Zu Druid & Definition Roundup

In separate threads about the Zu Druid V and Zu Definition 3 & 4 in this forum, several questions have been directed to me about the comparative merits of these models, supertweeter capacitors, and a variety of other variables. Rather than bury comments in those threads, I thought it better to start a new thread and focus any follow-up comments or questions in one place.

Over the past few weeks, I helped a new Definition 3 owner install and setup his speakers, after earlier having setup his loaner Def3s that had an earlier iteration of the supertweeter network. Additionally, I made a capacitor change on the high pass filter to the supertweeter on my own Definition 4 and Druid V speakers. For further perspective on this, I have lived with my Definition 4 speakers for the past 13 months, and my Druid Vs for the past three months. Prior to that, I have migrated through the Definition 1.5 > 2 > 4 upgrade path, and Druid “3.5” > 4 > 4-08 > 5 upgrade path in two discrete systems since 2005. Any search on Zu topics or my handle here will serve up plenty of commentary on Zu speakers, cables, suitable amplification and other related matters, so I am not going to attempt to repeat all of that here. But I am going to roll up a collection of observations in response to prior questions, that might help Zu owners understand the relative value of current options in the upper half of Zu’s range, as well as people who have never owned Zu but who are considering their speakers, to better grasp what they might gain.

Druid 3, 4, 5

My first Druids were a used purchase from a prior owner here in Los Angeles. It turns out they were one of the first 10 pairs of Druids made. They had been sent back to Zu in late 2004 to be upgraded to then-current configuration plus had full internal Ibis cabling. The first 10 Druids made had the Speakon connector for full B3 geometry from amp to drivers when using Zu cables (I did), along with parallel Cardas posts for connecting any other cable. When I bought this first pair of Druids, they were shipped to me from Zu, in what Sean called a configuration he approximated as “version 3.5.” That speaker hooked me on the holistic Zu sound, but it had a euphonic warmth and soft top end that was forgiving and not fully revealing. Nevertheless, that v3.5 Druid was addictive for its unity of behaviors, and much like the original Quad electrostatic its ample advantages made it easy to overlook its limitations. The v4 upgrade opened up the top end marginally and was welcome, but the Spring 2008 v4-08 upgrade to Druid was a big leap toward bringing Druid closer to the liveliness and open top end of Definition. Then Druid was taken out of the Zu line. I let the Essence aberration pass by. Sean got back on track sonically with Superfly but I preferred the Druid form factor so stuck with the dead-ended Druid 4-08 for my secondary system, all the time lobbying Zu – along with other Druid owners – to restore Druid in more modern form in their line.

We got exactly that in Druid V late last year. For 4-1/2 years, while Essence came and went, Superfly got the HO FRD and then Nano, Druid was static and falling behind. Version 4-08 still had some tone-density and focus that was sacrificed in Superfly in favor of that speaker’s livelier, burstier dynamics and somewhat more expansive scalar projection. Superfly also had a slightly more extended top end than Druid 4-08 so to most people it simply sounded more like a modern speaker should, than Druid 4-08. It also had a more complete Griewe implementation, for faster and more textured bass than Druid. Druid V addressed all that, and more. The more advanced multi-composite cabinet with integral full Griewe and the mechanical grounding of the thick aluminum plinth would have comprehensively improved Druid even if the old Druid drivers had been installed. But the advance of the Nano FRD and the Radian 850 in supertweeter use gave us a Druid form factor speaker that has the linearity and finesse of Definition, with the traditional focus, unity and tone density of Druid even more present and obvious than in any prior version. Druid V *is* the modern equivalent to the original Quad ESL, without the extreme beaming, the bass limitation, dynamic restriction and fragility. It just happens to deliver Quad-like unity and speed from dynamic drivers with much higher efficiency *and* power handling. Druid V is finally an uncompromised and uncompromising speaker that despite its price can be justifiably driven by the very highest quality amplification at many times the cost of the speaker, yet can put modest amps in their best light. Why would anyone drive Druid V with amplification that costs lots more than a pair of the speakers? Because the total design can leverage stellar amplification, and no other speaker today can duplicate the full combination of attributes that Druid V delivers. You can get even greater focus and unity, ironically, in Zu’s line from the ~$60,000 Dominance, with its radiused front baffle and three FRDs, but not with Druid’s lightness of mass, presence and drivability. No Magico at any price can deliver Druid’s pure unity of behaviors regardless of what you try to drive them with, and no Magico is as musically satisfying with such a wide range of amplifiers. Druid V laughs at the cacophonous disunity of a Wilson speaker. Druid V ridicules the dynamic choke points imposed on Focal speakers at the crossover points. In the same way that no one appreciative of the unity of the Quad ESL heard any musical value from the Infinity IRS or a Duntech Sovereign back in the day, a Druid V owner today can pretty much ignore the rest of the alleged “high-end” speaker market inflicting damage upon our hearing, with the exception of other Zu speakers.

Because of the newest Nano FRD’s ability to reproduce more musical scale than prior Druids, for the first time in version V, Druid is a credible HT2.0 speaker in addition to being a great 2ch music speaker. Also for the first time, Druid is now quite good for listening to a full orchestra, whereas earlier Druids fell short on scale for orchestral purposes. Druid V is the first “no-apologies” Druid. That’s not to say that Definition doesn’t have advantages for more money – it certainly does. But Druid V is now a true all-music, all-purpose speaker with no real musical limitations in practical domestic use, and if a lower linear limit of about 35Hz isn’t deep enough for you, there’s always Zu’s new subwoofers. It’s also extremely amplifier-friendly. And the Griewe implementation does a fabulous job of extracting solid, tuneful bass from low-damping-factor/rising-deep-bass-THD SET amplifiers. Druid V gets qualitatively better bass from 2a3, 45 and 300B SET amps than any unassisted (no powered sub) speaker I can think of.

Definition 1.5, 2, 3, 4

The 2004/5 era Definition 1.5 was a revelation in its day, for its combination of speed, transparency, resolution, scale, bombast and finesse while having very good unity behaviors and terrific amplifier friendliness. It was sharply different from the same-era Druid because of its extended top end, almost tilted a little bright, and for its impressive sub-bass foundation. It was a relatively big, bursty, lively speaker even driven by modest power. It also had two clear deficiencies: first the sub-bass array amp had no level control (later and quickly rectified for everyone after I pointed out the glaring omission upon receiving my speakers), and second, that v1.X Definition’s MDF cabinet “talked” at high SPLs, marring the clean and incisive sound with an overriding glare. In Definition 2, cabinet talk was dramatically reduced by introduction of the birch-ply cabinet structure, stronger baffle, more robust plinth and associated damping techniques. The voicing of the speaker also tilted somewhat darker but the net result was a Definition absent ringing and glare, cleaner at moderate SPLs and far less fatiguing at high playing volumes – even fair to say altogether unfatiguing. While Definition 4 introduced many simultaneous improvements, Definition 3 shows clearly how much cabinet talk was left in Def2’s “silent” cabinet. Def3 starts with a Def2 cabinet and gets additional bracing and damping during the upgrade and it is plainly apparent when you first fire up Def3s after being familiar with Def2, that sound emerges from cleaner, quieter noise plane in the newer speaker. Def3, while retaining Def2’s 4x10” sub-bass line array on a rear baffle, gains seriously-improved deep bass by virtue of replacement of the Def2 plate amp and level control with Def4’s D amp with parametric controls. The Dominance trickle-down Nano FRD gives Def3 a close facsimile of Def4 performance from lowest response up to 10kHz or so, but Def3 uses the older-generation Zu supertweeter, which cannot begin to match the beauty, finesse and spray of the Radian 850 supertweeter used in the upper range Zu speakers. Def3 sub-bass performance is not equal to Def4’s but it is surprisingly competitive. In the Zu FRD range of roughly 38Hz – 12kHz, Def3 is very close to Def4, separated by clear differences in cabinet construction and internal configuration that give Def4 advantage as should be the case. As you get above roughly 8kHz, where the Radian 850 in Def4 begins to slope in, the upper range of the FRD in Def4 through the Radian’s exclusive extension on the top are in absolutely every way contributive to an elevated sense of musical fidelity and realism.

Definition 3 would be a market-wrangling speaker not surpassed at 3 or 4X its price if Definition 4 did not exist. But it does. As good as the new sub-bass amp and parametric controls are for the older 4x10” line array on the back baffle of Def3, the 4x10” rear-firing cones can’t load the room as evenly and deliver the incisive unity of Def4’s downfiring 12” driver. As closely as Def3’s Nano FRDs match the same in Def4, the completely re-architected cabinet of Def4 allows the drivers to perform with greater neutrality and freedom from distracting resonance. And the Radian 850 sprays the loveliest and yet most objective harmonic content of any tweeter I can think of today. The combined effect of Def4’s improvements over the Def2/3 design make it a compelling upgrade worth every penny to anyone who can afford its price compared to Def3, and yet the bargain roots of rendering Def3s from donor Def2s yields a speaker that is astonishingly great for its sub-$10K price and is necessarily limited in the number that will be produced. Notwithstanding that Omen Def is probably the peak value point in a two-FRD Zu speaker, for true high-end applications, Def3 is the high-discretionary-income value point and Def4 above it is the luxury alternative that nevertheless has no non-essential waste in its composition or price.

Definition 3 or Druid V?

I get this question privately from time to time: “For less than $2K difference, Druid V or Def3?”

These two speakers suit different priorities. Ask yourself the following:

1/ What is your application? That is, do you use your speakers strictly for 2-ch music or is your system doing dual duty for 2ch music and HT2.0?
2/ How important is the bass region between 16Hz - 35Hz to you?
3/ What are you using for amplification?
4/ What is the size of the space you have to acoustically load, and how far you sit from your speakers.
5/ What are your music listening habits, and what are the 3 - 5 sonic attributes you most value to feel satisfied?

There’s not a straightforward answer to this question, without knowing the above, but it’s easy enough for anyone reading this to self-sort. Druid V will give you focus, tone density, top end finesse and beauty that Def3 can’t quite match; Def3 will give you spatial & dynamic scale, deep bass foundation, resolution and horizontal dispersion that Druid V can’t equal. Overlapping both are the speed, agility, transparency and shove of the Zu Nano FRD. So, having the honest self-awareness to know what satisfies you most if your finances force a choice, will yield a crisp answer. If you can’t live with the trade-off, that’s your signal to save, and save, for Definition 4s.

Supertweeter Network Capacitors

Recently, there has been a lot of new interest in capacitor upgrades for the supertweeter high pass filter in Zu speakers, particularly the Druid and Definition. I have not been able to listen to all the available and oft-discussed options. My Def2s and Druid Mk 4-08s had Mundorf Silver-in-Oil caps. I had my Definition 4s built with V-Cap CuTF as an upgrade over the Mundorf. My Druid Vs were built with Mundorf Silver-in-Oil. In January, at Sean Casey’s recommendation, I had Clarity caps installed in both Def4s and Druid Vs. My Duelund capacitors are back-ordered (well, Zu urgently needed my pair for a more demanding customer), so I await them. I have heard Duelunds in non-Zu speakers. There are a few things I can say about capacitors at this stage, with more comments to follow as I put more contenders head-to-head.

1/ Every capacitor brand, formulation and composition brings specific attributes and a sonic signature. None are perfect. Not even Duelunds. You tend to think that what is best in current experience is as good as it gets until you hear something better. I can understand why someone feels ecstatic allegiance to Duelund caps, while at the same time appreciating why someone else prefers V-Cap TFTF or CuTF or some other alternative to them. For example, Sean Casey takes the position that Clarity caps bring 85% of Duelund’s sound quality to Definition 4 and Druid 5, for less than 1/3rd the retail cost. Elsewhere on this forum, another poster relates a conversation wherein Sean said something similar about the Audyn True Copper caps (90% for 10%). I haven’t heard the Audyn capacitors so have no comment right now. I will say that if Clarity is close to Duelund results, then both are a clear improvement over Mundorf Silver-in-Oil. The Clarity cap is both revealing and exceedingly smooth. But the case for Clarity (and by extension Duelund if Sean’s assessment holds) isn’t a slam-dunk compared to V-Cap CuTF or TFTF. There’s such a thing as too-smooth. This is reminiscent of the same disagreement I have with advocates of “slow” voiced SET amplifiers compared to the quick and transparent Audion SET amps that are so unlike most other SET brands. Some listeners are strongly attracted to a too-smooth representation. A lot of instruments have some harshness and rough texture in their output. The Clarity sands a touch of this off, just like (but less than) the round-sound old-school SET amp voicings some listeners favor. The V-Cap has more snap & tooth in its sound, but it is also less forgiving. I’m still in trial with a decision about whether to stick with Clarity or return to V-Cap CuTF or TFTF – as well as Duelund – pending. No, don’t bother assuring me that I’m going to love Duelund caps. Just consider me open to being convinced, but also not assuming a priori I will be.

2/ All of these exotic film caps take time to settle in. Clarity sounds great fresh but then they put you through a few weeks of meandering performance. They seem to be sensitive to temperature during the infant hours of use. We’ve had an unusually cold December and January here in Los Angeles, and I don’t use much furnace heat (you northerners and east coasters should see what people in SoCal consider a “furnace…”). A day of 64 degrees in my house sets breaking-in Clarity caps back a couple of steps. A warm day with internal temps in the high 70s pushes them forward. Then they go through a period of sounding beautiful on simple music, but shut down with congestion and blur on complex music. And then they start being reborn again to reassert their original convincing impression, and more. You have to be patient with any change.

3/ The Radian 850 in supertweeter application in Druid V and above in Zu’s line is intrinsically smooth, articulate, detailed and lovely. Frankly every cap sounds great into it, with the worst and the best still within the realm of excellent. You’ll hear differences and likely develop clear preferences, but even the basic Mundorf Silver-in-Oil sounds fully credible and completely acceptable in the absence of hearing something better. But the advantage of upgrading the Clarity (or Audyn True Copper, I imagine) is unmistakably beneficial to Def3’s supertweeter, and any earlier Definition or other Zu speaker using it, is fairly dramatic insofar as you are paying attention to top end harmonic character and are influenced by it. Clarity really tames much of the comparative roughness in the pre-Radian Zu supertweeter, compared to all the stock cap choices put in those speakers. What I’m saying is, pick your cap for Def4 and Druid5, knock yourself out. Some will sound definitely better but all will sound very fine. But if you have a Zu speaker using the older supertweeter and have an appetite to give them a worthwhile refinement, get a Clarity cap network upgrade. The cost is very reasonable and the benefit is disproportionately large at the price.

4/ There may be a cheap sleeper in capacitors. I was discussing film cap upgrades with Bob Hovland a couple of weeks ago. He mentioned that his more recent research indicated that the material consistency of the dielectric in film capacitors (even thickness & density, absence of pinholes) is more influential to sound quality than specific materials themselves. He wasn’t suggesting that all more exotic capacitors might not deliver someone’s preferred sound, but he does believe an excellent sounding cap can be made from prosaic materials. SuperCaps has a relatively new family of “Robert Hovland Edition” film caps that are highly affordable. They are handmade in the US, comprised of non-exotic materials, highly inspected during build and sealed tightly. I got some samples from Bob to try in my tube-output DACs and the results exceeded my expectations by a wide margin. They are more than good enough to settle on, and are staying in the DAC (mhdt Havana Balanced). He is next very eager for me to try a pair of 1uF/1000v versions in my Zu high-pass networks. I don’t know what to expect relative to Mundorf, Clarity, Audyn, Duelund but it’s a trial too interesting to not undertake. I’ll post back results, perhaps after I can put Duelunds in the mix, too.

Enough for now. I’m happy to add comments if questions are posted. I am sure I will remember something I intended to write here, but forgot.

Phil