Meanwhile, there’s another, critical aspect of rack performance that we haven’t touched on yet – levelling. Given that almost all racks have adjustable feet, that might seem like a given, but it’s not quite that simple. In any rack that uses a single monolithic structure (where the shelves are not independently adjustable) levelling the top shelf does nothing to guarantee that the other shelves will also be level. Bolt together racks, built from machined components are no better in this regard. If you don’t believe me, then just put a precision level on the different shelves of any rack you choose. Unless the frame is assembled on a perfectly level surface, with equal torque on all the fixings, the chances of achieving multiple, perfectly parallel surfaces is vanishingly small – considerably less than the chances of hearing a string trio being played at an audio show. And that’s if the shelves or supporting platforms are rigidly mounted. If they’re decoupled with a soft interface, such as the HRS isolation feet or the Sorbothane rings used by Grand Prix Audio, then even if the supporting surface starts off level, uneven weight distribution in the supported component will throw it out of true – another reason I like the move towards solid feet on the HRS platforms.
From a functional point of view, an ideal rack should either ensure sufficient precision in manufacturing and assembly that the various levels can be considered parallel, so that levelling one shelf levels them all, or, each shelf or platform needs to offer individually levelling. From an ease-of-use perspective, the former is certainly preferable, even if it is a vanishingly rare occurrence.
A level playing field?
Most racks are levelled using adjustable spikes or feet – which brings us to another great audio misnomer: the mechanical diode. I’m not sure who started this one, but let’s be clear: a mechanical diode would pass energy in one direction only – and a spike ain’t it! The idea that standing a speaker or rack on spikes allows energy to drain into the floor, while preventing energy from flowing from floor into the spiked structure, is complete horse feathers. But despite that, spikes are not all bad. Potentially, they offer point contact, ensuring a stable, consistent interface with the supporting surface. But like everything else in audio, the devil is in the detail. The efficacy of any spike is defined by the pitch and diameter of the thread (which in turn define its stability and adjustability), how easy it is to turn (does it have a through-hole or flats), the material it’s made from (stainless steel or titanium are good – raw or hardened steel, not so much), the spike’s profile, how easy it is to lock (precisely) in place and how securely it can be locked. That’s plenty to get wrong and it’s amazing how many spikes do just that.
Something often overlooked when using spikes is the resilience of the supporting surface. Over time, mechanical vibrations passing through the spikes can erode or displace the material directly beneath the spike’s point of contact, impairing mechanical coupling/grounding and absolute level. This is a big problem with speakers, but shouldn’t be overlooked with racks. Level needs to be checked and adjusted on a regular basis. The ease and precision of this adjustment have their own influence over the rack’s and thus the system’s performance.
The meat in the (rack) sandwich…
Having talked about the equipment to rack interface and the rack to floor junction, what needs to sit between them? That’s a more complex question. On the face of it, the answer is that the rack structure needs to be rigid and stable (in order to properly support often heavy equipment) but it also needs to be non-resonant and dispersive. The degree of dispersion will depend on the available material to do the dispersing – and just how good it is at the job. That’s one of the principal issues with suspended or flexibly coupled support surfaces: the support surface or platform becomes solely responsible for the dispersion of energy from the supported chassis. At the same time, monolithic rack structures risk inflicting their own resonant signature on the supported system as a whole. While it might seem like a good solution to build a massive frame and use it support individual, dispersive levels, that assumes that the energy entering the rack is going to behave in exactly the way you want it to, your shelves or platforms isolating the equipment from the outside world while also dissipating internal energy, the frame providing stability while its mass helps absorb energy entering from the floor. The problem is that when it comes to the frame, you need an awful lot of mass to be an effective absorber, while the supporting surfaces need to convert energy with almost supernatural efficiency. What actually happens is that both ‘systems’ end up storing residual energy that bleeds back into the mechanical environment that surrounds the signal. Add to that the widely varying energy spectrum that the rack might need to confront and targeted damping becomes impractical.
