Definition has always measured quite flat and yet it is a deeply toneful speaker. Flat-measuring speakers that get that way through complex crossovers with steep driver contouring do tend to sound like anything *but* flat and are irritating for many people (like me) to listen to, for other reasons than measuring "flat." It's all in how the designer gets to "flat" on the graph.
The Druid in some respects ruined Zu's credibility on measurements because Zu didn't measure it early, while some reviewers did, not understanding how to measure its real-world use. The infamous Canadian review of early Druid, in which the publication complied with Canadian practice of measuring a speaker suspended in mid-air, yielded a disastrously ugly graph that had nothing whatsoever to do with how a Druid sounded in a real residential room when properly set up. But the empirical audiophile brigade couldn't (or wouldn't) grok the fact that Druid's partial Griewe loading requires placement on a floor (you know, the way you'd actually use the speaker) with some attention paid to a fussy floor-to-plinth gap, in order to sound and measure right. Since then, Zu owners haven't cared a whit how the speakers measure, and Zu critics won't bother hearing them, or grant that they can possibly sound good, let alone great.
Superfly sounds like it measures, objectively. But since it gets to a reasonably flat graph without a crossover and most of its range is produced by one coherent driver, it won't sound like most flat-measuring speakers sound.
As the graph indicates, Superfly will be revealing. It is a bit more "accurate" speaker than Druid 4-08. A grungy amp will sound grungier on Superfly than on Druid. But the high impedance makes most amps, even grungy ones, sound better than they normally do, and not driving a crossover before the voice coil puts them on better behavior too. So Superfly is both revealing and forgiving, all things considered. But conversely, an excellent amp will be even more appreciated on Superfly than on Druid 4-08.
The changes over Druid aren't that small, given how much the FRD has been massaged. The full Griewe model seriously improves and extends bass, but it also affects the driver's performance well into the midrange. The cone and motor changes to the FRD make it considerably more vivid. The high-pass filter to which the supertweeter is wired is both more linear and cleaner owing to the selected components. And the cylindrical phase plug markedly linearizes and improve dispersion of the FRD's high-frequency output from the whizzer. What a lot of other designers would try to upgrade through multiple drivers and a more complicated crossover, Zu has attacked by refining the main driver's capabilities directly.
A frequency response graph is only a facet of a louodpeaker's performance indicative of an ability to transduce convincing music. It tells you nothing whatsoever about tone density, timbral acuity, dynamic aliveness, transient speed, spatial projection, phase coherence, or other non-frequency fidelity factors. And since the speaker's interaction with actual room and amplification factors aren't at all captured by frequency response measurements, you can safely limit how seriously you take such representations of performance.
Do I believe "those numbers?" Numbers imply a precision to speaker performance that's not remotely sustainable in actual deployment into a customer's system. It's not that I don't believe them. I believe Zu attained that curve. I just don't think it's relevant. Superfly sounds neutral and extended in the ways Zu's graph suggests. That's easily the least influential reason to buy the speaker over the myriad other speakers that also measure "flat." in the same price range.
Phil
