The Soffit Simulation
The Sound Field
Further Discussion
A Final Word
The Soffit Simulation: Next: The Sound Field
The JBL 4412 acoustic pink noise response at the simulated listening position in our control room at Greenledge Studios on May 23, 2005, HIGH=-4 MID=+0, measured using PRAXIS @ 1/24 octave, 65536 points. Raw data obtained using a calibrated Beyer MC740N Microphone set to Omni mode (green trace), and a beyer M88 unidirectional microphone (pink trace). The notch in the response at about 120 Hz is phase cancellation caused by the reflection from the front wall (specifically the wall behind the speakers). Ostensibly this notch vanishes with flush mounting of the speakers.
The response on the whole is excellent, with the exception of a possible issue at 90 Hertz (the notch at about 90 Hertz may be a ceiling reflection/phase cancellation). These notches are comb filter effects which create a series of notches with a uniform frequency spacing at higher frequencies as well. In this case, 90 Hertz corresponds to a wavelength of 1130/90 = 12.55 feet, which means a 6.275-foot path difference, corresponding to 6.275/1130 = 5.55 milliseconds. Higher frequency cancellations are 1130/6.275 = 180 Hertz apart, or 270, 450, 630, 810, 990 etc... Examination of the response curve at these frequencies shows distinct notches in the regions 270, 630 and 990, with 450 a null result. Further examination may reveal more.
The peak at about 47 Hertz (the front-to-back-wall natural resonance) is significantly problematic because of its relative size especiallly since it is a characteristic of the room design. Low-frequency absorbers or diffusers are quite space-consuming in this range of frequencies (about 24 foot wavelengths mean 6 foot traps or diffusors). The resonant frequency of the rear (east) wall may be about 37 Hertz which already may be helping to dissipate this mode. The front wall design may be tuned more precisely to dissipate energy in this range, using the west walls as low-frequency panel absorbers. At this frequency we must be attentive to other possible ways to make a reduction. A possible solution is to choose woofer positions to correspond to a comb filter effect beginning at 47 Hertz (e.g. 6 feet from the floor). This may be far from practical though. We may be facing an entirely different situation when the side walls are built, and it would be wise to exercise considerable caution in the design of the south wall and north cupboard, not to mention the speaker positions relative to these (side) walls.
The Sound Field: Next: Further Discussion
It is important to note that the amplitude of the 47 Hertz peak depends rather strongly on the position of the listener. In the front-to-back center of the room (see the graph below) there is no real problem at 47 Hertz (just a small peak). Since the listening position for soffit mounting is closer to the center than for the simulation, this graph may be what happens. However, when other listening positions are involved (a desk further east, perhaps), then these are real room mode issues which may be dealt with, for example (as discussed previously) in wall construction. The 60 Hertz notch (seen below) is likely a floor reflection or absorption of that frequency by the west wall, or both. The wall could be finished so as to add just enough weight to trim the wall resonance to 47 Hertz, cancelling the 47 Hertz room resonance, if it's not too much to hope for!
Ideally, what we desire for this studio is a slightly rolled-off pink noise response, meaning that higher frequencies should gradually fall off in a linear fashion. Specifically, in the bass region (i.e. below 300 Hertz), it looks like we may benefit significantly from the 6dB bass boost which the soffit would provide (the notch at about 140 Hertz caused by the front wall reflection is eliminated with a soffit mount). The result achievable with a soffit is, cautiously, rather good, likely better even than a number of commercial studios. This is very exciting news and speaks in favour of the soffit in our case. Further measurements, calculations, and conclusions are warranted, however, since the cautious elimination of potential issues can greatly reduce the severity of future problems. But, YES! Without the notch at 140 Hertz and with a slight bass boost, this looks pretty flat, maybe within plus or minus 5 dB! But remember, caution is a wise course.
The JBL 4412 acoustic pink noise response at the front-to-back center of our control room at Greenledge Studios on May 24, 2005, HIGH=-4 MID=+0, measured using PRAXIS @ 1/24 octave, 65536 points. Raw data obtained using a calibrated Beyer MC740N Microphone set to Omni mode (green trace), and a beyer M88 unidirectional microphone (pink trace).
Further Discussion: Next: A Final Word
Mounting the JBL 4412's in the front wall and flush with it (soffit mounting) addresses the front wall reflection, and depending on the height of mounting, it may alter the floor and ceiling reflections in terms of of where cancellations occur in the frequency range. The low frequency range is a cause for concern in any mastering studio, where predictable low-frequency response is essential to the consistency of the sound mastered on one system and in one acoustic environment and later played on a wide range of consumer sound systems and in a variety of acoustical environments. Bass is particularly problematic because of the high energy of the low frequencies which makes them difficult to absorb by simple room treatments. These considerations are better dealt with in a room's design than by remedial room treatments. With Jehovah's help we will consider this further.
Higher frequency problems may be improved by incorporating serious diffusion into the front (west) wall, either with or without Helmholtz absorption (Helmholtz resonators are actually very space-saving diffusors as well as narrow band absorbers, which could be used in the bass region to quell undesirable room modes). Alternative ways exist to deal with high frequency problems also, such as graphic equalization. Adjustable acoustics may be provided in the form of track-hung sonex panels and even specifically tuned resonator panels of the same thickness. Adding equipment to the room should be done with forethought, since this affects the acoustic. There is little you can do to improve on a well-designed room, but even equipment affects a room's character by reflections and bass diffractions. Rooms which were designed using the principle of a reflection-free zone (RFZ) are adversely affected by equipment placed behind the console. Diffusion at the monitor end of the studio, including the side walls, may create early sound scattering (ESS), and thereby an accurate, reliable stereo soundfield.
Whether a soffit-mounted monitor system is provided, or whether the speakers are placed on stands in the room at carefully chosen positions is really a matter for prayer. One further note about flush mounting: the bass in a flush mounted speaker achieves 6dB of bass boost due to the fact of all the energy radiating forwards rather than half of it backwards. (It is only the bass that is affected in this way, because bass is omnidirectional by nature of its large wavelength. Keep in mind that, in general, in an acoustic environment, smaller wavelengths behave as particles and larger waves behave as waves. You can't hide from bass, whereas treble is easily deflected by smaller objects.) Soffit mounting is much more costly because of the need to reinforce the walls for solid sound projection. If free-standing speakers turns out to be our choice, soffit mounting may be reconsidered at a later date.
A Final Word: Back to Top
It has been a privilege to share these thoughts with you. In the words of the Lord, "He who believes in me will do greater works than mine." (Jesus' words, John 14:12)
Read the latest poem