Magic Bus / Measurements

A Bruel & Kjaer 2270 Handheld Sound Meter, mounted on a custom fixture, was used to measure the reverberation spectra inside the Magic Bus. After the balloon was popped, it acquired sound decay data at the rate of 400 samples per second. The reverberation times reported are the times necessary for the sound to decay to one-millionth (-60 dB) of its original intensity, but were derived from decay data from -5 dB to -25 dB, and extrapolated to -60 dB using B&K's powerful software.

A calibrated Earthworks M23 measurement microphone, mounted on a custom fixture that positioned the tip of the microphone between the listener's ears, was used to measure the frequency response inside the Magic Bus. Literally hundreds of sine sweeps were sent through the loudspeakers to obtain optimal crossover topologies, equalization, and time alignment. The final measurements were derived by a combination of objective analyses and subjective listening. The process of objective tuning and subjective evaluation was repeated a multitude of times until legendary audio component designer, Steve McCormack, was satisfied.

The photos below show the fixtures in use.
   
cache/wst.opf.2880826.xml
The images below show the results of the measurements.
Figure 1. Reverberation spectra.

Figure 1 shows reverberation time, in seconds, as a function of frequency, in Hertz ("Hz"). The reverberation time, by definition, is the time it takes sound to decay 60 dB. In practice, it is difficult, and sometimes impossible, to measure 60 dB of sound decay. Typically, a portion of the sound decay (10, 20, and 30 dB drops, for example) is measured then mathematically extrapolated to 60 dB. In the case of my measurements, the "T20" method was used. The Bruel & Kjaer Handheld sound meter determined the reverberation times for frequencies ranging from 50 to 10,000 Hz in 1/3-octave intervals. The gray lines on the plot define the upper and lower limits for reverberation time, according to the International Telecommunication Union ("ITU"). Not only did I measure the spectra of reverberation times in my Sprinter (indicated by black solid circles), but also for other vehicles, including a BMW X5 (dashed yellow line), a VW Beetle, (red crosses), and a Jaguar X-Type (green triangles), all for the sake of comparison. For further comparison, I also obtained published data from Robert Harley's personal listening room (blue diamonds).

First, I'd like to discuss the reverberation times in what represents a typical selection of automobiles. The plot clearly shows that all typical automobiles have insufficiently short reverberation times (about 0.1 seconds) throughout the midrange and treble, that are roughly half the magnitude required to comply with the ITU guidelines. This would give typical automobiles a "dry" lifeless sound, and little indication of spaciousness. The plot also reveals problems with the bass in typical cars; notice how the reverberation times below 100 Hz rise rapidly. The disproportionately long reverberation times in the bass, relative to those in the treble and midrange, are undesirable, and lead to sonic degradation. 

Next let's discuss the reverberation times in the Magic Bus. Notice how the spectra of reverberation times are essentially within the guidelines set for by the ITU. If one wishes to be critical, it is true that there is a small peak at 250 Hz, a small dip at 160 Hz, and another slight dip between 500 and 630 Hz, but these are minor imperfections when compared to the gross deficiencies of conventional automobiles. Also notice how the reverberation times in the bass rise only modestly relative to those in the midrange and treble. I'm quite pleased with the outcome of my acoustical treatments. My preliminary listening impression validates that they are as important as any component in my audio system, and have resulted in a dramatic improvement in the sound quality. With regard to the value and effectiveness of room treatments, Robert Harley summed it up best, "If given the choice between listening to $40,000 worth of preamp and monoblock amplifiers and no room treatment, or a $2,500 integrated amp and the ARS package, I would choose the lesser electronics in the treated room. Room acoustics are that important, and the ARS package is the ultimate realization of the goal of minimizing the room's effect on music reproduction". Although reverberation time is only one important measure of a room's performance, I'm delighted the reverberation times in my Magic Bus essentially comply with ITU's guidelines and compare favorably to not only Robert Harley's extraordinary listening room containing treatments valued at $33,000, but also to some of the finest recording studios.
Figure 2. Frequency response for left (blue) and right (red) channels.

Notice how the frequency response diverges at higher frequencies from the left channel - this was an intentional result derived from reading dozens of research papers about the HRTF ("head-related transfer function") and its relevance to the asymmetric listening position, typical of most automobiles. At frequencies below about 624 Hz, the wavelengths of sound are much longer than the dimensions of a human head, and the inter-aural intensity between the left and right channels must be identical, while the inter-aural time arrivals must be precisely tuned. Conversely, above 624 Hz, the inter-aural intensity begins to play a greater role, especially at frequencies much smaller than the dimensions of the human head. The Magic Bus features superb intra- and inter-channel time alignment along with supremely smooth frequency response to achieve unrivaled subjective stereo imaging.

 
Figure 3. Crossover points and slopes for left channel.

The tweeter, midrange, woofer, and subwoofer responses are shown in yellow, blue, magenta, and red, respectively. The overall response is also shown in blue. Notice how the acoustical slopes are all 4th-order Linkwitz-Riley, and the crossover points reside at "textbook perfect" points 6 dB lower in magnitude than the summed response. 
Figure 4. Step responses in the time domain for left (blue) and right (red) channels.

The first small spike is the tweeter, followed by midrange, followed by the woofer, followed by the subwoofer. Notice how each driver "hands off" to the next driver lower in the frequency range, which results in superb frequency domain integration. This is classic textbook 4th-order Linkwitz-Riley crossover design. 
Website Builder