frequencies seemed to last longer in our ears in opposed to the lower frequencies which seemed to decay quickly. The lower notes felt drier. Although, the noise felt like it "stayed" in the room and does not just disappear, in a sense. It seems as though the wooden, reflective wall behind the stage sends sound to the back of the room.

Clarity

The initial strike of the various notes sounded very clear regardless of the pitch. The music and its melody were heard very well and each note was very distinct from each other. Although the clarity seemed high most likely due to the fact that there are not many absorptive materials in the room and each subject listening had a unblocked path of direct sound.

Loudness

The lower pitched notes sounded much more quiter than the higher pitched notes. The higher notes also seemed to "pop" more, meaning they pierced the ears slightly more than the mid/low notes. The room supported the acoustic guitar well, particularly at higher frequencies. Because of this notion, amplification is unnecessary or undesirable.

Warmth

The room sounded very warm, not cold nor harsh sounding. The acoustic guitar has a very warm timbre whiched helped to draw this conclusion because the instrument felt as though it fit the room well. Other instruments might not have this sam effect such as electric guitar or drums. The higher notes sounded brighter than others, but notes played sounded particularly dull.

Envelopment

The sound of the guitar felt like took up the entirely of the room from both the front of the room and the back of the room. Although, in the back of the room it felt as those the sound level was much lower or quieter. It was more noticeable when sound felt louder. There were not distonct reflections felt back to the guitar player but reverberation was still felt as present.

Discussion

COMPARISON: MEASUREMENTS vs. LISTENING EXPERIENCE

Comparison of Front and Back of Room Microphone Locations

The plot to the right demonstrates the generally higher measured reverberation time in the back of the room, compared to listening in the front of the room. It is interesting that for each parameter (EDT or T20), the piecewise slopes are similar for both receiver locations; however, the values are shifted upwards to higher reverberation time for the back-of-room receiver. Noticeably, the EDT in the front of the room is lower than the EDT in the back, which aligns with the listening experience of slightly more reverberation in the back of the room. This higher EDT in the back of the room also proves the theory that a receiver in the back of a room, further from the direct sound source than a receiver in the front, gets more reflections compared to direct sound.

Measured T20 vs. Estimated Sabine RT

The plot above demonstrates the slightly lower reverberation time using the Sabine equation for some octave bands, compared to using the measured T20. It is interesting to see that there is about 2.3% difference between reverberation times for the 125Hz and 2kHz octave bands while other octave bands have about 13-15% difference, 250Hz and 1kHz primarily. The trend of higher octaves having a longer reverberation time is seen on both the Sabine Equation and the Measured T20 times.

Both of the T20 and Sabine Equation results fit the qualitative listening experinece of being in the room has a noticeable reverb and having a reverb of 1-1.5seconds is understandable when there is one individual speaking just like our experience. These parameters do not do a good job at describing perceptual attributes like envelopment. To better describe envelopment, directional microphones would be better used to find the sound energy from different directions of the receiver in the room. Clarity would also need another metric as it the sound energy over time is not fully portrayed by T20 or Sabines Equation. A graph of loudness or sound energy over time of the source would be better at making a perceptual correlation of the clarity of sound.

RECOMMENDATIONS FOR IMPROVEMENT

Noting the multipurpose nature of the Crane Room, several modifications could be made to enhance both the concert-going/music-playing and classroom lecture experience. In general, the Crane Room is quite reverberant and the clarity is quite low. While this serves acoustic, small-group musical performance well, the many reflective surfaces in the room could cause an overly-loud and muddled sound for amplified, larger-group musical performance. Additionally, from the experience speaking from the stage, the lack of clarity of the room was distracting, and noticeable by the group members standing at the two microphone receiver positions. Thus, it would be beneficial for all users of the Crane Room to be able to adjust the room's configuration to increase C50, which correlates with human perception of clarity.

The most effective way of incrreasing clarity would be to reduce the late reflections and increase early reflections. This can be done by adding absorption in the back of the room, further away from the sound source on the stage. This can be done by replacing the reflective wood on the side walls in the audience area with an absorptive material such as perforated walls. It would be beneficial as well to focus the absorption in these walls on lower octave bands to reduce muffled noise for deeper instruments when higher frequenices are taken care of by other materials which are better at absorbing higher frequencies.

A reflective panel close to the sound source above the stage area would also be successful at increasing early reflections from the instruments or speaker. This would increase the ratio of early to late reflections and improve the clarity both quantitatively and qualitatively for a better listening experience.

EXPERIMENTAL PROCESS

Limitations of Study and Improvements

Although this room was fairly simple to take measurements from, there is always room for improvement when conducting any experiment. In order to get more accurate measurements, the group could have used a more precise tool like a laser measuring tool instead of a tape measure. With the tape measure and simple standing height, the accessibility to the higher measurements in the room was limited such as the upper walls and the ceiling. Some of these difficult to gather measurements were estimated for the Sabine equation and could have proven a more accurate reverberation time. Other ways in which we could have improved our results was determining the exact type of wood was used on the walls, measuring exactly how thick the carpet was, and determining for certaing if the space behind the walls was hollow or not.

Some limitations of study including being unable to get extremely accurate measurements of the room, discounting certain objects in the room such as the desk, projector, and lecturer stand, and not knowing the exact materials being used around the room. It is important to note that the surfaces of this room has to be assumed, guessed, or estimated due to the fact that we had no accurate documentation o what the materials are. We used observation usualy visual and audio cues to try to match what we saw with what we heard. This estimation provides more room for error and is a large limitation of our study. Finding the offical materials would drastically increase the accuracy of this study.

Another unique limitation of this multipurpose room compared to typical concert halls is the absence of absorptive, cushioned audience seating that could mimic some absorption that people provide even when the room is unoccupied. The recordings from this case study were

Additional Measurements

Additional measurements, apart from improving accuracy of existing measurements, could be made to better characterize the behavior of the room as it serves different programs. Parameters of interest include:

strength (perceived as loudness), to determine a healthy limit for the number of amplified musicians playing at the same time in the room (i.e. at a rock concert)
stage support, to discover how well an ensemble (or even lecturer) could hear (early) and how much they perceive reverberation on the stage. Gaining more insight on the lecturer/musician experience could be helpful in determining where to position added adjustable absorption in the room to ensure that sound sources are less distracted by reverberation while still being able to hear themselves on stage.

The strength parameter could be attained by comparing the existing impulse measurement in the room to a new impulse response measurement of the balloon pop in an open field. The "free field" measurement should be taken with little background noise and an omnidirectional microphone 10 feet away from the balloon source.

The stage support (early & late) parameter could then be found by performing the same balloon impulse response measurement from the stage in the Crane Room as before, but with an omnidirectional microphone located on stage, rather than in the seating area.

Lessons Learned

Throughout this study, we learned about how a shoebox shaped room reacts to certain sounds and how different materials absorb and reflect the octave band frequencies different. The group also gained experience with the process of recording and analyzing and impulse response measurement. We became more familiar with microphone technology, Matlab program coding, and using online spreadsheet programs to calculate the Sabine equation. We have a much better understanding of reverberation and how sound decays in a room.

MODEL SUMMARY

VISUAL REPRESENTAION

Crane Room

W/ Perforated Wood on Lower Walls

W/ Perforated Wood on Lower Walls and Reflective Ceiling Panel

AboVe are the CATT models displayed the sound source locations for the various redesign auralizations as well at the two different receiver locations in blue.

DISCUSSION OF DESIGN PARAMETER: CLARITY

The goal for the redesign models of the Crane Room is to improve the room's clarity, qualitatively ad quantitatively. Clarity qualitatively refers to how clear sound is heard. This is especially important and more noticeable when multiple sounds/notes are being listened to simultaneously. C50 is the quantitative measure of calrity and is a logarithmic ratio between early (0-50ms) and late (after 50ms) reflections. C50 is measured in decibels, as it is a ratio.

Because Crane Room is used as both as a lecture space as well as a space for amplified music, it is imperative that the room works acoustically for both applications. With regard to lecture speaking, the room in its current state is successful at providing a space in which unamplified solo speaking is supported well enough by the reverbertaion of the room and also clear enough to a student sitting in any location in the seating section. Although, with regards to louder music performances with many instrumemts, such as amplified rock bands or lively jazz bands, the room's early and late reflections blend together creating a muffled or jumbled sound of the instrument notes as they play together.

To improve the clarity, two new design models were simulated and tested, one with a smaller amount of changees to improve clarity and the other with a higher amount of changes for even better clarity. The redesigns were created with the idea that amplified music with a significant amount of lower frequency noise would be playing on the stage.

REDESIGN #1

PERFORATED WOOD ON SIDE & BACK WALLS IN AUDIENCE SPACE

In this redesign, the wood placed on the lower walls is replaced with perforated wood in just the audience/seating area. The goal of this redesign is to drastically decrease late reflections off the walls farthest from the sound source of the music instruments on stage. The side and front walls in the stage area will remain as is with the reflective wood on the lower walls. By decreasing the late reflections, the C50 ratio of early to late reflections will increase.

The perforated wood is designed with larger holes to account for the lower frequency octave bands, which are the frequencies that become the most muffled between notes. The higher frequencies are more easily absorbed by the other materials so the main concern is the lower frequency sounds.

REDESIGN #2

PERFORATED WOOD ON SIDE & BACK WALLS IN AUDIENCE SPACE + REFLECTIVE PANEL ABOVE STAGE AREA

This redesign keeps the perforated walls as in Redesign #1, but to further improve the clarity, there is a reflective panel hanging from the ceiling above the performance stage. The perforated walls will decrease late reflections, but the reflective panel will increase early reflections due to its close proximinity to the sound source. Both of these combined will further increase the C50 ratio and create a very clear, unmuffled sound for both the audience and musicians alike.

SOUND SOURCES & LISTENING LOCATIONS

There are 12 differennt auralizations run and they are displayed below.

The there are two sound sources that are used in the auralizations in CATT Acoustics. The Jazz Band Audio ("JazzBand_stereo.wav") is the anechoic recording most similar to an amplified rock band recording which is the music truly played in the Crane Room. Although the jazz band recording is different music and is umamplifed, it is the most comparable because both styles of music are generally louder, include multiple instruments playing simultaneously, and has a large presence of lower frequencies. The male speech audio ("alternateMaleSpeech.wav") used as the sound source is the most similar to a lecturer speaking to his class in the Crane Room. The sound sources are both omnidirection sources that are placed in the center of the statge - where the musicians play and where the lecturer speaks.

An auralization was done for two different reciever locations, as well. One location is in the front of the audience/seating area and the other in the back of the audience/seating area. These two reciver locations simulate the two realistic locations with the most variability from each other.

SIMULATION RESULTS

AURALIZATION

THEORETCICAL CONTEXT

An auralization, as defined by Mendel Kleiner in 1993, is the "process of rendering audible, by physical or mathematical modeling, the sound field of a source in a space, in such a way as to simulate the binaural listening experience at a given position in the modeled space". In other words, an auralization is the modeling of how a sound in experienced in a certain room.

The three main components of an auralization are an impulse response which is measured or simulated, a convolution of the impulse response with an anechoic audio, and a spatial rendering of the convolution. All three of these components have been achieved and modeled in SketchUp, CATT Acoustic, and Matlab. All three softwares are used together to achieve an accurate auralization.

Jazz Band

Room 1

Room 2

Room 3

conv_timedomain_jazz_room1_recf.wav

Front

conv_timedomain_jazz_room2_recf.wav

conv_timedomain_jazz_room3_recf.wav

conv_timedomain_jazz_room1_recb.wav

Back

conv_timedomain_jazz_room2_recb.wav

conv_timedomain_jazz_room3_recb.wav

Voice

Room 1

Room 2

Room 3

conv_timedomain_voice_room1_recf.wav

Front

conv_timedomain_voice_room2_recf.wav

conv_timedomain_voice_room3_recf.wav

conv_timedomain_voice_room1_recb.wav

Back

conv_timedomain_voice_room2_recb.wav

conv_timedomain_voice_room3_recb.wav

DISCUSSION

IMPACT

The auralizations of the two redesigns were successful in improving the clarity to the extent that was assumed. Quantitatively, the C50 values for the first redesign were extremely higher in comparison to the that of the original Crane Room. The first redesign auralization increased the C50 values considerably while the second redesign increased the C50 values even more. This make sense as the second redesign decreased late reflectsions as well as increased early reflections. The improvements in clarity were mostly seen in the lower frequencies, especially the 125Hz and 250Hz octave bands. This is shown in the graph below labeled "C50". The reverberation time also decreased with the addition of the redesigns, but this was not shown in any paticular octave band.

Qualitatively, the sound in the room improved with clarity and varied depending on the different sources and receivers. With both the jazz band audio recordings and the male speaking audio recordings, the clarity of individual notes and sounds was much clearly, more for the second redesign than the first, but both redesigns proved to be an improvement. Regarding the sound receivers, the redordings with the reciever in the back of the room had much less sound energy and were significantly quieter than those with the sound receiver in the front of the room.

SUCCESS?

The goal was to increase the clarity and therefore decrease the effect of the lower frequency instruments which sounded much more muffled in practice and in the auralizations. This was succeeded as shown by the results, the audio sounded much clearer for the redesigns with less audible muffling for both higher and lower frequencies. The calculated results prove this successfully, as well, as described above. The lower frequencies of rock bands or loud music would be less muffled and more enjoyable for an audience. A class of students would be able to clearly hear their lecturer articulate and sound would carry efffectively. The results of the auralizations of the redesigns help us to undertsand that the modifications would truly work in the space for the given activities such as rock bands and lecture speaking. The iterations would make the Crane Room more acoustically sound for these purposes and the plan for changing is feasible to construct.

FEASIBILITY OF RECOMMENDED CHANGES

The cost of implementing perforated wood is fairly high, but there would be no other great physical limitations for redesigning if money is not an issue. Stripping the room of its original wood could perhaps taking away some of the room's character from the original architectural design, changing it to a more modern feel. Adding an adjustability feature to the perforated wood would prove to be even more expensive, but perhaps worth it for the multi-purpose room. Adding the hanging panels would also prove to be pricey. Another potential issue regarding the creative integretity of the room is that the hanging panels ccould block the sight of the people on the floor to the ornate ceiling above. But perhaps, the panels would hang low enough that the ceiling cold still be viewed in its entirety.

SUCCESS & LIMITATIONS OF THE AURALIZATION PROCESS

The software programs used to do auralizations does a very good job at accounting for many of the details that go into simulation sound in a given space. Exact dimensions and details can be achieved in SketchUp to create the most accurate version of the room studied. CATT allows you to know exactly the number of sound rays that are being used to study the room and how they react with the shapes and number of ray splits. CATT allowed us to use specifics directionability for the sound sources and receivers to model how a human would hear most accurately.

There are always limitations from the auralization process, and the largest disadvantage from usually binaural audio in the study of the Crane Room is the lack of in-head localization, meaning the natural way that human beings can identify where a sound is coming from regardless of other sense is eliminated. This reduces the accuracy of how the sound truly sound when one would be inside the room. Using binaural audio, and specifically when one listens with headphones, the ability to account for head tracking is lost. This and in-head localization go hand-in-hand in the fact that the human brain is able to sense where a sound is coming from regardless of the direction in which the body is facing.

FUTURE ITERATIONS

Other ideas for the redesign of this room were discussed. One way to further increase the clarity of the room would be to keep the perforated walls in the audience/seating area and the panels as discussed, but already reconstrucing the wood of the lower walls in the stage area to be even more reflective material. This would increase the energy of the early reflections even further. Another very simple way to increase clarity and decrease reverberation time is to add thick curtains on the windows all around the room. Although, the windows do not cover a lage amount of surface area in the room so the effect of the curtains might not be too large. The curtains however could be an easy, cheaper solution to increase clarity slightly. Another idea that would require a large amount of reconstruction is to replace the thin wood walls with thicker, less reflective wood and increeasing the air space behind wood. This woud be difficult to tackle lower octave band frequencies but would still reduce the sound energy reflecting around the room.

Page updated

Google Sites

Report abuse