Ceiling Boundary Ambience Enhancement (tm)

by Dick Olsher (5/99)

 1. Introduction

The primary goal of CBAE (tm) technology is to enhance the stereo listening experience. The problem with two-channel audio is that both the  recording's direct sound and ambience are projected from a plane between the speakers. All sound appears to come from the same distance as the loudspeakers. There is no sense of being enveloped by the sound. This is highly  artificial because in a live venue, reflections arrive with a range of time delays and with a specific tonal balance. To enjoy a similar sensation in the home, the proper soundfield must  be recreated at the listener's ears.  Surround sound technologies are quite powerful in reproducing a feel for the original acoustic because they literally surround the listener with front and rear speakers. CBAE (tm)  technology provides a means of integrating two  audio channels into a domestic environment so that they flesh out an optimum soundfield. The soundfield components in a listening room consist of:

• direct sound
• early reflections (< 10 ms)
• late reflections (10 - 30 ms)
• reverberance (>30 ms).

Reflections convey both distance and spatiality information to the auditory system. It is well known that distance perception is a function of the  direct to reflected sound ratio (1). However, researchers at Aalborg University in Denmark (2) have recently investigated the effect of "fine structure" in room reflections and have shown that the presence of early  reflections actually improves distance perception within the soundstage. Barron (3) and Barron and Marshall (4) have investigated the subjective effects of late lateral reflections in concert halls by experimenting with simulated reflections in an anechoic chamber. Their major finding was that lateral reflections, delayed by at least 10 ms were crucial to obtaining a desirable spatial impression. Early reflections (< 10 ms) were found to cause image shift. Ceiling reflections were not as effective, but appeared to elevate the perceived image. By spatial impression is meant the feeling of "being in a three-dimensional space." Marshall (5) includes a good description of spatial impression or ambience from the manager of the Concertgebouw Orchestra in Amsterdam, who explained that it was "the difference between feeling inside the music and looking at it, as through a window." As Barron (3) explains, spatial impression is very different from reverberation (time delay >80 ms) which tends to provide a certain degree of   envelopment in the sound, but mainly gives an impression of distance from the source.

Research has also shown that strong early reflections are never desirable: they can cause image shift, coloration, and loss of clarity. This argues for diffusive side walls and wall treatments such as RPG gratings. The crucial question is then how to harness room reflections constructively so as to increase the palpability of "being there" without loss of clarity. CBAE does this by generating a unique soundfield at the listening seat, in terms of both the tonal balance and time signature of reflected energy. CBAE designs delay room reflections in the critical upper midrange by at least 10 ms relative to the direct sound, while producing a uniform power response up to a frequency of about 1 kHz.

It would be wise to heed Arthur Benade's guidance (6) in the matter of reflected energy: "… in the real world of practical music, the room plays  a role that is so important we will call it essential. Far from being  confused and misled by the complexities of room, the player and the listener find it very difficult to operate comfortably unless they have the room as part of the signal path. Musicians absolutely hate to play outdoors, and they have been known to refuse to play in an anechoic chamber: they like and need to have help from the room itself." This is also the case for the playback of music in the home. It's a shame that the mindset that "all reflections are bad" has motivated some designers to consider highly-directional (high Q) speakers as ideal for domestic environments. This is consistent with the paranoia of many audiophiles about reflected energy. The obsession with excessive room damping and the elimination of room reflections has turned some listening rooms into virtual anechoic chambers. The basic idea in both cases is to make the speaker's on-axis response the dominant factor at the listening seat. I have never enjoyed listening in such an environment.

2. The Old Paradigm

Because frequency response of loudspeakers in a living room is a complex three-dimensional problem, most loudspeaker designers still follow the  traditional engineering paradigm:

• Consider the speaker strictly as a transducer
• Reference design to free-field conditions
• Use an anechoic chamber (or semi-anechoic measurement - e.g., MLSSA) for evaluation
• On-axis frequency response is the primary figure of merit.

Clearly, in the low-frequency range (below 200 Hz), it is impossible to predict a speaker's in-room response just on the basis of its power response  or directionality. The position of the speaker in the room creates large differences in bass balance. There are very few room modes to excite below 200 Hz, which makes bass response very uneven and room dependent. Floyd Toole's  past research at the National Research Center in Canada  has indicated that room placement impacts listener preference more so than even the speaker model. However, above 200 Hz it was shown by P.H. Chapelle as early as 1973 (7)  that the traditional on-axis frequency response does  not give a good approximation of the spectral balance of a speaker at typical listening positions in a living room and that a speaker's power response is more predictive of  the spectral balance at the listener's ears.

 Although audio magazines continue to make a living off on-axis frequency response measurements, it is actually well established that speakers that  measure identically on-axis in an anechoic chamber sound much different in a  typical domestic environment. These differences are due to how these speakers radiate sound off-axis as a function of frequency. Even at a listening  distance of only 8-feet, the soundfield at the listening seat is already  more than 50% room reflections in the bass and midrange. Speakers whose power response is deficient in the midrange will color the soundfield in the same  direction. For example, a two-way design with an 8-inch woofer crossing  over to a 1-inch dome at 3 kHz, may measure fine on-axis (anechoically), but show a 10 dB dip in the midrange off axis. Such a speaker will sound recessed  and lifeless in a real-world room.

 The late and great James Moir writing in Wireless World (8) raised the notion of an "optimum polar distribution for a domestic  speaker." He pointed out that dipole radiators such electrostatics will sound rather dry in some rooms, especially those with short reverberation time. He also rejects omni-directional radiators as "almost equally  unacceptable," because the stereo image is "diffuse and only vaguely located." He suggested that there may be some optimum "distribution of sound energy in front of a loudspeaker if the stereo image is to be  well defined and the sound quality is to be 'soft and non-tiring' to the  listener." CBAE (tm) technology implements just such an optimum directionality. The basic idea is to mimic the power response of a musical  instrument: omni-directional to about 2 kHz and becoming progressively more directional with increasing frequency.

3. The New Paradigm

If the old design school has a motto, it must be: "if I can't measure it, I can't hear it." To paraphrase the late Richard Heyser, these  are the same folks who would attempt to explain to a musician that his guitar amplifier cannot possibly produce good music because it has more than 0.001% harmonic distortion. The new paradigm focuses on the fact that the end  product of audio is the listening experience, which involves perception and cognition. The ability to decipher the music's emotional content or to immerse oneself in the original acoustic is valued more highly than static   specifications. The new paradigm mandates that the room and the listener be placed into the design equation. One of my speaker evaluation tests is to listen to a speaker from outside of the listening room. Does it give the  impression of a live musical instrument? How believable is the impression? Of course, if a speaker fails to energize a room in a similar fashion to that of a musical instrument, it fails the test. After all, live music fills  a  room and spills out in a way that we all instinctively recognize as the real thing.

Thomas Edison was the first to use speakers to stand in for musical instruments by placing them on stage as one would acoustic  instruments. For this  to work, for a speaker to imitate life, it must emulate the power response of real instruments. This is the cornerstone of Ceiling Boundary Ambience Enhancement(tm). The speaker is regarded as a  soundfield reproducer and is  expected to reproduce both the recording's direct sound as well as at the correct tonal balance and time signature of its ambience. Dr. Amar Bose (9) was the first designer to concern himself  seriously with recreating a concert  hall acoustic in the listening room. His work lead directly to the Bose model 901 loudspeaker design in which 89% of the sound energy is radiated toward the rear and side walls. This 8:1 ratio of reflected to direct sound  energy was derived on the basis of concert hall measurements. However, this attempt to squeeze a concert-hall acoustic into a living room generated far too much early reflected energy that  diffused image outlines. In addition  to the tonal balance of reflected energy, it is critical to consider the time signature of room reflections. Spatiality is in the fine structure of these reflections.

 4. The New Technology

My own research over the past three years has highlighted the importance of late reflections in a domestic listening room in enhancing the  palpability of the soundstage and the sensation of being  there. The engineering issues of how to achieve the required time delay with two-channel stereo and over what frequency bandwidth are at the heart of CBAE (tm)  technology. As implemented in the Magic Cube and Ichiban loudspeakers, the ceiling boundary is used as a reflective surface in order to delay and fan out sound energy to the sidewalls over a two-octave bandwidth (1 kHz - 4 kHz). The woofer/midrange is oriented toward the ceiling. Just about all living rooms possess a hard, reflective ceiling at a height of 8 to 12 feet off the floor. Using the path length delay -- 1 ms for each foot of travel -- sound energy reflecting off the ceiling is automatically delayed by at least 10 ms relative to the direct sound. On-axis, relative to the listener, dispersion is controlled in the upper midrange to avoid early reflections that would smear image outlines and cause timbre colorations. The direct sound in the range from 1 kHz to 4 kHz is slightly recessed by 1 to 3 dB (as measured under anechoic conditions) but is progressively filled in by reflected energy from the ceiling and sidewalls. It is important to note, however, that the measured in-room frequency response is flat on-axis.

Below about 1 kHz, CBAE (tm) speakers are omni-directional radiators by virtue of listening to the woofer/mid at 90-degrees from its main axis. Both  acoustic and electrical means are used to contour the directional  response of the speaker so that it's omnidirectional up to 1 kHz, and progressively more directional up to 20 kHz. The intent is to give a reflected sound  balance that rolls off with increasing frequency. A room curve that does not roll off in this manner gives an overly bright impression.

The speaker's overall power response mimics the way an acoustic instrument radiates sound into a room. Another advantage of this topology is that  woofer breakup modes, which are higher-order harmonic products, behave much like treble energy and are beamed along the woofer's main axis -- away from the listening seat. Hence, there is less distortion reaching the listener  and musical textures are purer and smoother sounding.

The most striking external feature of CBAE (tm) speakers is the ceiling-firing woofer. As such, it bears a superficial resemblance to previous  designs that have used this sort of topology. However, no other reflected/direct design incorporates the unique power response of CBAE (tm) speakers.

For the record, the ceiling-firing woofer originated with Korn (10) in the 50s. He cited two specific  advantages for this configuration: (1) the true  360-degree distribution of the direct sound in a horizontal plane, independent of frequency, and (2) increasing the ratio of the reverberant to the direct sound at mid and high frequencies. The basic idea was also embraced by  Stewart Hegeman in the 60s, and even Roy Allison (11) is a recent advocate.

To properly implement the CBAE alignment requires special drivers that meet specific acoustical  criteria. The tweeters are state-of-the art  dual-magnet designs with silver voice coils and a very low free-air resonance. Both the Magic Cube and Ichiban loudspeakers use rigid aluminum coned woofers capable of true pistonic motion within their operating range.

5. References

1. J. Blauert. "Spatial hearing," The MIT Press (1983).
2. J. Michelsen and P. Rubak. "Parameters of Distance Perception in Stereo Loudspeaker Scenario," 102nd Audio Engineering Society  Convention, March 1997.
3. M. Barron. "The Subjective Effects of First Reflections in concert Halls -- The Need for Lateral Reflections," Journal of Sound and  Vibration, vol.15 (4), 475-494 (1971).
4. M. Barron and A. H. Marshall. "Spatial Impression Due to Early lateral Reflections in Concert Halls: The Derivation of a Physical  Measure," Journal of Sound and Vibration, vol.77 (2), 211-232 (1981).
5. A. H. Marshall. "A Note on the Importance of Room Cross-Section in Concert Halls," Journal of Sound and Vibration, vol.5, 100 (1967).
6. A. H. Benade. "From Instrument to Ear in a Room: Direct or via Recording," Journal of the Audio Engineering Society, vol.33 (4) (1985).
7. P. H. Chapelle. "The Frequency Response of Loudspeakers: On-axis Response or Power Response?," 44th Audio Engineering Society  Convention, February 1973.
8. J. Moir. "Speaker Directivity and Sound Quality," Wireless World, October 1979.
9. A. G. Bose. "On the Design, Measurement, and Evaluation of Loudspeakers," 35th Audio Engineering Society Convention, October 1968.
10. T. S. Korn. "A Corner Loudspeaker with Coaxial Acoustical Line," Journal of the Audio Engineering Society, vol.5 (3) (1957).
11. R. Allison. "Imaging and Loudspeaker Directivity: To Beam or Not to Beam," 99th Audio Engineering Society, October 1995.