Some simple comments, which I think are reasonably correct.
(1)Speaker inductors and capacitors in crossovers do not pass through electrical signal instantaneously. For example, a capacitor is two parallel plates. Electrical charge builds up and electrons flow across. With alternating current, the energy goes back and forth. The build up and passage of energy takes time, not much, but a discrete amount of time. Therefore, frequencies can be time delayed. With time delay comes phase misalignment. So time misalignment causes phase misalignment in this case.
Other things can cause misalignments too. For example, no passive crossover is perfect. Some frequencies will come from both drivers. If the drivers are different distances from your ear, the same frequency reaches your ear at different points in time. Your ear perceives this difference in timing. The same frequencies from different drivers may also interfere in phase so you can get phase misalignment too.
(2) A sloped baffle can help to alleviate misalignment. You typically sit with your ears at tweeter height because higher frequncies are more directional. Off-axis seating is more problematic for tweeters than woofers. Sitting with tweeters at ear height means that the woofer is farther away. If you slope the tweeter back so that it is the same distance away from your ear as the woofer, when sitting with the tweeter at ear height, you can compensate for the phase misalignment somewhat.
(3)First order crossovers do not eliminate misalignment. In fact, it is the opposite. With a more gentle roll-off, and with some frequencies produced by both drivers, you will get greater overlap. Time and phase errors will be exacerbated, not reduced, in the absence of other compensating factors in the speaker design. If somebody tells you that first order crossovers, by themselves, eliminate misalignment, tell them to send their research to the Nobel prize committee for evaluation so that the world can share this revolutionary discovery in the laws of physics.
(4) The answer about coincident drivers should be obvious from the points above. If you have the tweeter and woofer radiating from the same point, one cause of misalignment will be eliminated since any overlapped frequencies from the two drivers will be the same distance away from your ear.
To supplant, perhaps, the Thiel info:
1)& 2) (Time alignment): in practical terms, the acoustic centres of the drivers (i.e. where sound originates) are aligned in the vertical plane. As tweets are typically more shallow than, say, woofers, their acoustic centre is more forward vs. the woofers'. So, sloping the baffle helps bring the acoustic centres CLOSER to being aligned. Soemtimes even, aligned. Snag: one listens to OFF axis sound (axis being the line of sound emmission from the transducer -- being vertical to the plane of said transducer).
(Phase shift & 1st order xover): all xovers intriduce phase shift -- that's what they're used for, it's part of their job, that's how they attenuate the frequencies you ask them to attenuate. The 1st order is considered less deleterious because of the gentle slope (90d /octave on paper) which, it seems, is less annoying to our ears. Snag: this puts a strain on the two drivers playing together and can cause intermodulation distortion. "Puts a strain" also means that the drivers+speaker system are/is engineered to play well in a 1st order situation: expensive or very cheap (and poorly performing).
To imagine phase, consider how we depict a sound wave -- i.e. two consecutive semi-circles on either side of a horizontal axis. Consider the crossing point on the axis as "zero" time. If the two drivers cross zero at the same time point, you are in phase.
3) There are many obvious advantages to coincident drivers, sound emanating from the same point ("single point source" i.e. "coincident") being a major advantage. There are many challenges: there is a xover anyway, intermodulation distortion, the outer cone can be a wave guide for the inner driver... (as Theil notes).
The Theil driver, BTW, is quite good and very expensive.
Ears are roughly 6 inches apart....a lot of phase correction is done when the brain combines both signals....so don't get too hung up on this issue alone, especially above 2Khz where our ears will quite likely get out of phase sound at certain specific frequencies anyway (depending on precisely where a person is seated...nobody keeps seated to within an inch position).
The lower mid range, however is an area that is more sensitive to phase issues....particularly the cross over from mid range to woofer in a three way design (two ways have an advantage here). At frequencies below 600 Hz the wavelengths are similar and then larger than the distance between our ears. Notice that at roughly 90Hz we completely lose the ability to locate sound. Misalignment in phase between two drivers in the lower mid can produce "comb filtering", a bumpy frequency response at the crossover and an "airy" unfocused sound; if male vocals lack the same focus as females then this may indicate a problem.
At 340 Hz the wavelength is around 1 foot; spatially speaking, mid and a bass woofer drivers that crossover near this frequency should not be more than a few inches different in distance from the ears (at least less than a quarter wavelength or 3 inches). This is rarely a problem if the drivers are mounted vertically....as the listener must go to extremely odd seating postions to have one driver more than even an inch different distance to the ears compared to the other driver.
Provided separate drivers are reasonalbly aligned in phase and spatially, most designers believe it is enough to have a smooth gentle phase behaviour over several octaves rather than preserve absolute phase (hence the preference for 1st order crossovers). Nothing abrupt. This approach is supposed to preserve the timbre/transient reponse of individual instruments by not smearing their partials with respect to their fundamental harmonics.
Active speaker designs have helped speaker manufacturers tightly control phase between separate drivers, gaining the advantages of separate drivers (less distortion, less beaming, less cone breakup, better dynamics and efficiency) whilst minimizing the drawbacks of having a crossover region.
Roy gave me this very sensible response to the coincident driver issue: "Regarding co-incident drivers: Any 1" dome tweeter, below 5kHz, wants to radiate in an omni-directional pattern. This means that much of the sound you would hear from a co-axial mid/tweeter driver in the lower treble comes to you secondarily by way of bouncing off of that mid's cone. Even a short horn around the tweeter cannot prevent this (because it is short). A high-order crossover can help reduce that secondary 'splash' off the mid's cone, by chopping off the low-end of the tweeter even more rapidly."
Hopefully he wouldn't mind me posting it (can't see why not).
With Vandersteen, Thiel and Green Mountain you are exploring some of the speaker manufacturers who have been most successful in coping with the inherent difficulties and problems of crossover design. There is a growing movement toward single driver systems that entirely eschew the whole crossover controversy by excluding them entirely from speaker design. Your posts suggest that you may well be interested in exploring such an option. Naturally there are tradeoffs involved but the Zu products have minimized the sacrifice necessary for such an exchange. Since buying Zu Druids about a year ago, I have become active with the company and have a loose relationship with them that has business potential. I am not a dealer, per se, but I am in a position tp profit if a demonstration in my home ever leads to a factory direct sale. Otherwise I am no different than any other audiophile with preferences. This info is offered to indicate the potential for bias in my comments due to factory affiliation.
I would recommend that you call Sean Casey at Zu and ask these questions of him. He has always been very helpful.
I have Zu Druids in my 2nd system. They are very nice speakers. However, they are not in the SOTA league of such $10K+ speakers as the new Thiels and higher-end Green Mountain C3s.
I already have great speakers in my Hyperion and Zus. I'm kinda looking for, now, what might be the be-all-end-all, truly SOTA speaker. At this point, to me, that means full freq-response, multidriver, dynamic, and time and phase-aligned.
So, yes, I'm familiar with Thiel & GM, but have never heard Vandersteens.
I sort of have two thread going now with very similar topics... sorry about that.
(Speakers which are) full freq-response, multidriver, dynamic, and time and phase-alignedThat's a tall order, my friend.
I strongly recommend you forget it: multidriver, & phase + time aligned don't go together. You will laways have phase matters with many drivers.
Well performing full-range passive spkrs are extremely expensive. Often they are not practical; heavy, difficult to drive, to place, subzero waf, etc.
If I may be so bold, ENjoy what you have. Cheers!
Paul, I own the new GMA Calypso's and they are very, very impressive speakers. I agree with others that say search the archives here for comments by "RoyJ" (Johnson of GMA) and also by "Karls" (Karl S. of Audiomachina), both very well versed on time/phase coherence. To my ears (30 year musician) a time/phase coherent speaker like a GMA brings you closer to the music than a non time/phase design. Once you own a pair of time/phase coherent speakers, you will find it pretty much impossible to go back. My opinion only.....for what it's worth to you. E-mail me if you have any questions I can answer for you about the Calypso's.
Paul, you are certainly on the right track in terms of the design principles that should ultimately lead to true fidelity in a loudspeaker. Unfortunately, there are always tradeoffs in every speaker design, and every implementation of a given principle has variables that lead to differentiating sonic results. In other words, there are no perfect speakers.
I am a huge fan of time alignment and phase coherency, evidenced by the fact that for the past 27 years I've owned a pair of Beveridge Model 2 SW loudspeakers. These are full range electrostatics with no crossover ( I eliminated the high pass X-over) save for the low pass on the subwoofer. This speaker probably comes as close to a perfectly time aligned and phase correct speaker as any design ever made, but it still errs in ways important to music reproduction.
The bottom line is, whether a speaker's design is critically time aligned and phase correct or not, ultimately it gets down to how you perceive its ability to reproduce music. I have long admired Theil's work, but at Theil's RMAF demonstration this year I was most disappointed in the sound of the new 3.7s. To me, they had a coloration to their sound that was akin to holding a long piece of tin foil between two fingerand shaking it. Everyone who heard it at the same time that I did reported hearing the same coloration. However, someone from our group who heard it later didn't experience it that way, so maybe there was something else contributing to that phenomena that was later remedied.
The GMAs are very fine speakers, and the C3s, when set up properly, rival my old Beveridges (I spent quite some time auditioning them with Roy last summer at his facility in Colorado Springs). Roy is an extremely competent designer, as is Jim Theil, but between the two, Roy's would certainly get my vote. Of course, YMMV.
songwriter72, what are you driving your Calypsos with?
Theloveman (hmm.. I feel slightly uncomfortable calling you that), I thought the Thiel 3.7s were very impressive. Not as good as the C3s, but very, very good speakers.
Sounds like you are referring to Al driver "ringiness"? It has never bothered me... I can notice it when I try, but it has never bothered me.
I've read that long thread that included the "battle" between Roy & Jeff Joseph. Kind of ironic, their opposite takes on T&P, as some of my favorite speakers are made by both. I used to have Joseph Rm25s which are very good speakers. I've also heard the Pearls, which were pretty amazing. That said, and although I'd heard them years apart, the Pearls do not stand up to Roy's all-out effort, the C3s, which, of course, take T&P into account.
I bought a pair of Continuum 1.5s used. May keep them, may not. May end up in my 2nd system. We'll see.
Heard some VMPS RM30s today at a local owner's house that were pretty impressive. Very good sound all around.