I have built underdamped pipe speakers, and speculate that the Druid's enclosure may be a variation on an underdamped pipe.
The 150 Hz notch would indicate an internal path length of about 90 inches. When the path length is equal to one wavelength, the backwave energy emerging from the end of the pipe is 180 degrees out of phase with the front wave of the driver, so cancellation occurs. The depth of the notch makes me suspect the line is underdamped. Note that the ear is much more tolerant of frequency response dips than it is of peaks; without that SoundStage measurement, I bet very few people would know it's there.
Brave move on the part of Adam and Sean, sending their speaker to be measured when I'm sure they knew full well that its measured response would raise eyebrows.
If the 90 inches ballpark path length is correct, we'd expect to see a local response maximum at about 75 Hz, with response rolling off slowly to about 37.5 Hz and then falling off rapidly below that. This is reasonably consistent with the SoundStage measurements.
To my ears, the Druids sound like they have much more bass than the SoundStage measurements indicate is possible, even optimistically factoring in boundary reinforcement.
Normal boundary reinforcement will of course bring up the response in the lower octaves, but not all that much. I do not understand why the gap between the vent and the floor is so critical, but can accept that it is. Zu claims to use muffler technology invented by a man named Griewe, which I have been unable to find details of so presumably it's not patented. Mufflers.... hmmmm.... Jackson Browne.... hmmmm....
There are two ways to bring up the apparent loudness of a signal. The most obvious is to increase the intensity, or sound pressure level, of the signal. A less obvious way is to make the signal last longer. If presented with two brief signals of equal intensity but different duration, the ear will perceive the one lasting longer as being louder. Perhaps the muffler-like characteristics of the Griewe enclosure take advantage of this psychoacoustic phenomenon, filling in the perceived bass a bit by sustaining bass energy for just a little bit longer. "Oh won't you staaaaaay, just a little bit longer...." This is just a SWAG on my part, and I invite clarification and correction from anyone who knows better.
Duke |
Thanks for the highly informative post, Casouza. Obviously I was wrong in my guess about the construction of the Druid, and I appreciate your setting the record straight.
Duke
dealer/manufacturer |
I don't really understand audiophiles belittling science. That's like people who eat belittling agriculture.
Now there are discrepancies between commonly used measurements and subjective preference, but that's hardly a secret. The applicable term is "psychoacoustics", and guess what - it's a science.
If a product works and sounds good, it's because somebody got the science right.
Duke |
Ait,
Thanks for the input from your perspective - that of someone with a professional scientific background.
Here's a link that talks about an effort to better correlate measurements with perception in audio by means of a psychoacoustically-weighted metric derived from several different measurements, but focusing on the shape of the transfer function rather than on spectral changes:
http://www.audiomasterclass.com/learn.cfm?a=3651
See also Audio Engineering Society preprints number 5890, 5891, and 6888.
Duke |
Macrojack,
The issue of the SoundStage measurements came up. I posted a possible explanation. Note that Zu hasn't really described what their enclosures do, so we are left to surmise - and note also my request for clarification and correction on the specifics. Is trying to understand the SoundStage measurements and correlate what little we know about the encloure design with subjective impressions really so out of place as to be called "hijacking"?
I'm going to do it again, just so you know. Shadorne commented on a similar-looking frequency response curve in a PMC loudspeaker as possibly shedding light on the Druid's design. I'm going to reply to him.
Also, just so you know, it's not the numbers themselves that I value. It's the knowledge and insight that can be conveyed if you understand what the numbers are saying. I wish to understand loudspeaker design (and the Druids are a particularly interesting design), and I wish to share my understanding with others who are interested in loudspeaker design. If you feel I have mischaracterized the Druids in some way, by all means take me up on my request for correction. If you just object to my interjecting numbers and science in general, I invite you to start another thread and we'll discuss it.
Shadorne,
Thanks for bringing up SoundStage's measurements of the PMC speaker. I think there's some information there that may shed light on the mysterious Druids.
The PMC is clearly a tuned pipe enclosure - in this case a transmission line having a factory-specified line length of "2 meters". The midbass notch location would lead me to expect a line length of about 2.15 meters, so the specified line length corresponds pretty darn close with the location of the notch. I think the notch is too severe and too low in frequency to be related to an internal resonance. It looks like what I used to get in some of my transmission line designs. I'm pretty sure the PMC woofer has a higher Qts than the fullrange driver in the Druid, which would be consistent with the PMC's measured response holding up better below 100 Hz.
From the location of the notches, we can conclude that the PMC and Druid have similar path lengths. The PMC uses a three-fold transmission line. We don't know what's inside the Druid yet. Maybe one of these days I'll try to map out the internal partitions by knuckle-rapping.
There's something else we might gain insight into from a comparision with the PMC measurements. Take a look at both impedance curves. You see how the Druid's impedance peaks rise up much higher in the bass region? My guess is that's partly due to the Druid having relatively little damping material in its enclosure. Now some of the difference could also be due to high a high mechanical Q (Qms) of the Druid's driver, but even a little damping material in a tuned pipe should bring those peaks down considerably. I'm not saying there's no damping material in the Druid, but what's there is evidently not very effective at low frequencies (intentionally, I would guess - the impedance curve may be an indication of the design's muffler heritage).
By looking at both speakers' frequency response and impedance curves, I think you can see a family resemblance. Something similar to the PMC's transmission line - but obviously not identical - seems to be going on in the Druid's enclosure.
Duke |
Kehut,
The original poster said that the Druids "have peaked my interest". I didn't read that as excluding interest in all commentary other than a direct comparison with Usher and/or Klipsch speakers.
And... well, I don't know any other way to say this. Since when is talking about speaker design on the "Speakers" board in a thread about a speaker with an intriguing and mysterious design "grandstanding"?
Kehut, not everyone shares your belief system that "scientific measurements dont offer any help at all as to the sound of any speaker". (Have you run that by Sean and Adam?) I guess you missed it, but what I was doing was explaining to the best of my understanding (given the lack of published design details) how the Druids can still have decent-sounding bass when they measured 25 dB down at their rated low frequency limit (38 Hz) and have a similar notch at 150 Hz. On paper, that really is abysmal performance. Come on man indeed... I was trying to keep your speaker in the running in the eyes of someone who may still give science some credibility.
Duke |
Kehut and Macrojack, glad to see we can still be friends even though if we don't eat the same vegetables!
Flaudio wrote something that I'd like to comment on:
"My point is that you cant draw subjective results from objective data. In other words, you cant say that product "X" sounds "Y" because of "Z"."
My question is, how can product "X" have sound "Y" without there being a cause, a "Z"?
Now I can understand how your experience with the Druids seems to support the position that measurements are useless. The measurements say one thing, and the reality you observed is something else - and you have chosen to trust your ears, not the graph! Good for you! But, maybe the conclusion you arrived at regarding objective measurements is off by a few degrees. There's another way to look at it: The measurements are showing too small of a piece of the puzzle to draw reliable conclusions from.
When there's a discrepancy between an accurate measurement and observed reality, the meaurement is either inadequate or irrelevant. You might want to go back and re-read the post by analytical chemist and Zu druid owner Ait. Among other things, he points out that a synthesis of seemingly unrelated data can give a useful (though imperfect) understanding of what's going on. [I grasp at frequency response and impedance curves, allusions to muffler technology, and subjective impressions (including my own) in trying to understand the Druids' bass system.]
With loudspeaker systems, the causes of perception are particularly likely to be complex and/or elusive. I can go into detail as to why if you would like. But given enough relevant pieces of the puzzle in my opinion it is possible to assemble a useful understanding of what's going on - what "Z" is.
Duke |
F1audio, thanks for taking the time to reply. And I certainly agree that the rich variety of loudspeakers is much of what makes the industry so exciting. There's a wide variety of highly capable designers taking vastly different approaches - and perhaps I get overly fascinated by that. And often it is small companies - like Zu - whose innovations are pushing the boundaries.
I don't claim that measurements can reliably predict listener preference among dissimilar systems (electrostats vs horns would be an example), but that doesn't render them useless as an evaluation tool for audiophiles.
Take for instance the subjective assessment "speaker X sounds amazing!". That "amazing" can be expressed with greater precision: Voices and instruments sound real; you can close your eyes and hear the exact location of each instrument; you get the feeling of actually being in a concert hall; you can hear all the tiniest details even at low volume levels; the music conveys emotion and makes you want to dance; the bass hits you like Mike Tyson; and/or you can listen all day long and never get tired of it (perhaps dependent on how long you can stay in the ring with Mike!). Each of these "amazing" characteristics can be traced to something that the speaker is doing right, and it's doing it right because of its design.
"Just trust your ears" is great in theory, but what if the only audio stores you have reasonable access to are Best Buy and Circuit City? You'll have to rely on something other than your ears to decide which speakers are worth a major trip to audition, or are worth the risk of buying unheard. Reviews can be helpful of course, but once you're to that stage the more knowledge you have the better choices you're likely to make.
So the simplified version of my theory is this: If you know what qualities matter the most to you, and if you know what type of design is likely to produce those qualities and/or what measurements indicate their presence, then you can make better choices of which speakers are worth the trouble to audition and/or risk of buying without an audition (which is the circumstantial reality for many people).
In my opinion as brick-and-mortar stores become more of a rarity, "know thyself" and "know something about speaker design" are likely to become more important to the person embarking on his or her Great Loudspeaker Quest.
Duke |
