Loudspeaker design is about making choices and compromises. To make the choices intelligently, the application must be clearly defined. Once defined, design parameters for accomplishing the goal need to be determined and prioritized. Without strick definitions and explicit prioritization, the system will be confused and second rate. Compromises are unavoidable. Every decision to go in one direction involves giving up something else. Discipline is necessary, or else less critical factors could be put ahead of more critical ones causing unnecessary compromises.
Following is a list of loudspeaker design principles placed in prioritized order. It begins with the basic requirements of the system, then proceeds from the most audible factors, those most important to get right, to the less audible and less critical factors.
- 1) Application : The design parameters for the system must be defined first by the intended application, such as home stereo, recital hall, auditorium, stadium, music, speech, sound reinforcement, etc.
- 2) SPL : Sound pressure levels required of the system, background music, realistic acoustic music levels, high level amplification, etc.
- 3) Bandwidth : Speech has the narrowest bandwidth requirements; music is much wider, hearing range is approximately 20 Hz-20 kHz. Lowest note on the piano is 27.5 Hz, some organs may go down to 16 Hz, although infrasonic frequencies are felt rather than heard (we hear only the harmonics). Realistic minimum range for music is from 30 Hz to 16 kHz flat. (Flat to those points consequently provides extension higher and lower.)
- 4) Linearity : For realistic reproduction of music the bandwidth amplitude must be linear. This will not make a speaker great, yet it is the foundation. Both experienced and inexperienced listeners are clearly able to recognize and prefer a flat, full bandwidth speaker, as evidenced by double blind studies. (See link below.) Deviations in linearity are easily audible as too bright, too dull, or just off the mark.
- 5) Low Distortion : Harmonic and intermodulation distortion need to be as low as possible. Speakers produce the greatest amount of distortion compared to any other component in the audio chain, and not just marginally. It's by orders of magnitude greater. Speakers are, without exception, the single most important part of an audio system—they produce the acoustic energy that our ears receive. In the midrange and treble distortion is readily audible when exceeding 1%.
These are the first principles for high fidelity, realistic audio reproduction. Other parameters important for creating the illusion of live music are finer points that will distinguish the outstanding from the merely good.
- 6) Resolution : Acoustic resolution is comparable to image resolution. The higher the resolution of an image, the more detailed and lifelike it looks. To get high resolution in audio, the drivers must be capable of responding quickly to the electrical signal in order to resolve all the detail (start and stop instantly). They must not resonate (ring), nor 'rounding off' transients (leading & trailing edges, attacks and cut-offs). No less important are cabinet resonances. Resonances obscure resolution. Some are from the internal waves reflecting back through the drivers, some from the cabinet walls. Cumulative Spectral Decay (CSD) is a way to measure this form of distortion (resonant behavior) of the driver or cabinet. The higher the Q of a system, the more it stores energy and resonates. Stored energy in the driver and cabinet continues to vibrate after the signal has stopped. Ported and passive radiator designs use this stored energy to boast bass levels, and consequently sacrifice transients and resolution. The accumulation of these added vibrations blur, smear, or muddle the sound and reduce clarity. Another contributor to the loss of resolution is a passive crossover and its interaction with the amp. A speaker needs to be a reproducer of the audio signal, not a resonator. For these reasons ported bass reflex (and passive radiator, bandpass, transmission line, horn), and passive crossover configurations are unacceptable compromises because they add their own sound to the audio signal.
- 7) Dispersion : A broad, even, linear dispersion pattern, the polar response, and the power response are critical for two reasons. First is for the listening position. Limited dispersion necessitates being in exactly the perfect spot to get the full spectrum that the speaker is capable of producing. Sit a little taller, slouch, move to the left or right, and the tonal balance changes. Listeners sitting off-axis won't get the full sound. Second, nonlinear off-axis dispersion effects the power response—a measure of the total power output relative to frequency. The reverberant or ambient soundfield created by the speaker's interaction with the room will consequently be out of balance with the direct (first arrival) sound if the off-axis response is nonlinear. The reflected sound should mirror the direct sound so that the soundfield in the room has the same tonal balance. Realism and neutrality is greatly improved with a linear off-axis response. This is because our perception of tonal balance is a function of the sum of the direct and reflected sound. Very few manufacturers ever mention the off-axis polar response in their marketing, instead touting some other single feature far out of proportion to its actual significance. Poor polar/power response is a major weakness of most speakers and is one of the reasons why most sound like boxes.
- 8) Waveform Fidelity : The most contentious parameter. No doubt, the goal of high fidelity is to perfectly reproduce the original waveform. Unfortunately, waveform distortion begins with the recording process and continues through every step of the entire audio chain. The biggest culprit, on the reproduction end, is the speaker and its crossover, especially passive crossovers. The psychoacoustics of phase and time incoherence, and of how the brain interprets music, must be taken into account. Our focus should be put on the parameters that are most easily recognized by the ear as effecting the perceived fidelity. Studies have shown that our brains are insensitive to the distortions induced by phase discrepancies. We are still capable of interpreting the audio cues even when the waveform, as seen on an oscilloscope, is distorted by phase. It is a wonder how this is possible. It is no less a wonder that we can discern individual instruments out of an entire symphony orchestra while they are all playing at the same time. That we can tell the difference between a single piano note or two notes played an octave apart is no less amazing. Why, for instance, do our ears not separate the individual harmonics of a single note, and yet do separate two unison notes played simultaneously? Since music becomes a single waveform, how do our brains distinguish individual instruments? Brain scans using fMRI have shown that sound is processed independently in several different areas of the brain. The signal is essentially deconstructed with different aspects of the sound, contour, rhythm, pitch, timbre, etc., being processed in different areas and then 'reassembled' to produce our perception. This reassembly puts the pieces back together based on our previous experience and expectations. It is certainly a lofty goal to attempt keeping the waveform intact and in absolute phase, but the brain has tools to circumvent this if the signal is in relative phase. There is one more thing that is constantly overlooked by time & phase coherency advocates. Time/phase shifts will result in a nonlinear frequency response. A linear response is the only way to achieve tonal balance and it will be, by default, [relative] phase correct. To place this parameter higher up the priority list at the expense of other more audible distortions is just a form of marketing hyperbole and a costly endeavor that sacrifices fidelity rather than preserving it.
a quote from Manny Carrubba, audio recording engineer, The Plant Studios
We tend to concentrate our attention on the axial response of the speaker as a matter of test convenience and simplification. And in fact, the direct sound of the speaker is very important. If there are response problems on axis, it will not be an excellent sounding speaker, period. What is frequently overlooked, also as a matter of engineering convenience, is the off-axis response, the extreme off-axis response and the overall power response of the loudspeaker.
Conventional loudspeakers have off-axis response curves that are increasingly rolled off as we move off-axis around the loudspeaker. Worse, because of the different directivity patterns of the individual drivers, most loudspeakers have increasingly lumpy response curves as we move around to the side of the speaker. All of this lumpy low-pass sound is emitted into the room.
Conventional wisdom says that directional loudspeakers and rooms that damp or diffuse early reflections are good. Usually, all that is being done with such a treatment is to add even more low pass filtering to the "lumpy low-pass" reflected sound and to the room tone in general. The loss in high-frequency information particularly hurts the localization of phantom images and phantom reverberance [sic] cues, as well as darkening the overall perceived timbre. However, with a loudspeaker that does not have the limitations in its dispersion that conventional speakers suffer from, accurate lateral reflections are maintained, yielding better images, particularly of ambience, and a brighter, more open, more spacious and natural range of timbres.*
* http://www.moultonlabs.com/more/making_music_sound_good/P3/ © 2006 David Moulton
References :
|
