|Broadband Noise Reduction
Any pretty good audio equipment rack will achieve some level of noise reduction at some frequencies. The problem, however, is that many rack systems reduce vibration at some frequency ranges only to amplify it at others. The key is broadband noise reduction. Broadband noise reduction is easy to achieve in theory, but very difficult to achieve in practice - especially with respect to mid and high frequencies. In general, the key is to create a large mismatch between any system's excitation frequency (Fe) and its natural frequency (Fn). In principle, Fe should be greater than Fn. Theoretically, this should be no problem. In practice, however, the individual components that make up a system and the compliant elements themselves have their own vibration modes that get excited by mid and high frequency energy. So if you construct a system that is isolating at low frequencies and should be isolating at mid and high frequencies as well, it may actually be increasing output at those frequencies because of the vibrational modes of the components and their compliant elements.
This means that any way of achieving broadband reduction in practice has to begin with a much more complicated set of mathematical models that take into account all of the frequency modes of each component in the system and how they react individually and system-wide over a broad range of input frequencies. Not to put the point crudely, but there is literally no hope of achieving genuine broadband noise reduction through simple engineering solutions. The failure to achieve broadband reduction is sonically very easy to pick up, and may even be somewhat attractive to the ear at first. I have heard a much praised system isolate at low frequencies while actually amplifying mid and high frequency information. The initial sonic signature of such a system is an apparent increase in high-energy information and detail. What one is actually hearing is spot-lit and highlighted information of just the sort that sells equipment in high end stores, but which is in fact simply untrue to the music. Racks can trick just like cables and speakers can. The feature that most listeners are drawn to is a large soundstage, the sense of extraordinary high frequency information and detail. I just never hear live music in this way. In fact, I'm with Anthony Cordesman on this. In two recent reviews in The Absolute Sound, Cordesman quite rightly identifies these features of playback as artificial and bearing no relation to accurate musical reproduction. On the other hand, high-end systems tend in this direction and many purchasers feel that they get their money's worth only when they hear picked-out details that would otherwise be submerged in what they mistakenly think of as less 'revealing' systems.
A system that reduces noise broadband should have the following characteristics. There should be increased clarity throughout the frequency range. High frequency information should be naturally and fully resolved and should appear as part of the natural mix and not as if someone went in with a microscope and picked this or that 'detail' to highlight or draw attention to. Nothing in the playback should draw attention to itself. The overall sound should be relaxed yet dynamic, tuneful and well paced, but never unnaturally excited or energetic. The noise floor throughout the frequency range should drop precipitously and you should find yourself with two somewhat incompatible-seeming urges: the first being to turn down the volume control to secure the same level of volume from the system; the second being the urge to turn the volume up since the elimination of 'nasties' makes it considerably easier and more enjoyable to listen at louder volumes. There should finally be an increased sense of coherence through the entire range, a sense of immediacy and presence as yet another layer of artifacts is removed between you and the music.
I was able to experience a variety of particular instances in which the HRS system dazzled in these and other regards. When Mike first set up the rack, I had the Wilson Sophia loudspeaker with its inverted titanium tweeter. I quite liked the speaker; indeed it remains the only Wilson speaker I could personally live with long term and I did. In any case, I am no fan of any of their tweeters (though I do prefer them to the unfortunately ubiquitous beryllium tweeter of the day). While Mike was setting up the rack, a group of us were talking about the Wilson tweeter, with no one singing its praises. Mike took a contrary view. He asserted quite boldly that the ringing we often hear and associate with metal tweeters of any sort is largely an artifact of a failure to properly isolate associated equipment. Isolate the equipment, he alleged, and the ringing and harshness will disappear and in its place you will find a smooth, extended top end. The minions were skeptical to say the least.
In what has to be one of the great demonstration moments of all time, when the rack was set up completely and all the equipment back in place, lights off and the system fired up and music playing on TT and CD player alike, guess what happened? A transformed Wilson loudspeaker. While most of us sat with our jaws dropping and our mouths agape, Mike acted like this was a common occurrence. The HRS rack by the way sits in the super high-end room of New York's most prominent Audio Emporium where I have heard it work the exact same magic on the beryllium tweeter. How do I know the effect is real? Easy, I've heard the same tweeter in other rooms in the same store and not once did I find it a satisfying experience. On the other hand, I could listen to the top end in the back room all day; not only without fatigue but deriving real pleasure from doing so. I wouldn't want that confession spread though.
The Isolation Platform/Base
The shelves of each HRS rack are separate M-3 Isolation bases. Bases are specifically configured to work with the customer's equipment, in sizes that correspond to component size; within the size parameter, they can be constructed to handle components of different weight. Mine were 19" x 21". All three were designed to handle the components that were part of the reference system. Changing the weight that an isolation platform is optimal for is a piece of cake. If you purchase heavier equipment requiring a different isolation platform structure, all you need do is contact your dealer who will come to your home and replace the feet under the platform. In effect this means that the HRS platform can last a lifetime. The only changes necessary are to the feet supporting the isolation platform. Mike Latvis estimates that normal changes in product purchases ('normal' by audiophile standards I imagine) should not amount to much more than another $300 in costs over the lifetime ownership of the rack. Speaking of which, each platform in designed to maintain its resonance control properties without deterioration indefinitely.
There are two main structural elements in each isolation platform: a machined aluminum frame and a granite slab. Structural simplicity hides the engineering that lies just below the surface.
The 0.75"-thick granite slab, which is placed inside the frame, is in fact decoupled from it and sits on a proprietary polymer substructure that has two purposes: 1) is to support the weight of the granite and the component, (2) is to decouple the granite from the frame and dissipate residual vibration energy and control structural resonances. The entire base rests on four custom billet-machined aluminum isolation feet precisely shaped to ensure minimum connection with the surface below. Each HRS Isolation Base foot is a six-degrees-of-freedom isolator. I will return to this feature of the footers when discussing the brackets. One of the secrets of the HRS rack is the manner in which the footers connect to the brackets which approaches a theoretical ideal that is yet another engineering feature designed to reduce the transfer of energy through the system.
The entire idea behind the HRS is to make sure that very little if any structural-borne vibrations get past the primary isolation stage. Any vibrations that penetrate the aluminum frame heading in the direction of your precious preamp, CD player or amplifier face a series of mechanical chokes and other energy reduction devices machined into the aluminum frame. The phrase 'mechanical choke' is used to describe a mechanism of mechanical constriction. A mechanical constriction occurs when you change a relatively large cross-sectional area into a relatively small cross-sectional area. The idea is to take mid and high frequency energy that is traveling through the frame and have it confront a mechanical constriction or barrier. Designed properly, such a constriction will reflect back to the source, thus reducing the amount that passes through. In toto, seven different materials are used within the aluminum frame. The materials chosen and their densities and locations within the frame are determined by their various anti-vibration properties. Final adjustments to materials and choke locations and material combinations are made by extensive listening tests.
Now imagine that some vibrations make their way beyond this point. If so, they will run into the proprietary polymer substructure that further reduces unwanted vibrations. In fact, the proprietary polymer substructure serves two purposes: first as I mentioned, to prevent vibrations that have come this far from proceeding further. This is accomplished by decoupling in ways that are not unfamiliar to those of you who have experience with Sorbothane feet. The HRS polymers are proprietary of course and their performance is said to exceed that of others.
The more interesting feature of the proprietary polymer substructure that decouples the aluminum frame from the granite is that it is designed to control the resonances of the granite. Like all other materials, granite has a natural resonance frequency. Black granite is chosen for several reasons. A relatively small piece of granite goes a long way in being able to stably support components. In addition, granite ranks a 9 on the Mohr's hardness scale. Only diamond ranks higher (10). Among other things, this means that you can use metal cones of your choosing to further decouple your components from the isolation base without worrying about cutting, scratching or scraping the surface.
One thing though, if you choose to further decouple your components from the granite, Mike is among those who would discourage you from making a three-point rather than a four (or more) point connection. Kiuchi-san of Harmonix, among others, also rejects three-point isolation. In any three-point approach, two corners of the component will be cantilevered. This creates an opportunity for the so-called 'diving board' effect, which may well significantly amplify any input vibrations that make it this far, including the air-borne vibrations that reach the skin of your components directly.
The issue with three-point isolation may be more telling when such connections are used as way of fastening shelves to racks, especially if the shelves themselves have long horizontal cross braces. Shelves connected at three points are in principle very problematic (other things being equal) when components with four feet are placed upon them. This invites a more substantial impact of the diving board effect (cantilevered mass). I have no doubt that designers who rely upon three-point connections work hard to overcome the diving board effect and it would make interesting reading to learn more about how they do so.
There is no doubt that one could have chosen a material other than granite for the base of the platform: wood, acrylic and carbon fiber come to mind. In addition to having the properties I discussed above, granite adds a partial mass-damping element to the overall design. I have heard the HRS rack described as a mass-damping approach to resonance control. If I have succeeded at explaining how the isolation platforms work, one thing should be clear and that is that Mike Latvis employs a mixed approach with very little actual emphasis on mass damping. One has to be careful not to confuse the fact that while the HRS rack is heavy, that's different from claiming that it kills resonances through mass damping. Mike Latvis is quite explicit in suggesting that mass is a spectacularly inefficient way of controlling resonances.
Whatever its virtues may be, Latvis is fully aware of the fact that granite has a sound of its own: that is, it has a natural resonance frequency of its own. The proprietary polymer substructure between the granite and the bracket is designed to control the untoward consequences of the granite's resonance properties.