Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

CONTROL ROOM DIAGNOSTICS

Santiago Iturburu, Lucas Fernandez Garabenta, Pedro Venezia


santiturburu@gmail.com, lucasrafaelfg@gmail.com , venezia46238@estudiantes.untref.edu.ar
Universidad Nacional de Tres de Febrero, Ingeniería de Sonido, Caseros, Buenos Aires, Argentina

This study presents the diagnostic of a control room frequency response, where an average of different
positions and a scanning method are used. While using fixed positions can give the response in particular
spots of the room and phase information, the scanning method shows a similar result as the averaged
measures regarding frequency response, but lacks of phase information and the capability to recognize
individual modes. The analysis provides insights into the room’s frequency response, the presence of
modes, the impact of listeners positioning, and the potential of filtering techniques to achieve a desired
frequency response. The primary limitation of the analysis is the unavailability of dismantling the
absorptive panels, resonators, and diffusers, making it challenging to formulate accurate proposals for
modification.

1. Introduction
In this article, the acoustical measurement and analysis of a recording studio control room are conducted
with the objective of evaluating the system response and proposing possible modifications and solutions if
needed. The studio being measured is Mosco Studio, located in the middle of Buenos Aires City. The
article will discuss measurement techniques, analysis methods and programs, characteristics of the monitor
system, and the acoustical properties of the room and its treatment.
The main parameters that will be discussed are the frequency and phase response of the room and
loudspeakers as a whole system. It will be explore which filters should be used to achieve a better response,
how this system interacts with minimum phase, early reflections, and modal analysis, among others. The
intention is always to understand how these factors affect the perception of the listeners in the room.

2. State of Art
Regarding to how control rooms should sound in order to minimize any change in the perception of
the recorded sound, Phillip Newell details apparently every thing to consider in his book Recording Studio
Design [1] . At chapter 11, Newell indicates the importance of the interaction between the loudspeakers and
the room, and how the room reflections and resonances contribute to the overall response perceived by the
listeners. In addition, he points out what to consider when applying correction of the frequency response
by equalisation and how this is related to minimum or non minimum phase situations.
In their paper [2] , Fazenda, Holland and Newell, highlights the low frequency time decay as a fundamental
aspect when it comes to direct sound perception. They explain that there is sufficient evidence to support
the hypothesis that the perception of resonant problems is significantly associated with the time domain
response and not just with the frequency domain response of the system. Furthermore, it appears that any
resonance that exceeds the modes detection threshold, given by it decay time, would be undesirable even
though it could contribute to the overall flatness of the room’s amplitude frequency response.

1
Acoustical Instruments & Measurements. 17/10/2023
3. Theoretical Framework
Acoustics have played a crucial role throughout human history and culture, enhancing speeches in both
informative and religious contexts. The manipulation and understanding of sound have been employed to
maximize the impact and effectiveness of public speaking. The implementation of stages, amphitheaters,
and churches served as practical solutions to improve auditory perception. These acoustical phenomena
were recognized and utilized, despite the lack of a comprehensive physical explanation at the time. The
development of dedicated music chambers and venues, along with the observations made by musicians
and investigators regarding various parameters, led to significant advancements in acoustics. This progress
encompassed physical discoveries, mathematical equations, and theories, shaping the modern theoretical
analysis of acoustics.
Control rooms are born as a special type of room, since its design must give the best listening condition
for someone listening through loudspeakers. Using only loudspeakers as the source is an important condition,
since the idea is to listen to something happening somewhere else or that happened sometime in the past (or
both). This means that the perfect listening condition exists without the interference of the actual control
room. If the control room or the loudspeakers system affects in some way the sound program, then the
ideal condition fails. It is ideally expected to listen the same signal as it was when recorded. To fulfill this
expectations, the control room must be acoustically invisible.
Of course it is impossible to carry out such an ideal room, but some may get somewhere near with some
considerations. It is very difficult to achieve this criteria in every spot of the room; however, doing it in only
one spot, usually called the sweet spot, can be done. This way the focus goes only in one small volume of
the room. To accomplish this, control rooms nowadays must "handle" at least three phenomena: the room
specific modes of resonance, the reverberation time and the early reflections.

3.1 Modes
The acoustical modes of a room are the stationary reflections that arise from the rooms geometry. Those
frequencies where the wavelength have a natural relation with the dimension of the room will produce
stationary modes, just like resonances in any other wave system. The main problem in this frequencies will
be the abrupt difference of level, depending the position inside the room. Pressure nodes and maximums will
appear, changing drastically the listening condition depending the position. This phenomena in noticeable
up to a frequency where the modal density is not large enough, and it can be obtained through the Schroeder
formula (Equation 1), where T60 is the reverberation time or 60 dB, V is the volume of the room and
FSchoroeder is the aforementioned frequency.
r
T60
FSchroeder = 2000 (1)
V
For a control room, the optimal scenario is the non existence of modes. However, this might be difficult
to achieve, so another solution is controlling only those that bothers at the sweet spot. The best way to
handle modes is to absorb them. Due to its long wavelengths not any foam-type absorbent will do the job.
That is why resonators are used, such as Helmholtz or membrane ones. In a rectangular room, equation 2
gives the frequencies of all the possible modes
s 
2  2  2
c x y z
fxyz = + + (2)
2 L W H
Where c represents the speed of sound in meters per second, L is the room length, W is the room width
and H is the room height. All dimensions are in meters.

3.2 Early Reflections


Early reflection are those first order reflections that still maintain high levels in relation to the direct
sound due to short travelled distances and hard reflective boundaries. For these first order reflections, the

2
Acoustical Instruments & Measurements. 17/10/2023
level difference with the direct sound comes from the absorption of the reflected surface and the extra
distance travelled.
Avoiding or attenuating early reflections is crucial in a control room since they are easily noticeable and
significantly impact the original sound introducing frequency coloration and modifying the time-related
aspects of the sound. Additionally, imbalanced reflections between ears can present a serious problem in
achieving an accurate stereo image.

3.3 Reverberation Time


The reverberation time of a room is defined as the time that takes for the sound pressure level generated
by an impulse to decrease 60 dB. It is usually known as T60.
For a control room, the T60 must be shorter than the typical decay time found in musical programs. For
example, if the room where a recording is made has a certain amount of reverb, or if the audio has a certain
amount of reverb due to external processing, the control room must have a shorter reverberation time to
avoid any impact on the perception of the program’s original reverberation.
This means that the T60 in a control room depends on what type of musical program will be listened. In
general, with an average T60 of 300 ms or less, assuming the room absorption treatment is evenly distributed
in the frequency range, with a natural decrease over frequency, making the T60/frequency curve smooth,
this conditions is satisfied.

3.4 Minimum Phase


A minimum-phase system is one in which the phase shift associated with the amplitude response is
the minimum that can be allowed while still exhibiting the properties of a causal system (one in which the
output never arrives before the input). Since there is a strict relationship between amplitude and phase in
such systems, correcting either one will inevitably tend to correct the other.
Non-minimum phase responses are those in which amplitude corrections do not tend to correct the
phase disturbances. An example of a non-minimum phase effect is the far-field response of a loudspeaker
in a reflective room. In this scenario, a delay occurs between the signal generation by the loudspeaker and
the superimposition of the boundary reflections on the overall response. The arrival times of reflections
create frequency and distance (time) dependent phase irregularities. Therefore, simple manipulation of the
amplitude response of the source cannot adequately compensate for these complex disturbances.

4. Measurement Procedure
The measurement setup comprises of two Behringer ECM-8000 microphones, one Beyerdynamic MM1
microphone, a Focusrite Scarlett 4i4 preamp and interface, and a commercially available laptop for data
collection. To assess the transfer function between the microphones, measurements were performed, revealing
minimal variation of less than 1 dB across all frequencies.
The software tool Aurora was used to generate and process a sine sweep signal, which served as the
excitation source for the room. Additionally, an inverse filter was applied to derive the impulse response of
the system. The resulting room impulse responses (RIRs) were further processed using the REW software.
Various analyses, including waterfall plots, frequency responses, and more, were conducted using the
information displayed within the REW program.
The objective was to characterize the control room, for which six possible listening points where
measured, one of them being the sweet spot. Figure 1 shows the mic and the speakers distribution in
the room. The ISO-16283-1 [3] was taken into consideration for this measurement, respecting each distance
between mics and surfaces. The six points are averaged to get the average response of the room on all
optional listening positions.

3
Acoustical Instruments & Measurements. 17/10/2023
3.50

1.10 1.20
1.70
1.90 M5 1.50
1.80

1.03 2.80 2.00


1.10
Position Height
M4
M1 1.2
M2 M2 1.1
4.70 1.40 M3 1
M4 1.9
M1 Sweet
M5 1.5
1.80
Spot 3.50 M6 1.7

M3
1.03 2.50
M6
1.90

2.10 1.20
1.00
1.00
0.80
2.70

5.00

Figure 1: Diagram depicting the positions of speakers and microphones on the plan. All distances are in meters,
following the SI.

To add information about the data displayed in the Figure 1, the studio monitors model is Neumann
KH310D, they are positioned at 1.25 m, and there is a big Soundcraft MH3 mixing console. There is also a
subwoofer Neumann KH805 positioned in the floor in front of the sweet spot.
After collecting the sine sweeps in all six points, a different method is used. The scanning microphone
method from the ISO-16283-1 is used to collect an average response of the room. This method is usually for
measuring the average sound pressure level in a room to later determine the isolation between two adjacent
rooms. In this case, a transfer function is applied, in other words, the difference between the pink noise after
and before passing through the system is calculated to get the systems response, and by using the scanning
microphone technique several points of the room are averaged in the same calculation.
Once all measures are processed and both the frequency and phase response are gathered, a target curve
is established as the reference, aiming to achieve a flat frequency response. To attain this target curve,
an automated filter technique is employed using the Room EQ Wizard (REW) software. The software
identifies and determines the appropriate filters that need to be applied for reaching the desired target curve.
This approach is employed in the Sweet Spot signal, the averaged signal from six points, and the noise
captured through the transferred scanning microphone technique.
By applying the aforementioned filters that result in nearly flat curves, it becomes possible to extract
the minimum phase characteristics. Specifically, in regions of the spectrum where the system’s minimum
phase exhibits constant group delay (referred to as linear phase), the system can be considered as minimum
phase. The phases of all three curves are compared to assess whether the frequency bands with minimum
phase exhibit uniformity or if they are affected by any acoustic phenomena.
In the Figure 2 are presented several photos of the control room showing the first three measurement
positions.

4
Acoustical Instruments & Measurements. 17/10/2023

(a) Rear view. (b) Front view.

(c) Sweet Spot. (d) Sofa mic positions.

Figure 2: Photos of the analyzed control room.

5. Results and Analysis


The frequency response of the Sweet Spot is depicted in Figure 3. The observed frequency range
spans from 40 Hz to 16 kHz, exhibiting a considerable difference of approximately 17 dB between the
maximum and minimum levels. However, it is worth noting that certain frequencies display spikes, which,
if appropriately addressed, could result in a narrower range of approximately 10 dB. Notably, prominent
spikes are observed at 68 Hz, 190 Hz, 1.4 kHz, and 5.6 kHz, deviating significantly from the average
response.
According to Equation 1, the Schroeder frequency is around 150 Hz. However, for that equation to be
valid, the room must be reflective enough to generate a diffusive field. In this case, it will be taken into
account the 150 Hz given by Schroeder’s equation, but adding an extra bandwidth of transition between
both statistical and model worlds, getting up to 300 Hz (2 · FSchoeder ).
It is suspected that both 68 and 190 Hz are modes. In fact, λ68 /2 is 2.5 m, exactly half of one axial
dimension, corresponding to a room mode. The same goes for 187 Hz, where λ187 /2 is 0.9 m, or 1/5 of 4.6,
another axial dimension. Whether these modes exhibit maximum or minimum intensity, and the specific
level of their intensity, depends strictly on the positioning of the listener and the sources, relative to the
axis associated with the respective mode. It is when the modal density becomes large enough that this level
differences are no longer perceived.

5
Acoustical Instruments & Measurements. 17/10/2023

Figure 3: Frequency Response in the Sweet Spot

Figure 4: Frequency Response comparison

When comparing various locations, this modal behavior can be observed, with the peak at 190 Hz
corresponding to the mode (0,5,0). This can be seen in figure 4 where the response at the sweet spot is
plotted with other one at different height ("standing sweet spot").
As seen in figure 5, the microphone position used to measure the sweet spot had the wrong angle
regarding the loudspeakers high-end way. Hence, since the directivity of the speakers increases, the band
from 3 to 8 kHz is not correctly captured by the mic, so the level in that band decreases. This can also be
seen in Figure 4, where the microphone was at a different height, placing it in another angle with respect to
the vertical axis of the loudspeakers and measuring a flatter response in the mentioned band.
On the other hand, the strong drop in 1.4k Hz can be happening for either two different reasons (or both).
First, the loudspeakers system directivity 4. Only by moving 20° off-axis the level in 1.4 kHz decreases
almost 12 dB, as shown in Figure 15 in the annex. This is suspected to be due to the crossover frequency of
two of the ways of the system (Low and Mid). Graphs of the directivity contour and SPL off-axis are shown
in the Anex in Figures 14 and 15. In both cases a drastic change in directivity performance is evidenced.

6
Acoustical Instruments & Measurements. 17/10/2023
Second, a reflection in the mixing console might be summing out of phase with the direct sound, generating
not only a lower level, but also group delay. This intuition becomes understandable through the presentation
of the Figure 5.

Figure 5: Sweet Spot view.

A probable solution is to find the axis determined and recommended by the system manufacturer and
align both speakers exactly to the chosen sweet spot area. However, the systems directivity issue with 1.4
kHz makes this area narrower for that specific frequency.
The reverberation time, shown in the Table 1, is as expected for a control room of this characteristics
and uses. It is almost constant around 300 ms for all frequencies higher than 100 Hz. Regarding lower
frequencies, the RT60 increases up to 500 ms. This is due to the resonant behaviour of the sub-woofer, in
addition to the modes already described. The energy decay is seen in the waterfall in Figure 7. To solve this
low-end issue a more detailed modal analysis must be done to understand weather its the sub-woofers fault
or an acoustic mode.

Table 1: T30: Average of the eighteen measurements (six points and three repetitions)

Frequency band 63.5 125 250 500 1000 2000 4000 8000 16000 [Hz]
Sweet Spot T30 547 286 219 247 309 338 358 319 251 [ms]
Sweet Spot Deviation 322 022 025 004 004 008 007 007 005 [ms]
Room Average T30 525 281 237 268 312 349 355 330 264 [ms]
Room Average Deviation 262 040 020 016 013 009 008 010 013 [ms]

7
Acoustical Instruments & Measurements. 17/10/2023

Figure 6: T30 estimation for the sweet spot measurements and the average of all measurement points.

Figure 7: Waterfall showing the frequency response over time.

Although the RT60 is correct in practice, and shows the treatment is well done, it might generate conflict
when measuring the room acoustics. Always the best way to measure a room is before and after the
treatment has been done, to see the difference and know if and where the treatment was effective. In
this case, it is impossible to do so.
The scanning microphone technique and the average of all six points are shown in Figure 8. This
graph evidence the similarity in both results, and, therefore, concludes that the scanning method can be
used to measure an averaged frequency response of the room. One setback is the phase, given that moving
the microphone around ruins the phase reference for the measurement for obvious reasons, leaving only
the frequency response alone. On the other hand, it also disregards or does not explicitly address modal
behavior, as it integrates numerous points within the motion technique itself. While one approach could

8
Acoustical Instruments & Measurements. 17/10/2023
involve partitioning the circle into multiple semicircles and quantifying the deviation in low frequencies,
this method offers limited insights into modal analysis.

Figure 8: Scanning Method and Averaged Frequency Response

The sweet spot response and the filter used to flatten the curve are shown in Figure 9.

Figure 9: Sweet Spot and Filter Frequency Response

The predicted curve after applying the filter can be seen in Figure 10. The deviation from the target
curve is minimal, with a maximum difference of only 1 dB. This indicates that the response is closely align
with the desired frequency response outlined by the flat target curve, meaning the phase linearity stipulates
where the system is a minimum phase one. To simplify this analysis the group delay is shown in Figure
11. From 1k up to 18k Hz the group delay is almost constant, so it is correct to say the system (room and
loudspeakers as a whole) is of minimum phase. Below 1k Hz this is arguable, since the ripple down to 400
Hz might be the 1 dB level deviation. Finally, the low-end below 400 Hz definitely tends to no constant
group delay and isn’t minimum phase.

9
Acoustical Instruments & Measurements. 17/10/2023

Figure 10: Filtered Sweet Spot Frequency response and phase

Figure 11: Filtered Sweet Spot Group Delay

It is interesting to observe in the following Figures 12 and 13, the filtering proposal of REW using the
average response for all the measurements and the scanning method response.

10
Acoustical Instruments & Measurements. 17/10/2023

(a) Average frequency response and filters proposed by REW

(b) Filtered average response and target curve.

Figure 12: Filtering the Averaged Response for all measurements

The version of REW used to obtain these filters and predicted curves differs from the one used to obtain
the filters for the Sweet Spot analysis. In this case, the software only applies subtractive equalization, which
requires the target curve to be set below the minimum level of the desired response in order to process it.
This is in contrast to the version used in the Sweet Spot analysis, which is capable of performing both
subtractive and additive equalization.

11
Acoustical Instruments & Measurements. 17/10/2023

(a) Scanning Method Frequency Response and filters proposed by REW

(b) Filtered Scanning Method Frequency Response and Target curve.

Figure 13: Filtering the Scanning Method Frequency Response.

6. Conclusions
The analysis of the room’s frequency response and acoustical characteristics revealed several key findings.
The frequency response in the sweet spot exhibited a range from 40 Hz to 16 kHz, with a significant
difference of approximately 17 dB between the maximum and minimum levels. However, certain frequencies
displayed spikes, notably at 68 Hz, 190 Hz, 1.4 kHz, and 5.6 kHz, deviating significantly from the average
response.
The measurements indicated the presence of modal behavior, with the 68 Hz and 190 Hz frequencies
identified as possible room modes. The sound pressure levels (SPL) of these frequencies depended on
the positioning of the microphone and sources relative to the associated axis. Additionally, the monitor’s
position and angle are not optimized for use with the determined sweet spot, affecting the capture of
frequencies in the range of 3 kHz to 8 kHz due to their directivity.
The reverberation time (T30) was consistent with the expectations for a control room in a recording
studio, with values around 300 ms for frequencies above 100 Hz and increasing to 500 ms for lower
frequencies. The presence of resonant behavior from the subwoofer and acoustic modes contributed to
the longer decay times in the low-frequency range.
The scanning microphone technique proved to be a viable method for measuring the averaged frequency
response of the room. However, it lacked the ability to accurately capture phase information and identify
individual modes. The filtering process applied to the sweet spot response successfully achieved close

12
Acoustical Instruments & Measurements. 17/10/2023
alignment with the desired frequency response, indicating minimum phase behavior for frequencies above
1 kHz.
In summary, the analysis provided insights into the room’s frequency response, the presence of modal
behavior, the impact of listener positioning, and the potential of filtering techniques to achieve a desired
frequency response. Further investigations could focus on addressing the identified issues, such as aligning
the speakers according to the manufacturer’s axis and examining the subwoofer’s contribution to low-end
behavior. Repeating the measurements after applying the proposed filters would be an interesting way to
continue this investigation.
As mentioned, a comprehensive analysis of the acoustic treatment in the room is lacking due to the
practical inability to remove the absorptive panels, cloud, resonators, and diffusor for evaluating the room’s
response with and without different components of the treatment. Additionally, an analysis of the loudspeakers’
response can be conducted through a nearfield quasi-anechoic measurement. These endeavors would enable
the recovery of valuable information to make informed decisions, provide well-founded recommendations,
and offer the client more accurate and effective solutions.

References
1. Newell, P. R. (2003). *Recording Studio Design*. Taylor & Francis. English.

2. Fazenda, B. M., Holland, K. R., Newell, P. R., & Castro, S. V. (2005). The Time Domain Performance
of Standard Listening Rooms: An Assessment of Current Rooms and Recommendations for Achieving
Improved Compatibility. *Journal of Acoustics*, *27*(Pt. 5).

3. International Organization for Standardization. (2014). ISO 16283-1:2014: Acoustics - Field


measurement of sound insulation in buildings and of building elements - Part 1: Airborne sound
insulation. English.

4. Spinorama. Neumann KH310A. Retrieved from https://www.spinorama.org/speakers/


Neumann%20KH%20310A/Neumann/index_vendor-gll.html.

5. Official Neumann website. Retrieved from https://www.neumann.com.

13
Acoustical Instruments & Measurements. 17/10/2023
7. Annex

Figure 14: Contour graph of the Neumann KH310A.

Data courtesy of Spinorama [4] and Neumann [5] .

Figure 15: Frequency response off axis.

14

You might also like