Now that we are all happily reading—as I hope—The Absolute Sound’s Illustrated History of High End Audio, Volume One: Loudspeakers, it is natural to think of speakers under review in historical context. In capsule form, speaker design in the era since the introduction of the long- playing record has gone through these stages: the box speaker era of the 1950 and 1960s; the planar, both electrostatic and planar-magnetic, era of the late 1960s and the 1970s and 1980s; and then in recent decades, perhaps rather surprisingly, an era of return to box speakers, albeit floorstanders rather than stand- mounted or bookshelf models of the earlier days. Of course there have been exceptions all along. But it is striking to note how much the later chapters of the book are about floorstanding box speakers.
Two things happened here to cause this return to boxes. One was that dynamic drivers improved and improved a lot. There were fine drivers in the past, but they were hard to make. Cones had to be “doped” by hand, tweeters had to be laboriously pair- matched, and so on. These problems are largely something of the past.
A second development was the increasing use of exotic cabinet construction. The BBC designs and their offspring had dealt from decades ago with the issue of cabinet sound via their damped thin-wall enclosures, starting at least as early as 1968 with the Spendor BC1/BBC LS3/6. But the industry, especially in the U.S., began more recently to be obsessively interested in deadness of cabinets. Some of the result involved truly heroic measures—the Rockports come to mind as do the Wilsons, the Wilson line having arguably initiated the trend.
But a funny thing happened on the way to dead cabinets.
People began to observe, some people anyway, that dead cabinet speakers sounded to them, well, dead. The BBC had already pointed out that increasing rigidity pushed such resonances as there were up in frequency. Their approach was to use relatively flexible cabinets but to damp the resonances, which were at rather low frequencies and thus were less audibly objectionable. In effect, one had to match the amount of rigidity with the right kind of damping. Rockport’s constrained-layer damping was another effective approach used later on where rigidity was higher than in the BBC designs. But outside of those theoretical considerations, the idea somehow reappeared that in musical terms there might still be some reason to allow the cabinet to be less damped and to participate in the sonic picture more.
This is not an easy matter for an outside observer to consider definitively. After all, one can hardly take a review sample, remove the drivers, put them in a different cabinet, and go from there. What is not hard to check is that cabinets matter. The Microscanner devices to absorb cabinet energy (one of the tweaks that actually did something—no plastic dots of negligible mass but a real device to divert vibrational energy into non-audible forms) really did change the sound. Opinions on whether they changed it for better or worse might have varied from person to person, but change it they did. Nonetheless, a definitive estimation of the effects of cabinet sound in any one speaker—that is hard to get.
Like the Sony AR1s that I reviewed enthusiastically in Issue 214, the GamuT Superior RS5s use the cabinet as part of the sound. In fact, in my estimation they do this even more than the Sonys. Like the Sonys, they have a wood cabinet, rather than a particle-board one, with the idea being that the natural resonance of the wood will be integrated into the sonic picture. And the natural resonance is given what seems to be quite free play here in the GamuTs. Rigidity enthusiasts have tried to claim that cabinet sound is always distortion, but of course there is no reason why this should be so if cabinet sound is done correctly. The approach is not unreasonable—just tricky to pull off. No one wants to go back to the buzzy cabinets of old table radios—and no one is. The idea of letting the cabinet participate sonically is serious, whether one agrees with it or not. And plenty of serious people are pursuing it.
There is a certain difficulty involved in analyzing what is going on here because one’s basic reaction to a speaker is dominated by frequency response and radiation pattern considerations— things which are only indirectly connected to the way the cabinet is made, though the shape of the cabinet has definite effects. So all one can really do is to describe the speaker’s “first-order” behavior—frequency response and radiation pattern—first, and try to infer what secondary effect the cabinet has on the rest.
Perhaps it does not really matter what causes what. Only the final sound is at issue for the consumer. But people should once and for all disabuse themselves of the idea that total absence of cabinet sound of any sort is some sort of Platonic ideal. An open mind is desirable here, as elsewhere, this situation being rather complex.
What the GamuTs Try to Do and Why
Lars Goller, the designer of these speakers, was very forthcoming about what he was aiming at, in the cabinet design in particular. Let me try to summarize for you what he said as I understand the picture. The set of ideas embodied in the cabinets of the GamuT arose from a general program, to some extent developed jointly with DALI’s Hotter Bak, of looking at what limits the resolution of speakers. The first round of this general investigation by DALI some time ago led to the idea of having drivers that avoided stick/slip friction at low levels. (My write-up of it for TAS can be read here: regonaudio.com/Tact%20and%20Dali. html.) The next and more recent stage was to look at cabinet effects from the same viewpoint. What was the cabinet doing to low-level signals? In particular, as the sound died away, was the cabinet preserving the decay correctly?
In principle, damping does not make this go wrong—not if damping remains linear at low levels. A linear damping process changes phase and frequency response but would not alter the correct decay behavior as such. According to Goller, however, research determined that, in fact, damping was non-linear when the levels became low enough. (Linear throughout this discussion means without distortion, not flat in response. People in audio journalism have taken to saying linear when they mean flat in response. I am saying linear in its true technical meaning of being a linear system, whether flat or not.)
This possible non-linearity is not something that is true of necessity—theoretical viscous damping is a linear process for instance. But Goller says that in practice, damped panels become non-linear at low signal levels.
So the idea of the GamuTs was to use rather less damping than usual, with the intention of preserving the natural and presumably recorded decay of sound. Also, as noted previously, wood rather than particleboard was used, this being regarded as having a more musical sound. To avoid coloration, the resonances of the quite lively cabinet are tuned to different frequencies. This is verifiable directly. Bang on the middle of the enclosure with a knuckle and you get an F, bang on the top and you get a C (more or less). By comparison, knock on a BBC damped-wall enclosure and you get an essentially un-pitched sort of thud with rather little specific frequency perceived at all. Knock on the walls of the Rockports, say, and you get nothing much at all. [Editor’s note. GamuT sent this clarification regarding the cabinets: “The description of the RS5 cabinet is not entirely correct, as it is described as lively. This is probably due to the fact that one only ‘sees’ the outer 12mm layer. The total board, however, is actually 28mm thick, and consists of an inner part made from 16 layers of 1mm veneers, the outer part is 12mm thick and made from six pieces of 2mm veneers. We apologize greatly for not having pointed this out more precisely.”]
The composite result from the GamuT approach is supposed to be a sound that is not overtly colored by pitched resonances— multiple resonances with different frequencies will sound less colored than a single resonant frequency for the whole cabinet— but at the same time “life” and correct decay are supposed to be preserved by the comparative absence of damping of the enclosure. Incidentally, the cabinet and crossover design must be taken as the justification of the price, since the drivers probably account for only a small part of the cost. The drivers are high- quality, and made by ScanSpeak to GamuT’s specifications. GamuT then modifies the drivers by impregnating the cones with a “natural oil.” Goller has compared this to the strings on a Stradivarius versus the violin as a whole, the violin being presumably the cabinet. And certainly the cabinet does function here as a musical instrument, in a sense. It plays a large role in the design thinking.
It all sounds a bit theoretical, but before you ever hear the speakers, just knocking around on the enclosure does give you the distinct impression that here is something new and different. And Goller, while he has theoretical justification, also appeals to the idea that there is something intrinsically less musical about damped sound than (distributed) resonance. The theoretical picture is a little indefinite because in linear-systems theory a (minimum-phase) system that rings equally at all frequencies, that has completely distributed resonance, does not in fact ring at all. But the idea in the GamuTs is to get sort of halfway there, to distribute the resonances enough to have low coloration but to retain the liveliness of minimal damping. That is my impression of what Goller is getting at, anyway. Apologies for any possible misrepresentation—it is not easy to describe what amounts to mathematical matters in informal terms without losing precision.