This review page is supported in part by the sponsor whose ad is displayed above

Reviewer: David Kan
Digital Source: Restek Radiant, Philips DVP-9000S, Micromega Microdrive, Deltec PDM Two DAC, Assemblage D2D-1/DAC-3.1 Platinum
Preamp: Symphonic Line RG 3 MKIII, Audio Zone PRE-T1
Power amp: Symphonic Line RG4 MkIII, NuForce Reference 9, NuForce Reference 9 SE, KingRex T20 integrated amp
Speakers: Apogee Centaur Minor, Loth-X BS-1, Unity Audio Clearwater [on loan], Unity Audio Inner Spirit [on loan], Unity Audio Whitewater [on loan], Mark & Daniel Ruby
Cables: Symphonic Line Interconnect, Luscombe LBR-35 interconnect, Unity Audio Solid Link interconnect [on loan], OCOS speaker cables by Dynaudio, Unity Audio Solid Link speaker cables [on loan], Orphee Audio digital cable
Power Cords: Symphonic Line power cables, Aural Symphonic Missing Link, Ensemble Powerflux, Unity Audio Link /Precision Link [on loan]
Power Line Conditioning: Tice Power Block IIIC, Monster Power HTS-3500 Mk II (modified by Nu Force), Monster Power HTS-1000 Mk II
Room Size: 15' x 13.5' x 8' diagonal set up, 12' x 24' x 9' opens to 12' x 17' x 9' L-shape, short wall setup; 13' w x 28' 8" d x 9' h in short wall setup, with openings on one side to hallway and staircase
Review components Retail: OCOS 2 meter/pair $299, OCOS Standard Adaptor (red) 30/ea., OCOS RFC Adaptor (black) 65./ea.; Speaker Enacom $99/pr, RCA Enacom $235/pr

It rings. What horror. And it's not the telephone.
High frequency ringing. Does that ring a bell? I haven't bothered with that for a long time until a reader recently wrote me about high-frequency ringing from his NuForce Ref 9 SE V2 and Mark & Daniel Ruby setup. The gentleman had invested in this setup at his own free will of course. Nevertheless, I felt guilty that I hadn't heard any ringing myself when I reviewed the Ruby mated to NuForce. For me, one easy way out would have been to say to this gentleman: "Hey, mine was a NuForce Ref 9 SE V1." But was there anything I had missed? I felt I owed everybody and myself an investigation, and hopefully, a post rationalization. By the way, to avoid misunderstanding the term 'ringing', I am referring to the super high frequency disturbing sounds that seem to rattle or shrill in your inner ears when it comes to some high-pitched soprano or even tenor singing (Placido Domingo at times), as well as higher-octave piano notes, usually when played in sharp, brisk staccatos (Bartok and Prokofiev for instance). In the worst case, ringing could also happen during clashing chorus, string and percussion interludes. It could even make woodwinds sound metallic and strident, especially the piccolo, which operates in the high registers. In milder cases, ringing might not happen all the time and there may only be a few CDs in your collection that trigger it off. Still scary, isn't it?

What causes ringing? By definition, ringing is a form of transient noise, a high-energy impulse waveform, an excitation of high-frequency oscillation. Unfortunately, this seems to be one of the less scientifically documented subjects in our theorized-to-death audio symposium. From resources limited by my own laziness, consensus seems to be ultrasonic RFI/EMI distortion that is inductively and capacitively coupled to the amplification circuit. Sources for these unwanted bugs to propagate range from appliances to radios, motors to computers. In short, anything that involves electromagnetic force. Electromagnetic interference can be induced internally in audio equipment as well. Class D (or power switching) amplification in particular induces ultrasonic noise during pulse-width modulation and requires output filters to finish the job. Not only is this type of amplifier sensitive to radio tuners nearby, it can transmit RF signals if the parameters of the output filter are not meticulously calculated. When coupling with a loudspeaker, the crossover network of the speaker becomes an extension of the amp's output filter. For that reason, single-driver speakers or high efficiency speakers without crossover networks are more susceptive to ultrasonic noise pollution. They lack a second line of defense. That partly explains why we should exercise caution when mating class D amps to such speakers. To a certain extent, that applies to class T as well even though they modulate analog input signals and control output switching timing more accurately for much lower electromagnetic emission (at least as claimed by Tripath). On the other end of the spectrum, RFI/EMI can start doing damages as early as in the recording stage, interfering with microphones, recording equipment and mixing consoles. Recordings with this kind of unfortunate hereditary ringing will sound unpleasantly bright and harsh no matter what.

Traditionally, the use of bypass capacitors (to decouple the active device that generates EMI/RFI) and series resistors (to attenuate the unwanted high-frequency signal strength) are the active measures to neutralize the intrinsic common mode noise that causes distortion and ringing. Passive tactics to reject external interference include shielding (such as EMI/RFI gaskets or ferrite magnet suppressors that clamp around the end of cables) and braided outer conductors of a coaxial cable. Interestingly, more and more silver interconnects these days do away with braided shielding. That includes my Luscombe LBR-35 and the Slinkylinks recently reviewed by Frederic Beudot. One other inexpensive passive shielding add-on device are the non-shorting RCA socket caps to stop EMI/RFI from creeping in through the back doors. I tend to put them on every RCA input socket not in use.

Two other seemingly insignificant components have existed in my audio systems for twenty years. They blend in so well that they virtually sink into oblivion and I hardly pay any attention to their existence, let alone their benefit. The first is the RFC Filter or RFC Impedance Stabilizer which is an integral part of the OCOS speaker system. The second is the Enacom speaker end compensator.

OCOS, the acronym for Optimal Connection System, was developed by Dynaudio with German know-how and manufactured with Swiss precision. Laughing in the face of exotic materials and esoteric construction, OCOS doggedly chose to be a simple coaxial, highly purified copper cable with straightforward winding geometry of non-acute-angle twisting for the core conductor and non-fancy braiding for the outer conductors. Just like any generic interconnect really. Measuring only 6mm in diameter, the cable is fitted with a special graphite-composite insulation which is deliberately conductive within a predefined limit. This protective sleeve is resistant to abrasion and scratches and resilient to bending. I swear that OCOS is the thinnest yet toughest cable I've ever worked with. Every time I tried to cut or rework this cable to custom lengths even with all the proper tools, my finger tips felt terribly sore afterwards. The OCOS cable is terminated on both ends with proprietary male coaxial connectors finished in Sucoplate plating. Sucoplate looks just like stainless steel but is in fact a copper-tin-zinc alloy that, due to its non-magnetic, non-allergic property and equal-to-silver performance, has found its way into high-tech applications like off-shore cabling, telecommunications, airframes and space and test equipment.

What makes OCOS really special is the adapter that accepts the male coaxial connector and then connects to the speaker binding posts. This black female plug, also Sucoplated and labelled RFC Filter or RFC Impedance Stabilizer on the Teflon sleeve, connects to speaker binding posts through a pair of very short red and black wires. What it does is keep the impedance exceptionally low across the broadband frequencies regardless of cable length.

The above chart compares the characteristic impedance of OCOS vs. three other commercial speaker cables. The wave resistance of OCOS runs straight as a ruler at about 10 ohms from 10Hz to above 1000kHz while other cables rise above 400 ohms within the critical bandwidth. "Even with lengths up to 100 meters, the impedance retains its constant value and the sound remains neutral and balanced," claimed OCOS. That's why OCOS is particularly suitable for home-theatre applications where cables connecting to multi-channel speakers may be of different lengths without jeopardizing performance. In my NuForce Ref 9 and Mark & Daniel Ruby setup where the rig is tugged into the left corner, I run the left speaker cable at 10 feet and the right one at 23 feet!

In a 1990 Stereophile staff meeting to discuss the 'Component of the Year', the subject soon fell on cables. 1990, as we might recall, was the year extravagant cabling techniques began to proliferate. Writer Sommerwerck noted in the discussion that despite the presence of Audioquest's Bill Low: "The only cable manufacturer that has written anything vaguely convincing, one way or another, was OCOS." What he referred to was the matched impedances theory which forms the basis of the OCOS technology: Compatibility of source impedance (amplifier), wave resistance (cable) and load impedance (loudspeaker). In hindsight, it occurred to me; the real benefit of OCOS had to be more than just being able to run asymmetrical cable lengths.

Let's look at the second chart. (The original source material is in German. HFC adapter is the same as RFC adapter - Hochfrequenz is radio frequency.) "Loudspeakers do have an impedance character which is frequency dependent. This makes for a hard job for the amplifier. The following chart shows the upper end of the typical impedance curve of a dome tweeter. In the RF range, the impedance of the voice coil grows substantially. The OCOS RFC-adapter is the perfect link between the OCOS cable and the loudspeaker system. The lower curve shows the tweeter impedance with the RFC adapter."

Having a chance to revisit the OCOS materials and perform some final research on the subject sixteen years late as it were, I began to get the complete picture. "OCOS is the first and only loudspeaker cable with absolute constant impedance throughout the entire frequency range, matching that of a normal loudspeaker. The sound is completely transparent, neutral and balanced without harshness, edginess or nervousness. Through its neutrality across the entire frequency spectrum, a recording will be rendered accurately and realistically with its resolution and spacious imaging fully intact." Commercial overtones aside, the operative words are neutrality without harshness, edginess or nervousness. I'm sure all Dynaudio owners would gladly testify that these are indeed the intrinsic qualities and sonic characteristics of their loudspeakers.

After a series of simple with-and-without tests, I'm further convinced that this innocent-looking black adapter is the facilitator that does the real work. It works most effectively with two types of components: class D or power-switching amps; and single-driver high-efficiency bookshelf speakers. My other NuForce amp Ref 9 SE V1 has been happily driving the Dynaudio Facette above, which, like the Contour 3 and other Dynaudio models of that time, has built-in OCOS sockets and impedance stabilizers. (The latest Dynaudio models seem to have done away with OCOS sockets but the OCOS-tuned sound and the Dynaudio signature sound have been inseparable.) Each one of my eight Loth-X BS-1s has OCOS RFC impedance stabilizers permanently attached to them and use only OCOS cabling. Apparently I've been enjoying ring-free performance without knowing the true identity of my benefactor.

So, what exactly is this mysterious black adapter made of? The secret exposed is nothing more than a Zobel network which is basically a resistor and capacitor in series. Since the voice coil of a driver has the characteristics of an inductor -- meaning impedance increases with frequency -- a series resistor-capacitor network connected in parallel to the driver can neutralize the effect of voice coil inductance. This kind of network is sometimes called a 'snubber circuit' because it softens the edges of a switching waveform, resulting in slightly slower rise and fall times and therefore less power at its higher-order harmonic frequencies.

Inside the black OCOS adapter we find a 0.33uF/100 V capacitor and an 8.2ohm/1W resister precisely doing just that. As OCOS claimed: "An integrated circuit balancing the speaker impedance at the very high frequencies." Bingo. In Wes Phillips' Monitor Audio Studio 2 loudspeaker review back in Stereophile, February 1995, he also reported that "As much admiration as I have for my reference Transparent Audio Music Wave Reference speaker cable, the tweeter sounded peaky and less controlled when I used it. Changing to the OCOS/"black interface" combination, the peakiness disappeared so thoroughly that my first thought was that the cable or termination was essentially some sort of bandpass filter." Wes's reference amps didn't involve any power switching then. But I've heard people commenting about Monitor Audio's tweeters as being metallic sounding.

The red adapter that connects the OCOS cable to the amplifier, however, is a pure Sucoplate connector without Zobel. Dynaudio recommends double or even triple OCOS, using more than one cable for each speaker connection (biwiring a single terminal) where less resistance could result in more transparency, most noticeable in the deep bass region. OCOS has marketed a triple interwoven cable named OCOS TT. When going double or triple, one black adapter is adequate. The rest should be just the red adapter(s). I have installed two additional red adapters (naked, without the red sleeves) inside the Dynaudio Facette to accommodate three runs of OCOS to each speaker. The result is most rewarding.

So much for OCOS. Moving on to Enacom. I have been using Enacom Speaker End Audio Compensators and Line Input Compensators for about 20 years but never paid any attention to them. I've left them on my Apogee Centaur Minor and Symphonic Line RG3 Mk III preamp (with Symphonic Line interconnects and speaker cables). Before I acquired Symphonic Line cables, I used OCOS and Clear Audio interconnects in this setup. I never bother to do A/B comparisons then because I'm not a tweak man. Since this ringing wake-up call, I realized that I had two other pairs of speaker Enacoms collecting dust (but actually plugged in) behind my less significant Quest speakers which were primarily for home theatre. What a sacrilege I thought. I took them off and tried them on the Unity Audio speakers during the KingRex review. They worked reasonably well in terms of controlling the high-frequency ringing that was quite prominent during the audition, making the high octaves sweeter and more polished.

Enacom by Harmonix (parent company Combak, prior to that, Enokido) is "the easiest -- and least expensive -- way to filter out annoying RF noise as well as the ringing high frequency distortion caused by the inherent load resistance of interconnect cables and speaker wires." There is also an AC Enacom, which is a 2-prong AC plug parallel filter but I haven't tried that. The theory Enacom puts forward is very similar to OCOS. "When a signal travels beyond the output terminals through the interconnect cables and speaker wires, it is tackled by load resistance producing ringing distortion. To make matters worse, RF noise (200kHz - 1 MHz) absorbed by the cables and wires further deteriorates signal quality. This is where Enacom comes to the rescue." Tightly sealed in cylindrical copper casings, the contents of the Enacom are a well-kept trade secret but many people speculate that it is some form of parallel Zobel network (that's why there's no connection polarity to observe). When connecting to a bi-wire speaker, Enacom needs to connect to only the HF terminal.

Based on the Unity Audio Whitewater + KingRex T20 and Apogee Centaur Minor + Symphonic Line systems, I conducted the comparison between OCOS and speaker Enacom. In the end, I preferred OCOS in many ways. First, OCOS is a complete system with solid physical parameters and sound technical backup. It's a virtually transparent medium between amp and speakers. Second, when I use triple OCOS with the Dynaudio Facette, the improvement in bass extension is very noticeable. So it is with the Apogee Centaur Minor. I guess both speakers tend to be bass shy if your cables don't let them 'see' the amps. OCOS keeps the impedance low at all frequencies and never blocks their 'view'. The Speaker Enacom on the other hand involves another variable, the speaker cable. Third, OCOS, being a coaxial cable, theoretically has better RF resistance. Fourth, OCOS can be run in any lengths. Plus, switching cables from speakers to speakers is as easy as plug-and-play if you have enough adapters for each speaker. But of course, Enacom is more affordable and can be conveniently added to any existing cable systems.

The line input Enacom somehow does not seem to have the same decisive control over the ringing this article is primarily concerned with. Nevertheless, as long as they are not making matters worse, I'm happy to leave them on. Now, can the speaker Enacom work in tandem with OCOS? I did try just that with the Unity Whitewater. I'm happy to report that a double dose will not get you to overdose. In fact, that does give me the best control over my worst hereditary ringing CD, The Tribute to Stanley Black [Decca 473 940-2]. By the way, OCOS adapters as well as both speaker Enacom and line Enacom come with no connectors.

Both OCOS and Enacom are patented but that doesn't mean DIYers cannot build their
own Zobel filters. In fact when I asked Srajan if I should proceed with this 16-year late review, his response was encouragingly positive and he added: "On my big Zu Definitions, I use Zobel
networks from Walker Audio as well. The speakers don't ring but the Zobels clean up some HF hash to make the treble sweeter. Merlin too incorporates a Zobel network (or at least they did in the past, I haven't kept up)". So even if your speakers don't ring, a snubber circuit could still make them swing.
Quality of packing: OCOS was purchased in1990 in custom lengths, no packing. Enacom was purchased in 1991, packed in apple pie paper box.
Quality of owner's manual: OCOS is basically a technical text with charts, very detailed and informative. Enacom is just a one page summary, poorly designed.
Website comments: Both the OCOS and Enacom websites are up and running, with adequate information except for pricing. OCOS is stylishly designed. But, hello, anybody home?
Global distribution: OCOS, Enacom
Human interactions: Emailed questions to OCOS inquiring about pricing and availability - no feedback except for an automated answer that acknowledge receipt of message. Emailed questions to Enacom about technological information (what's inside the Enacom for example) - no feedback.
OCOS website
alternate OCOS website
OCOS US distributor's website
Enacom website
Enacom US distributor's website