ACOUSTIC DESIGN OF A NON-ENVIRONMENT CONTROL ROOM FOR TEACHING A A A Proceedings of the Institute of Acoustics ACOUSTIC DESIGN OF A NON-ENVIRONMENT CONTROL ROOM FOR TEACHING Glenn Leembruggen Acoustic Directions, ICE Design Australia Camille Hanrahan-Tan Acoustic Directions Osborn Fong Acoustic Directions 1 INTRODUCTION This paper discusses the design and acoustic performance of a newly-constructed control room at JMC Academy in Sydney Australia. As the concept of a non-environment room is not well known in Australia, this may be the first non-environment room in the country. In 2002, the principal author first heard a non-environment room in Portugal that Philip Newell had designed, and was extremely impressed with the sound quality, particularly the consistency of the low frequency sound energy in the room. Although the concept of the non-environment room was first developed by Tom Hidley, Newell has propagated non-environment control rooms in Europe and published a number of books on recording studio design [1]. 1.1 Design Goals and Constraints The design goals for the control room were: a) Non-environment room at mixing position, as propagated by Phillip Newell [1]. b) A reverberation time (RT) as low as possible whilst: • Retaining a hard floor and front wall • Maintaining a flat frequency characteristic from 30 Hz to 20 kHz c) Decay times consistent with frequency from 30 Hz to 15 kHz. d) Provide a half-space acoustic environment for the loudspeaker monitors to deliver the best possible bass response. For the loudspeaker to be free of image sources, the half-space area should be as extensive as possible. [2] [3] e) Room to readily support teaching with comfortable communication between mix position and rear bench seat. f) If required, a controlled image design as per the approach described by Walker [4]. g) Sound energy components arriving at the mix position should be more than 20 dB below the direct sound during the period up to 20 ms after the direct sound over the range 1 kHz to 10 kHz, as proposed by Walker [4]. h) Room modes to be essentially non-existent (consistent with a non-environment room). The constraints for the control room were: a) Existing shell with dimensions of 5.5 m x 5. 6 m. b) Limited width of the front wall to angle the walls to prevent reflections at low frequencies from the left speaker on the wall providing the half-space environment for the right speakers and vice-versa. c) Unable to accommodate the usual depth of 1 m for absorbers on walls and ceilings as typically used in non-environment rooms: • Maximum depth of sound absorption on rear wall: 500 mm • Maximum depth of sound absorption on ceiling: 500 mm to 1.2 m • Maximum depth of sound absorption on walls: 200 mm d) Bench seating for students to extend across the rear of the room, with cushioning up to the top of heads, which would obstruct the 500 mm deep sound absorption behind the seat. Vol. 42. Pt. 3. 2020 Proceedings of the Institute of Acoustics 2 DESCRIPTION OF THE ROOM Features of the design of the control room are: 2.1 Reflection free zone As noted earlier, the room has insufficient width to provide the wide, almost-flat half space walls that are a feature of many of Newell’s rooms. For example, see Figures 13.4, 13.7 and 16.1 in Newell [1]. This lack of width implied that the monitor wall needed to provide Walker’s controlled image environment. Figure 1 and Figure 2 show acoustic ray traces in the horizontal and vertical planes of the control room. Figure 1. Ray trace of the left monitor in the horizontal plane. Figure 2. Ray trace of the left monitor in the vertical plane. 2.2 Reverberation Times When the requirement for ultra-low reverberation times (RT) that are constant with frequency from 30 Hz to 15 kHz was combined with the available depths for sound absorption, we found that it was not possible to implement the 1 m (approximately) deep absorbers that are often used in non-environment rooms. The maximum available depths for absorbers were 500 mm for the rear wall and sections of the ceiling and 200 mm for the side walls, and the computed RTs with resistive absorbers with these depths was too high at low frequencies. To achieve a lower reverberation time at low frequencies, resonator absorbers were required. Bespoke low-frequency Helmholtz absorbers were designed using the transfer matrix method described by Colam and Leembruggen [5] and Cox and D’Antonio [6]. As part of designing these absorbers, we measured in our laboratory [7] the flow resistivity of the fibrous insulation that would be installed behind the perforated panels. Vol. 44. Pt. 3. 2022 Ay at AE M Ty us MY a ae Proceedings of the Institute of Acoustics Examples of the absorption vs frequency characteristics of the absorbers we designed are shown in Figure 3. In addition, measurements of example low frequency absorbers that we made in our impedance tube showed that a gap of approximately 50 mm was required between the perforated cover and the insulation behind it for the absorber to function correctly. The impact of this gap appears to be much stronger at low frequencies than at mid frequencies. Figure 3. Examples of sound absorption characteristics of the low frequency absorbers. An unavoidable consequence of using Helmholtz absorbers is their reflection coefficients are unity at higher frequencies, rendering the room vulnerable to unwanted reflections at high frequencies. To minimise the risk of problematic reflections occurring at the prime listening positions, the Helmholtz absorbers were located so that reflections from them would be directed into broadband resistive absorbers. As this was not possible in some areas, the Helmholtz absorbers were interleaved with broadband absorbers to provide some scattering resulting from impedance discontinuities of the disparate surface types. A 700 mm deep Helmholtz absorber is located beneath the rear seat. 2.3 Monitoring Loudspeakers We recommended the Genelec 8361 monitors ( https://www.genelec.com/theones t o JMC Academy for the control room; based on the following attributes of this loudspeaker: • Fully concentric design • Published directivity and flatness of frequency response extending to 20 kHz. • Low frequency cut-off of 30 Hz JMC procured a sample of the loudspeaker, and we measured its frequency response and confirmed that the manufacturer’s data was correct. Listening to the loudspeaker indicated that it had a remarkably accurate and uncoloured sound both on an off-axis. As noted above, to optimise the smoothness of the low frequency response of the loudspeaker, we desired to create a half-space acoustic environment (free of image sources) for the loudspeakers by flush-mounting them into the front wall of the room. However, the 8361 loudspeaker is a vented-box type with the vent on the rear face of the loudspeaker enclosure, and that location of the vent created a strong design challenge of how to resolve the two conflicting situations. We elected to design a bespoke enclosure in which the left and right loudspeakers would be housed, allowing the low-frequency sound from the rear vent to be transmitted through another vent in the front face of the enclosure. This design of this enclosure is discussed in Section 5. Genelec sell such an enclosure for the smaller model 8351, which we extensively measured and listened to, however concluded that it was too lossy and slightly too resonant to copy and enlarge to suit the 8361. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics 2.4 Room Structure Figure 4 shows a plan view of the control room and identifies various acoustic functions of the walls. Figure 5 shows the elevations of the two side walls and identifies the surface finishes, while Figure 6 shows the layout of the various acoustic elements on an exploded room plan. To produce a degree of scattering in those areas with flat Helmholtz absorbers (that ideally should have deep absorbers), the Helmholtz absorbers were interleaved with broadband absorbers so that sound wave sees discontinuous surface impedances. 500 mm deep broadband absorption Solid half space wall environment for loudspeakers These areas absorb reflected sound from the front wall, and limit the extent of reflections from the monitors Low frequency and broadband absorbers Figure 4. Plan view of the control room. Figure 5. Elevations of the two side walls of the control room. The light areas are solid timber, the striped areas are perforated timber, and the dark areas are fibrous insulation behind perforated metal. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics With its sawtooth-like sides that were required to produce a controlled image design, the room has taken on attributes that Newell attributes to Wolfgang Jensen’s designs in the 1970s. The perforated timber on the angled side walls helps to provide “life” to conversations between staff and students during teaching sessions. oe Figure 6. Layout of surface finishes in the room. Figure 7 and Figure 8 below show images of the completed room. When the photos were taken, the monitor loudspeakers had not been installed in their enclosure that would allow the rear vent sound to be delivered to the room. Vol. 44. Pt. 3. 2022 me ES ‘Broaand Type A100, 1x 10mm Marin XHDI00 bind 6 mm p.m sheet 40% Open rea and 24 mmole am) ‘roaband Type 8210, 2 100m Marin XHD100 bain 1 mn pr. metl het (40% Open Ne a2 mm le am) ‘roadand Type 6-00. 5 100 mm Marl M10 behind 1.6mm port. metal seo (10% Open Awa and 2.4mm le dam) ‘roaand Type D0, 1x50 mm Man XHOSO boi 1.6mm pert mea seat (0% Open Ara and24 mm ole ar) LoweFeg Type A50, 18 mmtmber aco (0.8%4 Operon 8m hole am) non of $0 mm avy and 450mm Marri MOD absorption Low Feg Type 8200, 18mm timber (0.7% Opendea, 8 male lam) nett 50 mm a caviy an 150 mm Martel XHOS0 absopion \LoweFeg Type 200, 30m inber(8 mm + 12mm) (08% OpenAen Ben ole am) in ont of 0 en a aut ar 150mm Marin XHO80 absorption Low re Typ 20,18 mn ier (3.8% Opentea, mle am) n Hott 5 mma cavey and 501m Mari XHOSO absepon ees ns nna 70 1 tne Cpe. 8 ann tute am a cata 60 mm Mai MDS» 350 Heavy plasteoard/2x 16 mm MO 16mm MOF te Proceedings of the Institute of Acoustics Figure 7. Image of the frontal part of the room. Figure 8. Views of the front wall (left image) and the right-rear corner (right image). 3 LISTENING IMPRESSIONS Although the science and architecture associated with the room are vital elements, the sound that is produced in the room is much more important. Before we look at the measured acoustic performance of the room, which is not perfect, it is worthwhile to review comments about the room provided by the audio engineer and teacher who had carriage of the project. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics Tom Payne writes: I've been lucky enough to listen to what are considered some of the premium studios in the country and I can confidently say that Studio A control room is the best listening experience I have ever had. The soundscape is extremely immersive and there is a level of depth in the phantom centre image that I have not heard before. The phantom centre meshes beautifully into the L and R signals, it really does feel like the whole front wall is one giant sound-source. Instrument panning and localisation can be critiqued from an area far outside of what would be considered the ideal central listening position which is invaluable in an educational setting where multiple students work across the width of the console. The accuracy of the listening space paired with the Genelec loudspeakers selected by Glenn’s team has immediately increased what is possible when it comes to critical listening. The studio is constantly revealing new information in music I am very familiar with. The control room design and speaker selection have removed the need for a subwoofer when working in a music production setting. To me, the most important part of this is how well and how evenly the room supports the low frequency energy, especially in the time domain. 4 ACOUSTIC PERFORMANCE Given Tom’s favourable review of the room presented above, how well does the room measure? 4.1 Impulse Responses Figure 9 and Figure 10 show the impulse responses of the left and right monitors at the central mix position filtered to bandwidths of 1 kHz to 10 kHz and 565 Hz to 4525 Hz. Walker’s criterion is within a hairs-breadth of being met, when considered from these perspectives. RChMonitor_MixPos (LP+HP, 1000 - 8000 Hz) RChMonitor_MixPos (LP+HP, 1000 - 10000 Hz) WinMLS Pro 0 -10 -20 -30 -40 -50 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Time [ms] Figure 9. Filtered IRs at mix position over range1 kHz to 10 kHz . RChMonitor_MixPos (BP, 1600 Hz, 3 oct.) LChMonitor_MixPos (BP, 1600 Hz, 3 oct.) 5 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 -55 WinMLS Pro 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Time [ms] Figure 10. Filtered IRs at mix position over the range 565 Hz to 4525 Hz. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics Figure 11 shows the impulse responses of the left and right monitors at the central mix position filtered to a bandwidth of 141 Hz to 1131 Hz. LChMonitor_MixPos (BP, 400 Hz, 3 oct.) RChMonitor_MixPos (BP, 400 Hz, 3 oct.) 5 WinMLS Pro 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 voit BES REREES YH -50 -55 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Time [ms] Figure 11. Filtered IRs at mix position 141 Hz to 1131 Hz . 4.2 Reverberation Times. Figure 12 shows the Schroeder decay curves of the left monitor loudspeaker at the mix position; as the right monitor is very similar, it is not shown. Figure 12. Schroeder decay plots of the left monitor in octave bands. When the reverberation times (RTs) are very short, the use of one-third octave filters can introduce an element of inaccuracy into the computed reverberation times due to the settling time of the filters, and there may be some inaccuracies in the RTs shown. Figure 13 shows the RTs filtered in 1/3 rd octave bands at the mix position, computed with filtering applied using backward integration. The RTs of a perfect impulse in third-octave bands are also shown for reference. gesesese? Figure 13. Reverberation times in 1/3 rd octaves at the mix position of the left, centre and right monitor loudspeakers. The RTs of a perfect impulse with the 1/3 rd octave filters are also shown. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics The measured reverberation time at low frequencies corresponds closely to the suggested decay time for music control-rooms for reliable and accurate monitoring as suggested by Toyashima and reproduced from [8]. Figure 14 below presents the graph from Toyashima shown in [9]. Figure 14. Suggested decay time for music control-rooms for reliable and accurate monitoring (from Toyashima). 4.3 Cumulative Energy Build-Up Table 1. Rise times to -3 dB point of octave band filters 4.3.1 Intrinsic Rise-Times The cumulative energy plot is a useful measure of the pattern by which the sound energy a measurement position builds to a steady state. If that analysis is carried out in different bandwidths, it is important to recognise that the filters will also have their own build- up time (i.e. rise time). To illustrate this, Figure 15 shows the build-up (or rise-time) of an ideal impulse when filtered into octave-wide frequency bands (6 th order filter slopes). Table 1 lists the rise times to a level of -3 dB relative to the steady-state level. When viewing the cumulative build up plots, the rise-time of the filters must be kept in mind, but note that the rise times of the combination of the room/loudspeaker combination and the filters are not necessarily fully cumulative. Rise time to - 3 dB Frequency (Hz) Time (ms) 63 20 125 10 250 5 500 2.5 1000 <2.5 2000 <2.5 4000 <2.5 8000 <2.5 Normalized Filtered Time Data WinMLS Pro 5 Ref bb IR (BP, 63 Hz, 1 oct.) Ref bb IR (BP, 125 Hz, 1 oct.) Ref bb IR (BP, 250 Hz, 1 oct.) Ref bb IR (BP, 500 Hz, 1 oct.) Ref bb IR (BP, 1000 Hz, 1 oct.) Ref bb IR (BP, 2000 Hz, 1 oct.) Ref bb IR (BP, 4000 Hz, 1 oct.) Ref bb IR (BP, 8000 Hz, 1 oct.) Ref bb IR (BP, 16000 Hz, 1 oct.) 0 -5 -10 -15 -20 -25 -30 -35 100 90 80 70 60 50 40 30 20 10 0 -10 -20 Time [ms] Figure 15. Cumulative build-up of an ideal impulse when filtered into octave-wide frequency bands. 4.3.2 Measured Cumulative Energy Figure 16 and Figure 17 show the cumulative energy in octave bands of the left and centre channel loudspeakers at the central mix position. As the right channel is very similar to the left channel, it is not shown. The rise time of the centre channel at high frequencies is prolonged by a flutter echo that was completely unexpected. Vol. 44. Pt. 3. 2022 Roverberation ime (sec) Frequency (Hz) ioe Proceedings of the Institute of Acoustics Left+ EQ centre mix pos (BP, 63 Hz, 1 oct.) Left+ EQ centre mix pos (BP, 125 Hz, 1 oct.) Left+ EQ centre mix pos (BP, 250 Hz, 1 oct Left+ EQ centre mix pos (BP, 500 Hz, 1 oct.) Left+ EQ centre mix pos (BP, 1000 Hz, 1 oct.) Left+ EQ centre mix pos (BP, 2000 Hz, 1 oc Left+ EQ centre mix pos (BP, 4000 Hz, 1 oct.) Left+ EQ centre mix pos (BP, 8000 Hz, 1 oct.) 0 WinMLS Pro -5 -10 -15 -20 -25 -30 100 90 80 70 60 50 40 30 20 10 Time [ms] Figure 16. Cumulative energy build-up of left channel monitor at the mix position in octave bands. Centre ch mix pos centre helmholtz reflection (BP, 63 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 125 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 250 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 500 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 1000 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 2000 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 4000 Hz, 1 oct.) Centre ch mix pos centre helmholtz reflection (BP, 8000 Hz, 1 oct.) 0 WinMLS Pro -5 -10 -15 -20 -25 -30 100 90 80 70 60 50 40 30 20 10 Time [ms] Figure 17. Cumulative energy build-up of centre channel monitor at the mix position in octave bands 4.4 Waterfall Spectrograph Decays Figure 18 shows the waterfall decay plots of the left and right monitors at the central mix position. Left channel Figure 18. Waterfall plot of left loudspeaker at the mix position. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics Right channel Figure 19.Waterfall plot of right loudspeaker at the mix position 4.5 Wavelet Spectrograph Decays 4.5.1 Wavelet Resolution A plot of temporal spectral behaviour which spans a wide range of frequencies usually has a time resolution that either too low at high frequencies or too high at low frequencies. A 100 ms window, for example, produces a 10 Hz frequency resolution, which at low frequencies is a large fraction of an octave (1/1.4 octaves at 20 Hz), while at high frequencies it produces a tiny fraction of an octave (1/1386 octaves at 20 kHz). More useful is a frequency-dependent trade-off between time and frequency resolution, implemented using a constant fraction of an octave for the frequency resolution, rather than a constant number of cycles. This trade-off provides a higher time resolution at high frequencies and lower at low frequencies. A constant Q wavelet transform achieves that, specifically a Continuous Wavelet Transform (CWT). A constant Q wavelet transform is mathematically equivalent to using a frequency-dependent window to produce the spectrogram. However, it is important to recognise that the use of the wavelet introduces an additional decay time into the results, which depends on the length of the wavelet. Figure 20 shows a spectrograph decay plot of an ideal impulse using wavelets with an equivalent frequency resolution of one sixth octave. Table 2 lists the associated decay times. Table 2. Times for ideal impulse to decay to -10 dB with 1/6 th octave resolution. Decay time to - 10 dB Frequency (Hz) Time (ms) 30 35 50 22 100 11 200 5 400 <5 800 <5 1600 <5 Figure 20. Wavelet decay of an ideal impulse to -20 dB with 1/6 th octave resolution. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics 4.5.2 Wavelet Spectrograph Decays Figure 21 shows the wavelet spectrograph decays of left and right channel monitors at the mix position. On a frequency-by-frequency basis, these plots show that the ITU and Walker criteria are only met out to 14 ms (shown with the dotted line), with reflections at 1.2 kHz and 2.5 kHz arriving directly after that time. The source of these reflections was not investigated. However, the presence of the mixing console and outboard effects rack introduce surfaces that we cannot control. Left channel Right channel Figure 21. Wavelet spectrograph decays of left and right channel monitors at the mix position. 4.6 Frequency Responses The loudspeakers were equalised using a process combining spatialised acoustic measurements, computations of filter parameters and critical listening. As the left and right loudspeakers direct sound over the meter bridge on the top of the console, that structure introduces inevitable acoustic diffraction Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics into the measured responses. As the centre channel loudspeaker directs sound downward onto the console, there is a strong reflection from the console reaching the listener. Given that the human ear is a highly sensitive receiver and that a flat frequency response is not always what is heard when reflections are intrinsic to the sound, we elected to use each calculated equalisation filter to suggest possible changes to the response and allowed our listening to be final arbiter of whether specific adjustments improved the sound. Boost equalisations were used only with wide bandwidth with one band per loudspeaker being available. Figure 22 shows the responses at the central mix position of the left and right loudspeakers, while Figure 23 shows the response of the centre channel loudspeaker. The time window is 130 ms with a half-Hanning window The responses of the left and right loudspeakers show evidence of the diffraction over the meter bridge and a reflection from the floor between the loudspeaker and mixing console. The response of the centre channel shows strong evidence of the comb filters introduced by the refection from the console. Figure 22. Frequency responses of left and right monitors after equalisation. “0 le 4 = Bia 4 PATE Vy Woy sy yoy zl, _\) fw 254 Vi Vv aT acl 30| au 7,000 Frequency (H2] 10,000 Figure 23. Frequency responses centre monitor after equalisation. Vol. 44. Pt. 3. 2022 Proceedings of the Institute of Acoustics 5 LOUDSPEAKER ENCLOSURE As noted above, the desire to locate the loudspeakers with a rear facing vent into a half-space acoustic environment introduced some design challenges. A design was developed using a specific volume behind the loudspeaker and two large openings in the baffle above and below the loudspeaker. The openings were deemed, without much thought, to be sufficient to not substantially affect the response of the loudspeaker. Figure 24 shows the cross-sectional view of the enclosure. However, technical hubris is often accompanied by a downfall and the measured frequency response of 8361 loudspeaker installed into the enclosure showed that the enclosure significantly degraded the low-frequency response of the loudspeaker. Figure 25 shows the measured response and with the problem area highlighted. a 500 mm Section 8-8. with rubber edging ‘around speaker 2 16mm boards ‘glued & screwed fogether 6mm MOF support shelt withstand 35x35 timber support brace to reduce shel vibration ‘eames ketion Figure 24. Sectional view of enclosure to provide half-space environment for the loudspeaker. Figure 25. Measured frequency response at 400 mm from the loudspeaker mounted in the enclosure. As substantial investigation using trial and error with the loudspeaker enclosure to understand the cause of the poor response did not yield a way forward, we elected to develop an acoustical- mechanical equivalent circuit for the system and predict the response of the system. That equivalent circuit is shown in Figure 26. Vol. 44. Pt. 3. 2022 AY Fence Proceedings of the Institute of Acoustics Figure 26. Equivalent acoustic-mechanical circuit of Genelec loudspeaker in the external enclosure. Based on the stated -3 dB frequency of the Genelec 8361 speaker and the external size of the enclosure, we estimated a set of loudspeaker driver and Genelec enclosure parameters that produced a Butterworth response with a – 3 dB point at 30 Hz. The acoustic parameters of the external ports and the enclosure volume were combined with an estimated loss factor for the external enclosure. The predicted response of the as-designed system is shown in Figure 27, which bears similarity to the measured response in Figure 25. Using the model, we developed a solution to ameliorate the problem, which consisted of blocking up one of the two front ports, reducing the internal volume of the enclosure and installing a large amount of high-density insulation into the enclosure to damp the resonance. The resulting response is shown in Figure 27, which also bears similarity to the predicted modified response. We attribute the rise in the measured response around 30 Hz to 40 Hz to the additional loading produced by the effective curvature of the front wall. Response of Genelec 8361A in bespoke vented enclosure I : i I —Mocies 6 oer eat owt zurt soe ever eet sor v6 vse ve ost 909 19 065 evs cor oat ase ost oe ow frenweney go AA A/a - 7 Figure 27. Predicted response of Genelec 8361 speaker in original and modified external enclosure. Figure 28. Comparison of original and remediated responses measured at 400 mm on axis. Vol. 44. Pt. 3. 2022 Seneec port and enclosure coneoupt ie Proceedings of the Institute of Acoustics 6 CONCLUSION A control room for JMC Academy in Sydney was recently designed and commissioned by Acoustic Directions. The room was intended to be true a non-environment room, but the available room dimensions limited the depth of low frequency sound absorbers and required the monitor wall to become a controlled image design. However, even with these compromises, the room achieves a very high degree of performance in relation to the following parameters: • Impulse responses at frequencies above 500 Hz are reduced to - 20 dB after 5 ms. • Reverberation times of approximately 180 ms to 150 ms are relatively constant with frequency. • Highly linear Schroeder decay plots for frequencies 63 Hz to 8 kHz. • Rise times for frequencies 250 Hz to 8 kHz are less than 8 ms to a level of 2 dB below the steady state level. • At frequencies above 250 Hz, the spectrograph decays are mostly 20 dB below the steady state level at up to 19 ms after the direct sound arrives. • The spectrographs and waterfalls show only very minor decay ridges at low frequencies below 150 Hz. • The longer decays below 30 Hz are likely due to the group delay of the loudspeaker. • The frequency responses at the mix position of the left and right loudspeakers are very flat but show the effects of comb filtering due to unavoidable reflections and diffractions from the mixing console. JMC staff have provided comments about the quality of sound in the room that are most heart- warming. Based on those comments, we conclude that this room that could not be made fully non- environment due to constraints is sufficiently non-environment in its performance to provide the same sound quality that the principal author first heard in one of Newell’s non-environment rooms in Portugal. 7 REFERENCES [1] P. Newell, Recording Studio Design, Focal Press, 2003 First Edition. [2] R. F. Allison, “The Influence of Room Boundaries on Loudspeaker Power Output,” JAES, vol. 22, 1974 June.. [3] G. Adams, “Time Dependence of Loudspeaker Power Output in Small Rooms,” JAES Vol 37 No 4 April 1989. [4] R. Walker, “A New Approach to the Design of Control Room Acoustics for Stereophony,” Preprint 3543 94th AES Convention1993 Berlin. [5] S Colam, G Leembruggen, “A Computational Method for Analysis and Design of Acoustic Absorbers and Low Frequency Transmission Loss,” Proc. IOA Vol 25 Part 8 2003. [6] T. Cox, P D'Antonio, Acoustic Absorbers and Diffusers, Theory, Design and Application, Spon ISBN 0-415-29649-8, 2004. [7] C. Hanrahan-Tan, G Leembruggen, “Development and Use of a Low-Cost Acoustic Flow Resistance Meter,” Proc IOA Vol: 42 Part: 3 2020. [8] P. Newell, K. Holland, “High Performance Studio Rooms for Simple Domestic Construction,” Proc. IOA, Vol. 43. Pt. 2. 2021. [9] K. H. a. P. Newell, “High Performance Studio Rooms for Simple Domestic Construction,” Proceedings of the Institute of Acoustics, Vol. 43. Pt. 2. 2021. Vol. 44. Pt. 3. 2022