Time coherence - how important and what speakers?

Time coherence is as important as the amplitude response measurements typically taken.

A time-domain snapshot would show the pressure spreading away from that cabinet- a disturbance that contains both high and low 'frequency' components, better thought of as quickly rising/falling pressures overlaid with slower rising/falling ones. When we hear them in their original sequence, we remark "what a gifted musician!"

There are serious challenges and outright limitations to achieving 'perfect' time coherence:

-We are limited by the drivers having finite bandwidths before any crossover is applied. We need perfect pistons in the treble, and response to DC in the bass.
-A driver's electrical characteristics change with the power applied (temperatures rise), which means the crossover points, thus the phasing, change dynamically.
-We have the issue of cabinet reflections. A tweeter in a large cabinetface is like putting a woofer in a corner, speaking wavelength-for-wavelength. Even if that face is beveled, or felted (felt does not absorb 100%)

How can you tell a cabinet face is a problem? Just pick any point on that face and compute the `round trip distance for sound to get over to there from the dome tweeter and from there on out to your ear. Compare that to the direct path distance from the dome to your ear. You'll see that the path-length DIFFERENCE is greater than 1/4 wavelength of any lower-range sound from the tweeter (or mid), which means the reflection smears over the direct wave- it is not coherent.

This tweeter (and mid) 'splash' off of the front panel is 'corrected'in those large cabinetface designs by crossing over the tweeter higher than the mid's crossover point (and the mid higher than woofer)- which de-focuses the image and makes the dynamics sluggish, as now the drivers are 'a little out of phase' over their ENTIRE ranges.

If it wasn't 'corrected' by staggering those crossover points, the tweeter (or mid) would measure too loud in its lower range, as the reflections boost its 'bottom end response' when measured with test tones or pink noise, or even MLSSA- which is why this phenomena is not discussed- it's not visible with std. tests. But it is audible.

To hear what really good time-alignment and lack of reflections do for the clarity of the performance and the musicality, please listen to a single element headphone- Grado's or some Stax electrostatics, etc. And listen to a single-mic recording on them, such as a Harry James Sheffield disc- you can clearly hear what each musician is doing, in any part of the spectrum, which is the benefit of a time-coherence transducer.

Then play a crummy recording and see if it is less irritating- it will be. That's because the transducer has less phase distortion, which only distorts the original distortion. You normally hear distorted distortion, which is a multiplicative process, never additive. And why it's better to improve the source components first, before say, changing the amplifier or speaker wires.

Speakers which are very sensitive to your choice of amplifiers have a phase-response that is all screwed up- which magnifys any problems the amplifier has. Also they use a wierd crossover circuit that's causing the phase problems to begin with. These circuits also put a difficult load on the amplifier, as their extra parts store energy instead of passing it on as soon as possible.

Finally, time-coherent speakers can sound great on cheap stereos, for the reasons above, but only if they employ very simple first-order crossover circuits whose few parts can actually respond to every nuance the amplifier can muster. The evidence is heard even in cheap Sony headphones (which have little phase shift) plugged into any stereo.

Also, most crossover-circuit parts cannot pass the most delicate signals, which makes the music bland. Most crossovers also use far too many of those lower-fi parts.

We don't see too much written in the press about time-coherence, as the math is confusing at first- not suitable for a casual article. I wrote one for Audio Ideas Guide magazine, and it's still hard for me to wade through without re-reading.

The best dealer has worked hard to hear `most every brand set up well, whether he carries it or not. This industry wouldn't be where it is without those retailers (which are few).

Hope this helps. Basically, trust your ears and use them to verify that a dealer knows what good sound really is.

Roy Johnson
Green Mountain Audio

greenmountainaudio.com will be published in the next few weeks.
Roy Johnson
Green Mtn. Audio

They are not tuned to your room, but "focused" to your listening position via moving the mid and tweeter in the C-2 and new C-3. You are setting what we call the "Soundfield Convergence" for time coherency at your seating spot. This is done with a tape measure and takes a few minutes max.

The phase response of the C-2/C-3 will thus be- by definition- compromised everywhere else. But by many HUNDREDS of degrees less than speakers with higher-order crossovers.

We do note that with a first-order crossover, off-axis comb filtering does not seem to be an issue except on test tones and pink noise.

Moving completely up and out of the main listening plane (where off-center listening was still fine), we hear the depth of field decreasing, as the time domain becomes progressively "warped" from bass to treble. But at least there are no sudden "jumps" in the relative acoustic phase from driver to driver as there are with high-order crossovers, no "twitchiness".

Standing up, the time delay warp across the spectrum is still far less than with high-order-crossover speakers. We do not hear the soundfield degenerate into "a wall of sound"- plenty of ambience remains.

High-order designers are getting better at smoothing out the abrupt jumps in phase- for less "separate tweeter-separate woofer" effect.

They are also getting better at damping the ringing present in the highest-order crossovers, but the result is a hard load on the amplifier.

What you hear from the smoothing and damping is less image depth, less dynamic impact, and less rhythmic definition (finesse) anywhere around those crossover points. Which is why, quite often, the approved "audiophile recordings" used to demonstrate them are so bland performance-wise. Not much there to challenge the speakers. Something aggressive won't be pleasant- no Zappa allowed...

I noted in some postings above, references to the possible lack of phase shifts in minimal-driver speakers- single panel electrostats, Jordan module, Lowther, etc.

In a single driver, phase shift won't be caused by an electrical crossover- there isn't one!. The signal remains free of crossover parts distortions/haze too.

However, any driver has mass, suspension, and damping (by the suspension's resistive losses and the amplifier). Thus it is a "damped harmonic oscillator"- in a Physics 101 book.

A harmonic oscillator has a 1/4 wave's worth of time-delay down at its low-frequecy resonance, compared to the midrange tones. For a sealed woofer with -3dB at 40Hz (close-mic'd measured), that means 1/4 of 1/40th of a second, or 1/160th of a second=6.25 milliseconds. That doesn't sound like much time delay, but it is ~7 feet of distance, at the speed of sound.

Put two microphones on a piano- one for the left hand, one for the right; both equally close to the strings/soundboard. Now, impose 6.25 milliseconds delay between those two mics- that is, between the lowest notes and the mid-scale notes.

Imagine what the piano would sound like if the right hand tones got to the microphone seven feet sooner than the left hand's lowest notes, because that's what's happening as you slide down the scale for ANY loudspeaker- and it's a gradual change in phase, which is why we don't complain too much. Any driver does this- headphones, Walsh, electrostats, Lowthers...

A damped, harmonic oscillator also has a high-frequency limit, imposed by its moving mass- which equals phase shift in the highs, or time delay.
It has phase shift in the low frequencies, because it has mass bouncing on a suspension (it's a mass/spring system), as described above.
And since it cannot be an infinitely-rigid cone, it has cone breakup too, which imposes a ragged phase error across the roll-off region, a raggedness that changes with loudness too.

If it has a whizzer cone for the highs, like a Lowther, then there is a time-delay (phase shift) between cone and whizzer, seen as a wiggle in the driver's impedance curve. At that mechanical crossover frequency, the idea is that the cone stops moving as the whizzer starts moving.

Yet the amount of time delay between those two parts is far more than some electrical crossovers would've imposed. Even as the whizzer moves, the cone is also breaking up- parts of it "rattle on" in non-pistonic motion, so the phase change is not smooth with frequency.

Finally, since the forward edge of the whizzer is un-terminated (not damped or otherwise constrained), it has its own breakup modes. Which makes complex, loud, high tones sound hazy, fizzy, fuzzy or dirty (depending on the whizzer's breakup modes).

Any mechanical transistion also changes its characteristics with loudness, humidity (possibly) and aging of the materials. You are asking a piece of paper, plastic or glue joint to flex, predictably, for billions of cycles (per week), and flex in a completely linear, proportional manner on the very softest sound and the very largest- often simultaneously.

A mechanical transition is also happening at the leading and trailing edges of the ripple moving down a Walsh driver, or spreading out across the face of a Manger diaphragm. It is hard to find driver materials that do not change very much in flexiblity with age, or humdity, or loudness- which means the designers of those two drivers were truly ingenious to get as far along as they did. What those drivers offer is minimal phase shift in their mid-bands- which is good. But neither one can handle low bass, nor is very sensitive.

To check the transistion to whizzer (happens in the high-voice range), see if something abrasive such as Janis Joplin, can be tolerated in that tone range. Then listen to her on a good headphone (has no phase shift in that tone range). The whizzer transistion could be apparent on massed, loud strings- as wiry, steely, or strident- it all depends on where that mechanical crossover point is in the tonal scale, and how much you aggravate it.

No single driver can cover the whole audible range including low bass, unless it is a large single-panel driver, for which you have to sit exactly in the middle- exactly.

The Lowther drivers and Jordan 5" drivers do have some bass, but not enough to balance out the voice-range when listening > 10 feet away, and they have loudness restrictions: Their high efficiency comes from low moving mass, due to a short voice coil = minimal stroke available for midbass and lower tones. Look up the x-max specs on the drivers- you'll be surprised.

If a design has a mid cone with no electric crossover, yet the tweeter does, then at the acoustic crossover point, you have:
the electrical phase shift of that tweeter's circuit,
plus the phase shift caused by the tweeter's having its own low-end resonance,
plus the phase shift and ringing at the mid cone's breakup modes(indicated by wrinkles in the impedance curve of that mid driver),
plus the emf sent back into the amplifier from those extra cone oscillations, which gets into the negative feedback loop.

If you have ANY kind of frequency-response roll-off, then you have phase shift (time delay) that gradually comes on as you approach those roll-off points, no matter what object in your stereo, or in the recording studio you examine- amp, mic, mixer, A/D & D/A converters, analog tape recorders, disc-cutting lathes.

But all those devices have very little phase shift in the main part of the audible range. What they do impose comes on gradually, octave by octave, as you approach the devices' -3dB points, with the exception of certain microphones, like a Shure SM-57/58, that have a ragged phase shift in the sibilance range- lending a hard edge to the voice, often intentionally employed by the recording engineer.

It is the speaker's phase error vs. frequency that is much higher than anything else in the recording/reproduction chain. It is caused by high-order electrical or mechanical crossovers that "twist" or warp the phase in the mid-bass or low treble- wherever those crossover points are. It is caused by cones breaking up (or going soft, ala KEF and B&W) in the middle of their range, before they even get close to crossing over to the next driver. It is also caused by the drivers not being the same acoustic distance from the ear.

To find out what effect any crossover (mechanical or electrical, or combination thereof), has on music, listen to simple sounds that move through those frequency ranges- there you lose depth, clarity and dynamic expression. Also listen to a lot of musical instruments in person, up close, perhaps at a music store on a slow afternoon. Talk with the store's percussionist- let him show you why musicians pick certain cymbals, bell-trees, drum kits, sticks, mallets. Have the guitar person play you some differences in his gear.

Any speaker designer worth his salt needs to know, quite intimately, what goes on in the studio, in the musician's hands. After all, that's what needs to be heard on the other end.

Roy Johnson
Green Mtn. Audio

Phase shift in the bass is a given.

When you suspend a mass, so it can move, but with proper damping (so it does not vibrate on forever), you've created a "damped harmonic oscillator"- the term in a physics or engineering book.

When one tries to drive that mass with a "tone burst" signal (which sounds like "OOOO"), the mass takes time to reach full stroke. How much time depends on how high or low on the musical scale the "OOOO" lies. And it takes the same amount of time to then stop. After all, the energy didn't go anywhere- it just got delayed. Late start = late finish.

Whatever the amount of time delay, we also call it "phase shift" (# of degrees, 360/cycle).

A "perfect" moving system always has 90 degrees of phase shift at its resonance- a frequency calculated by using the mass, compliance, and damping values. This is in physics 101 texts.

A smaller woofer in a sealed box, properly damped, has a 50Hz resonance (50Hz = 1/50th second per cycle). It also has 90 degrees of phase shift at 50Hz, which means it has a time delay at 50Hz = 1/4th of 1/50th second = 1/200th second = 5 milliseconds delay.

Compared to what?

Relative to the time delay in the midband of that woofer (assuming it has a decent cone and no crossover). Approaching the midband range, the amount of delay declines to only a few percent of the test-tone's period. When we get ~3 times higher than the resonant frequency, at ~150Hz the 50Hz woofer would exhibit ~4/100th of 1/150th second delay, ~.25ms delay.

And since time is proportional to distance traveled, then for sound, 1ms is worth about 13.5" of travel. Thus at 50Hz, the 5ms delay is about six feet. At 150Hz, 0.25ms delay is ~4".

This means the lowest bass tones are heard as emerging from another "woofer" 5+ feet behind the real woofer's upper bass/lower mid location. What that does to the formation of a sharp image or to any transient or harmonic relationship, one can imagine, but we do often mistake that 5ms+ delay for room problems (which have similar 5ms+ delay to/from the walls).

The time delay gets even longer when that woofer is equalized (like most powered subs). Throw in the sub's crossover and it only gets longer still (and rings)- which is why you see someone drag a sub all over the room until it "blends".

If a woofer is underdamped, it resonates on for several extra cycles, which again is nearly always mistaken for room problems because those extra cycles arrived late- from the cone, and from the walls. That underdamped cone also reaches its full stroke later- so it sounds sluggish or "behind the beat". Then the walls get those delayed sounds and put their own delayed reflections on top of them. Then add on the effect of multiple woofers headed to you and to the walls, as they are all differing distances too!

That was for sealed woofer designs. Perhaps the panel speaker designers would explain what they have to deal with.

If the woofer is ported or is mounted in a "transmission line" (a big mis-nomer), or loaded by a horn in the front or rear, you still have the same sealed-box-woofer time delay relationships for the sound from the front of its cone.

You also have the same time-delay relationships for the sound emerging from the port- as it's just a different mass bouncing on the same spring. And its motion is also inverted in POLARITY (not "phase") compared to the motion of the front of the cone.

When we measure the combined port/cone output using pink noise, or by MLS, or using steady sine-wave tones, what we "see" is 180 degrees of shift at the frequency which coincides with the port's max output. At a 50Hz port tuning, that would be 180/360 (=1/2) of 1/50th second = 1/100th sec = 10ms delay.

That is not what we hear.

We hear two different sources separated by the time delay from the extra distance over to the port, and by the extra time it takes to start to move the air at the end of a long "transmission line". And we hear that one of those sources is inverted from the other.... all which means less definition.

We wonder why speakers aren't perfect!

Dr. Butterworth developed the math predicting the response and phase shifts of electrical filters, and Dr's Theil and Small first applied that math to the moving system of speakers.

Wayne asks, "What effect does this phase shift have on bass definition?" Well- studios have to listen to it too, from their monitors- and thus mix for "it" on their pop/rock/jazz recordings. Only on classical, or other recordings where things are left alone (sheffield, telarc, delos, chesky, etc) can an approximate standard of reproduction in the bass be obtained.

We reduce phase shift/time delay at 50Hz ONLY by lowering that woofer's resonant frequency- by adding mass to its cone. Or one can change to another woofer having the same mass but higher compliance (softer suspension), a larger magnet and cabinet. Or use a larger diameter (= heavier) cone that has the same or higher suspension compliance, with a bigger box.

No matter how we decrease the resonant frequency, we hear tighter midbass, "faster bass", more spaciousness (ambience), easier room placement, a more realistic bass image, and better voice and highs. All but the latter two are from creating a better time alignment between the harmonics and their fundamentals. The last two are effects from bass output that reaches farther, as loudly, down the scale.

Of course, using a heavier cone with a longer-stroke (heavier) voice coil decreases efficiency. Which means we turn up the volume. That extra power means an even hotter voice coil on any peaks or sustained loud bass. The heat increases the voice-coil's resistance momentarily, which means less amplifier power is delivered, which means more "power compression". This sounds and measures like a softening of peaks and of any loud, sustained bass.

Some drivers have extra-large diameter voice coils that heat up less, as they are larger radiators of heat. Yet a larger diameter coil is heavier, which reduces efficiency. Unless the winding length is shortened to keep mass the same- which reduces stroke. Also, with a large voice coil there is more surface area in the voice coil gap over which to spread out the field from the magnet, so efficiency decreases, unless you use a huge magnet. A large coil also means less room for pleats in the important rear spyder suspension- and so the cone rocks more. The voice-coil gap has to be widened to prevent that coil from rubbing or jamming, which reduces magnetic field strength, thus efficiency, even more. But at least the large coil doesn't burn out, and it makes for good advertising...

Higher efficiency woofers are more efficient because they have they have less mass- usually via a shorter voice coil. But that means they run out of stroke, and won't play loud. And if you could reduce the cone mass so that one could keep the longer voice coil, then you often have other problems from the lighter cone. Also, for a truly lighter-weight cone/voice-coil combo, one must use a VERY compliant suspension to keep the resonance down at that same low frequency, and to keep the high-end response from tilting up, like a PA speaker's woofer (mid). Yet suspensions are already as compliant as can be made consistently.

Wayne asks, "Has anyone heard phase coherent bass?"
Yes- we all have- but only from any live instrument without a PA system, and from any live voice. Which is why we can hear benefits when we reduce phase shift- it sounds closer to the real thing.

Wayne asks, "If there is a phase shift and you can not go back in time, is it possible to phase shift the good guys to equal the sluggards?"

You bet- digitally time delay everything else in the speaker by the appropriate amount, so that the low bass is not "behind" anymore- and it does help a lot! But the correction machines cannot read the subtle time delay problems that occur higher up the scale- from bad cones and from cabinet reflections, so we wind up "tweaking" by ear. Sigtech-type devices also cannot separate the woofer from the room at the lower octaves, nor "fix" the port inversion problem. In fact, these devices "cutoff" in the midband, before the room starts to confuse their measurements.

If you are going to do the digital delay PROPERLY for a speaker, you have to do it for ONLY the sound which comes right from the face of each cone- no cabinet reflection correction would be allowed.

But ask where does one stop the correction? Besides the speaker's woofer, we have bass time delay from your phono needle, phono stage, any coupling transformers/caps anywhere in the chain, and from the analog master-tape copies, from the original analog master tape, and from certain mics.

Digital mastering (DDD) threw out all that bass phase shift except for three: your woofer, any transformers/caps in the chain, and the mics. This is one reason digitally-recorded bass is better in many ways than analog.

Good questions, Wayne. Complicated answers, sorry- but that's why you do not see "time delay" or phase shift properly covered in the press. The above is taken from what is being prepared for our website.

Best,
Roy

Thanks!

Regardless of one's opinion about motional or positional feedback applied to a sub, both suffer from the limitations of the accelerometer used, the second voice coil being 'read'. Both are transducers, thus with their own dynamic range limitations, frequency response, resonances, distortions.

Also, the sensor often cannot not pick up motion from more than one direction/dimension- it can't see the cone "rocking" very well, for example- it only sees that it has not stroked far enough. However, a 1-D sensor is cheaper. And marketable. It can help a poor woofer or bad box design.

It's far better to design the best possible woofer and put it into a proper cabinet (not easy), and get the phase accuracy of the electronic crossover correct, by choosing certain slopes of the filters:

Use a 24dB/octave slope on a sub and 12dB/octave on the speakers, if they are sealed or ported, and if the sub is not a resonant bandpass design (THX markets this standard pro-sound crossover as something unique). The crossover point belongs near the impedance peak of the sealed box, or near the upper impedance peak of a ported box.

Other option: If they are panel speakers/Quads- don't give them a crossover. Just put 6dB/octave electrically onto the subs only (must be stereo subs for this).

Either way, you get better phase alignment all through the crossover range- which SOUNDS LIKE far less room problems.

To fine tune either: listen with the sub at the same tape-measure distance from your ear as the main speakers' woofer centers. Then move the sub +/- 12" front-to-rear off that plane/arc. As you do so, listen for dynamics and then for tonal uniformity (separate tests).

Listen to something steadily percussive at the crossover point- a kick drum for 50Hz, a floor-tom drum at 80Hz. You will hear the best dynamic "alignment" as you change front-to-rear position and fuss (a little bit) with the crossover point.

Then listen to a string bass run the scale- try Christian McBride on his "Gettin' To It" album, or use a celloist's solo. This lets you find general tonal weaknesses/boosts most easily. Changing the crossover point may not help much, but try. Usually if there is a drop/boost in output in a range of frequencies, that's the room, as that is how rooms behave on music. To address this, pick up the entire speaker, sub and chair layout, and move it out further into the room by 18 inches and hear what happens. Don't touch the volume control or crossover settings- one variable at a time.

Your question about "proper" digital correction:

Walk around a speaker and ask what sort of things need correcting? Can you point to them?

Do we wish to correct digitally for cone breakup? But that breakup "nature" varies with the music's dynamics and tonal spectrum, and that we cannot measure with a test tone!

Do we wish to correct for a floor reflection? That is heard differently by ours ears than by the mic, so the mic's signal is not "accurate"- hence the LF cutoffs of digital correction. This explains a little of the higher crossover point choice: it narrows dispersion and thus reduces cabinet and floor reflections in that higher tone range.

Is it a "splash" reflection of many simultaneous tones, coming off a big, curved front face? If so, ask what is the 1/4-wave dimension of the panel? Because from that tone on up, you get the "splash"- a new bubble of sound launched off the face- and that happens all the way right down next to the dome of the tweeter. This is what the digital units try most to correct for in the direction of your chair only. But that leaves many driver non-linearites uncorrected, as to the mic, those were swamped by the splash. The mic cannot separate them from the splash the way our ears can.

With digital correction, what is going to happen to the sound everywhere else in the room? Not just what others would hear, but what will be coming back to my chair from the walls? Whatever that is, it can't be zero. And it's definitely not "corrected".

One has to ask ALL the right questions of the digital correction situation. Make the speakers better in the first place is my answer. Costs less, sounds better. Then try correction.

As far as placing the speakers out into the room? Yes, if they were designed from scratch for that location. You cannot have it both ways- i.e., a speaker design accurate when placed within a few feet of the wall, vs. many feet out. As that distance from the back wall is varied, the perceived output below 300Hz changes, unless the speaker is the size of a `fridge. This cannot be balanced out with a switch on the back of the speaker. And in my opinion, it cannot be corrected properly by digital means.

Actually, it's not my opinion, but what physics says we are still going to hear come off that back wall (even with digital correction). To any speaker designer, the first question after, "What kind of room do we have to work in?" is, "How far out from the walls can most users live with these speakers, so they can actually hear the depth of the image?" The third question is always, "How high up the scale is the woofer to go?"

There IS a distance from the back wall which is "far enough"- the Hass effect is an indicator of that distance.

Sorry I could not respond sooner- too much work. Website coming soon.

Best regards,
Roy
Green Mountain Audio

A side note: Anyone recommending a really wide-set speaker placement is throwing a bigger acoustic shadow on your opposite ear. Why do that? When they are set up in the normal 48-53 degree spread (the limits of our peripheral vision by the way), their image is unstable as you turn your head- usually because of double drivers, multiple drivers, or line sources.

As you turn your head, there remains a constant acoustic field UNDER YOUR CHIN, because you have a chest making lots of nearfield reflections. But for sound coming over the top of your head, the field is determined only by the direct sound from the speakers. And if there is sound being launched from higher than your head and also from the same height simultaneously, then the opposite ear hears more of the "higher" driver location as you turn- and the image jumps to that speaker. If the source of the sound is more compact than your head, the image does not jump.

Listen to a solo voice with your eyes closed, no eyeglasses, no coffee table in front of you, nor footstool, in a very quiet room, to identify any image jump most easily.

Audiokinesis- thanks for your kind comments about Jeff and me! Jeff and I have yet to wrestle behind the Alexis Park- we'd buy each other beers instead.

I find many of the smaller speaker designers (smaller than corporate entities such as Vandersteen, Theil, Martin-Logan) share tips and ideas. We'll give out supplier information, talk about how to join a particular wood, a brand of power tool not to buy, or a better way to pack for shipping. In tight spots for parts, some of us have been known to help out another. This does not happen in amplifier design or in digital or turntable work (it would not be kind to speculate why).

Consider that
-Jeff and I know that truly professional loudspeaker design requires a lot of hard physical work done in isolation, in many different areas from materials science to field theory.
-We rarely sit down with someone who has actually experienced and understands the difficulties of bringing any speaker design to full production.
-We read research papers of all sorts, looking for an insight on a particular measurement technique, or to find if a certain type of cabinet loading has pitfalls the researcher missed, so we don't have to stop and make that measurement or cabinet ourselves.
-We learn that materials and drivers suppliers don't know enough about their products to help us.

We share tips because advice is worth A LOT from someone whose judgment and experience we respect. Besides, any tip about who makes a good acoustic felt is never going to upset the competitive balance anyway. After all, we're taking on the big guns of "speaker design", with their not-too-insightful designs that border on outright laziness. We are the Panoz up against K-cars. ~twas ever so... The big firms attract customers who drop big bucks, only to be bored with the music; sound so un-inspiring that their friends hear there's no way they'd EVER spend that kind of dough.

There are no schools for speaker design, and no peer review as in other science or engineering fields. It doesn't help that magazine writers aren't technically competent like they were 25 years ago. So, just because you are the head of B&W, Bose or the Candadian Research Council, doesn't mean you know what you're doing. Nor will anyone find out... remember, few people can call us out on what we really know about sound.

Besides, when any design doesn't "sound quite right" or perform well on aggressive music, it must be the cheap amplifier or the dreaded "poor recording" and of course the room at the stereo show.

It couldn't be that the designer hasn't ever looked at time-domain math, or at cone breakup (putting in notch filters!!!), or looked for shear vibrations in the standard folded-up cabinetry that their cabinet shop assured was "the best way to make the cabinet". Or hasn't even looked up the absorption coefficients of the internal stuffings to see that his crossover is "correcting" for a 1/4-wave internal cabinet resonance that shouldn't have been there in the first place. Nor has he done the 6th-grade wavelength math that shows how non-parallel cabinet sides don't do anything to supress internal echoes. He only has to claim they do, `cause that seems to make sense- and he knows so much more than the listener or reviewer!!

There are poor speakers because there are poor designers. If these were cars, many of them would have five wheels, and three would steer!

So, the more customers that move over to Jeff's and my speakers (and to Soundlab), the more they'll enjoy the music- which keeps them customers of ours, and their friends in the market too.

There will be time to discuss some more of the ins and outs of speaker design- my thanks to all who have posted, because they at least listen, and think about what they are hearing.

However, anyone interested in understanding the art of speaker design needs to get a good grasp of the fundamentals of soundwave propagation. Step one is to know wavelengths vs. piston diameters, necessary to understanding the reasons for directionality, reflection, and "radiation resistance".

As an example, the designer who espouses a tweeter should be placed in the center of the mid's cone, ignores 6th-grade math that clearly shows that, in the crossover region, any tweeter will always try to be fully omni-directional (the wavelengths there are 3-7 times the dome's diameter!). The truth of that omni-directional "mathematical assumption" is always verified on sine wave tests, on noise tests, on impulses, tone bursts, on TEF, MLSSA, indeed on EVERY test.

Thus, being PROVEN an omni source, any tweeter's low end will ALWAYS "splash" off the mid's cone. Which means the designer will always screw around with the crossover, sucking out the bottom end of the tweeter until it "sounds OK on Holly Cole". It should also sound OK on Janis Joplin, and Billy Holiday, Dinah Washington, James Brown, Screamin' Jay Hawkins, Willie Nelson, Garth Brooks, the Klezmatics, Britten and Barber and Stravinsky, and the three tenors, and Tool, Primus, Metallica, ... but somehow it never does.

That designer needs to justify putting the tweeter in the center is a great idea, so he will manipulate the crossover to make 2 or 3 types of measurements look good, make the speaker sound fine on non-aggressive music, and ignore the other tests and other music.

It is also easy to say "we are not sensitive to phase", and to claim that after all, "the recorded sound is mixed up in phase before it ever gets to the speaker". There are several holes in that logic that I see. Can someone in the forum point them out? And it's not because the word "phase" is used instead of "time".

No reviewer will ever call a designer on his technical claims, from politics, from a lack of education. The designer of the concentric-tweeter model will then proudly display the complicated crossover that makes the design function on those few tests and recordings that "prove" how good it is. And the giant "wall of sound" that emerges is unlike anything else- so it must be better, as it's from a wide-selling, respected designer! And different is often confused for better- until we play enough variety of music (another thing that isn't done during reviews) to hear that speaker's signature.

I prefer time-coherent designs, simply because they are more revealing of the musical intent. On non-minimum-phase speakers, you listen to the separate parts of the sound, as the speaker picks it apart in time. The "separate tweeter" phenomena is one example, another is the image sticking to the face of the speaker (lack of depth), and another is the sensation of height- a tweeter out-of-phase artifact. But the real test comes on instruments with harmonic structures that span the crossover region. Without time-coherence, they sound flat, lifeless, definitely not "real".

To avoid listening to the sound of the sound (can I say that?) instead of the instruments, listen to ALL sorts of recorded music, at all loudnesses, on all kinds of stereos- ESPECIALLY music you don't much care for nor ever will. Try to experience many different types of live performances, from unamplified acoustic music in a living room, to a marching band. That's how you become an experienced audiophile- by knowing sound.

Thanks again to Audiogon for providing a place for such a forum. Jeff and I will be out back behind Audiokinesis' store taking apart a Soundlab if you need us.

Roy Johnson
Green Mountain Audio

PS: something to think about for what's "real sound" from a speaker- here, the working definition is "the perfect speaker gives us the clearest, single pinpoint image from each mic." ~cause the mic itself cannot pick up more than one dimension of the soundfield- and that dimension is distance. Which is time-arrival differences. Which is another indicator of how time-domain information is indispensible in enjoying music, and in reproducing the clearest pinpoint image from each mic.

We are stuck with pinpoints as "perfection". If you don't hear specific pinpoints (images that have no height, no width, only depth) in ALL the tonal ranges of the music, then you are hearing time-domain distortions in those ranges where the pinpoint is smeared out (up to "life size"!). Even in the nearfield of the time-coherent Soundlabs and Wisdom Audio ribbons, all we hear are pinpoints, especially when we remember to close our eyes to allow our ears to better function as location sensors. And you'll find that even surround sound is only 2-D, with only depth and angular location as the dimensions.

Cdc/Unsound you are too smart for your own good...

Cdc asks:
- Is time / phase coherence maintained through the whole recording / playback process so the speaker is the only thing messing up phase and time coherence?
The answer is no, but... ALL gear and speakers and mics add time delays at the high and low ends of their operating bandwidths. Those are (fortunately) gradual changes in the phase vs. frequency. ONLY speakers (and certain microphones) will jerk the phase around abruptly in the middle of the audible band.

- What if a driver naturally rolls off greater than 1st order. Does the crossover have to boost the driver? Is this bad?
Any driver is rolling off because it has a "mechanical crossover", which is usually > 2nd-order. It rolls off because it has moving mass and compliance. The compliance is the flexibility of the hinge-points on the cone as it breaks up. The moving mass is varying with frequency- the outer portion of the cone coming to a standstill.

With less effective moving mass for that voice coil to move, the frequency response peaks before it rolls off- you can see how the wiggles in a raw-driver's impedance curve correspond to the breakup modes, and to peaks and dips in the frequency response.

Cone breakup is a "dirty" or "ringing" sound at worst. Breakup modes are always excited by the fundamentals down at 1/2, 1/3, 1/4, 1/5... of each breakup frequency. A driver should not naturally roll off `till well beyond the actual crossover point. Any breakup modes should be well-damped so they don't ring.

Cone breakup can be disguised with very high damping- such as with the woven Kevlar and carbon fiber cones, and the mineral-filled poly cones. It keeps those drivers' frequency responses smooth, but crushes the dynamic response as the cone structure is absorbing the energy, instead of moving the air. So why use those cones?
a) They don't have the "peakiness" of poorly-designed paper cones.
b) They look hi-tech.
c) Making proper wood-fiber cones is a dying art.
d) Plastic cones are good for sales in tropical environments and cars.
So we lose dynamics, and inter-transient silence, and detail- who cares?

- Would the extra power handling of the voice coils in 1st order x-over degrade performance in other ways like higher inductance, mass, and hysteresis?
Inductance- nothing we can't adapt to.
Moving mass- higher means less efficiency (see my post today on the High Efficiency thread) and eventually more power compression.
Hysteresis- depends on what you're talking about. The longer voice coil does overhang into more of the fringe-magnetic field created by a poorly designed magnet structure. There are others, such as inadvertently changing the center of gravity of the cone/voice coil assembly- which will lead to unwanted rocking motions, which puts really strange eddy currents back into the pole piece, which...

Unsound,
Your observation that "active speakers don't necessarily need to be self contained" is true. It would have all the advantages you list.

But if a speaker manufacturer merely supplied or recommended a certain electronic crossover, then the customer must purchase several "matched sound quality" amplifier channels and interconnect cables and speaker wires. And those are very strong points for putting all that stuff back into the speaker cabinets.

Either way, my main problem: whose gear do we choose- and will they let us? I've spent a lifetime working with speakers, not amplifiers- the only amp I could design would be from a cookbook recipe, or I'd have to mimic one already out there. And we'd have to service it! Egads..

Thanks for the insightful comments and questions. I wish the press would talk about this stuff. Why do you suppose that is? (serious question for another thread)

Best regards,
Roy

von Schweikert designs are not time coherent.

Yet, Albert makes very nice sound because:
1) He knows how to get the most out of the 4th-order approach (the only one that makes sense if you have to/choose to give up time coherency).

2) He makes a good cabinet- with proper placement of damping materials inside (it sounds like), and they are strong, fairly slim.

3) He chooses linear drivers, unlike most of what is out there.

4) Really a part of (1)- Albert is one of a few designers that knows mathematically and empirically most of the factors that influence the selection of the actual acoustic crossover points. Not too many do, as it takes years of homework in Physics, and then years of listening, measuring and designing to prove to your ears what is right or wrong with that Physics-based math. Electrical engineers are not taught this math (since the world went digital)- which is why there are so many non-musical speakers out there.

Recommended recording- Hugh Masekela's album "Hope" (CD). Stunning dynamics and excellent musicianship! The Burmeister sampler uses one of the cuts found on this.

Hugh M. is a horn artist, so this disc is not pretty at nightclub levels on non-time-coherent speakers that have a crossover point in the 2-3kHz range. We'll play it in `Vegas on our Continuum 3s, at the THE Show.

Remember, Mr. Bischoff, that there are many physical, electromagnetic and acoustic factors behind achieving time coherence. Those specific design factors are a checklist of well-known concepts, any one of which ignored means no coherence.

They include factors most of us have heard about:
uniform diaphragm motion,
small diaphragm/cone/dome size vs wavelength,
diffraction reduction,
crossover time delays,
time delays caused naturally by the enclosure tuning.

What a layman lacks is a rule of thumb, some numbers, for when "uniform" is "good enough" (as nothing's perfect), or a diaphragm size is "small enough", or diffraction low enough, or what the actual acoustic delay times are vs. frequency.

Those give enough information for a layman to choose music that will reveal a speaker's flaws immediately- some of which he probably can live with. This info will be on our website shortly, along with a "recommended recordings" list that makes sense to challenge a system with.

In the long run, Mr. Bischoff, you would find that phase error in speakers affects dispersion vs. frequency. It also affects dynamics vs. frequency, and imaging vs. frequency. It affects hearing the rhythmic pulse, the emotional inflections. That is, if one plays a wide variety of music. But if a listener sticks to tried and true recordings only, then he is not investigating very far into any problem with the system or speakers. Using a wide range of material with which you've become familiar speeds up speaker evaluation and system tuning.

Anyway, thanks to all for their kind comments, and Happy New Year to everyone at AudiogoN!
Wish for Peace in the new year.

Best regards,
Roy Johnson
Green Mountain Audio

No dealers very near to you right now, although a lot of owners. Dealers down in Orange County and Central Coast area.
Call me if ?? 719 636-2500
Thanks,
Roy

Greenmountainaudio.com is almost done.

We leave to display our speakers at the T.H.E. Show in a couple of days. Don't know if we'll get the site up before I leave, or while I'm gone, but soon...
Best,
Roy

If you mean digital "tricks", I think the best answer is that the value and limitations of any digital manipulations could be more easily ascertained if they were not trying to correct for things they should not be- like cone breakup, cabinet reflections, gross phase shifts, interior resonances- non-linearities which should not have been in the speaker design in the first place.

If you mean "servo control" of the subwoofer, I have not seen that applied yet to anything close to an already linear design (a well-behaved subwoofer driver and rigid cabinet). I also have never seen anywhere close to the best accelerometers being employed... so expensive! Again, we really can't hear all the benefits and the ultimate limits of servo technology.

Maybe someone who designs servo subs for a living can come forward and explain more on that subject. Anyone out there up to that?

Best,
Roy

No. Yes. No. They could be with appropriate digital time delays (delays that are different for each frequency) built in. Those are available as studio-oriented x-overs with built-in adc/dac's.

Roy

Cdc, it is time to dispel the myth hidden inside their statement, as many other firms say exactly the same thing.

You say Revel states, "The crossover networks . . . maintain a 24db per octave, 4th order acoustic response..."

FYI: Acoustic response- this is the actual frequency vs. amplitude response a mic would measure, freefield at one meter on swept sine wave tones (no floor bounce). This is the right way to describe any crossover- by its final acoustic response (assuming you even listen at one meter... which is a whole other problem).

". . . the steep filter slopes ensure good acoustical behavior in the crossover regions, with a minimum of acoustical interference,.."

This is true -no question- when measured by swept sine waves or pink noise, w/o the floor bounce. FYI: "good acoustical behaviour" means there are major no peaks or dips in the frequency vs. amplitude response on swept tones or pink noise.

"along with low distortion and wide dynamic range."

Yes, because 4th-order rolloffs keep the drivers protected really well from low frequencies- the upper-range cones/domes don't even wiggle on a bass drum.

"The somewhat steep 24dB per octave slopes also provide the benefits of keeping ALL DRIVERS IN PHASE AT THE CROSSOVER POINTS.."

Yes they are in phase, which is a benefit, but THEY DID NOT START AT THE SAME MOMENT, NOR WILL THEY STOP AT THE SAME MOMENT. THEY ARE NOT TIME COHERENT.

With a tone generator, feed that woofer and mid a steady sine wave at the frequency where the crossover occurs. Put a mic out front- look at the combined single sine wave from the woofer and mid on a `scope. You cannot see the beginning or end of this steady tone- it's just a sine wave going up and down.

Then unhook the mid- look at just the woofer. Then look at just the mid. THEIR WAVE'S PEAKS AND VALLEYS LINE UP- they are in phase. But if you could see the beginning or the end, one starts A FULL CYCLE LATER THAN THE OTHER. And when they stop, ONE STOPS A FULL CYCLE LATER THAN THE OTHER. You could even say their combined output rings at that crossover frequency.

ALL of that is in any second or third-year electrical engineering "Filter Theory" book in plain English- that is the behaviour of the 4th-order filter. In phase, yes, but 360 degrees out of step.

Can you hear this? Yes. Just listen to a speaker (or headphone) without that time delay. How much time delay was imposed? Exactly one full period of the crossover frequency. If that was 400Hz, then the time delay between the two drivers is 1/400th second, or 2.5 milliseconds, which is ~32 inches for time of travel, acoustically.

You just smeared the guitar spatially by ~32 inches, front-to-rear, and transiently by 2.5ms, across its 400Hz range (just above middle C). It sounds like the upper strings of the guitar are "leading" the lower strings, or that there is more pick on the string. It also changes the wave envelope, which means a loss of clarity. And all that means it changes the musical message.

But how would you know unless you've heard the real thing? All you really know is that there are many "poor recordings" you can't play, can't enjoy. Many performances "you just don't get". Why? Because that phase distortion is distorting everything that comes into it- including any recorded distortions- so you hear distorted distortion- which is multiplicative, not additive.

You are being presented a speaker that warps not only the image of the guitar, and its dynamic attack, but mis-aligns the individual tones that go into shaping its harmonic envelope, which is its timbre- the very nature of why it sounds like a guitar.

"...The steeper fourth order slopes, however, avoid the power handling problems associated with first order crossover networks."

Yes, if you use less than the best drivers.

Cdc, you remark, "and comments made above about bad in-room frequency response (some Revel is very good at) with slow roll-off x-over design 4th order sounds very convincing."

I am not sure what you mean by "(some Revel is very good at)", but all you have to do is listen to end the debate. Music waveforms have little to do with the math Revel and others use, math based on sine-wave arguments, "proven" by sine wave measurements.

The standard response to all that I've stated above about the drivers not being in sync is, "Well, you can't hear the phase errors, anyway. There are tests that prove that."

The tests don't prove that- they are ambiguous at best, because they were based on clicks and other unfamliar sounds.

If a designer has never built a speaker that is minimum phase (the standard term for any time-coherent system), he has not even tried to hear the difference.

A 4th-order acoustic rolloff is used because:
--the designer believes the sine-wave arguments,
--it allows high-tech-looking metal cones to be used, as most all of those have a severe (>10dB) peak at the cone breakup frequency, along with 5-10% distortion caused by that peak. Look at the SEAS and others' metal-cone driver frequency responses- readily available. The 4th-order rolloff chops them off.
--a 4th-order rolloff lets the designer lower costs by using less rugged drivers.
--a 4th-order rolloff is often computer-aided in its design. Sounds high-tech, and good for advertising.
--the designer doesn't have to learn time-domain math, which is quite difficult.

All the math of acoustics and physics and music supports my claims, fully. What's known about how the ear uses the time domain for image formation and timbre retrieval supports my assertations. The actual times and timing of music's transients support them. The method behind how any timbre-containing wave envelope forms as time evolves, supports them.

As far as off-axis interference? 1st-order drivers overlap a lot. So those drivers must be really good- and it's hard to find "the best". However, they are overlapping TIME COHERENTLY, `most everywhere in the horizontal plane, so there is no interference or lobing. Period.

If you stand up- yes, they start to go out of phase by many degrees. But the 4th-order circuits are already HUNDREDS of degrees out of phase, no matter where you stand or sit.

Some would claim this time-coherence stuff is not a question of right or wrong, but a matter of taste. I would remind you that keeping the speaker a minimum-phase design (not possible w/4th-order) is the ONLY way to preserve the actual shape of the musical waveform received by the mic. For inside that "shape", that irregular, non-sine wave envelope, lay all the tones of the music, all the dynamic changes, all the musicality, and all of the musical message. Why disturb it?

I welcome any attempt to refute my points on any grounds.

Most of all- just listen. Non-time coherent speakers sound like a wall of sound -no depth- as the time domain is scrambled. Time is distance. Time delay is depth. Time is transients coming and going, and tones building then decaying. Scramble the timing between bass and treble- you lose the depth. You lose the timbre. You lose the musical intent. You lose access to many, many recordings.

Time coherent behaviour can be heard in any decent headphone- even a $30 Walkman headphone, let alone the Grado or Stax. Just listen, especially to music that has energy near the crossover frequency.

Thanks. I don't want to be the constant "Answerer" here, but I really hope this helps.

Roy

I should have included Linkwitz-Riley circuits in my post previous to yours. They operate in the same ways as described, but change phase approaching the crossover point in a different manner, still totalling 360 degrees difference at the crossover point. Which makes them in phase for test tones, but out of sync for transients and anything that forms what we hear as "space".
Best regards,
Roy

Best,

Thanks for the feedback.
I don't feel bad about answering- I just don't want to dominate this thread. And it would be more fun to ask the questions. But you're right, this is my profession. Much of what I learned came through sharing, so I'm glad you appreciate the information.

My words here have not been about reaching perfection- speakers are man-made devices, and I am not going to give away any trade secrets on how to better approach perfection. My intention is really to tell you about certain design pitfalls that could be avoided, except "the math's too hard!".

And out of respect for your and everyone else's intellect, I try not to make unqualified statements- so you see the reasoning behind my logic and of others, and in 20 years come up with a new way to drive the air. Mostly, I would hope any reader leaves with enough knowledge to recognize when "marketing" is disguised as engineering. Like the claims of cones dissipating energy. The lack of educational/technical writing in the press over the last 20 years is one of the main reasons hi-end has become a confusing, poor-value hobby, full of frustration for most participants.

Anyway, thanks again to you and the others for your interest and insightful questions. So I do not repeat myself for related topics, such as why a cone breaks up in the first place, please look at my posts at "the Vinyl Engine" at

http://www.nakedresource.com//yabb/cgi-bin/yabb/YaBB.pl?board=general;action=display;num=1038342561

About the Spica? It's been a long time since I studied some of what he claimed, but I do have several AES papers on similar computer simulations done in the early eighties that do not support that claim of virtually no phase anomalies. Actually his claim could be true- because of the word "phase". Remember though, that "lack of phase anomalies" does not mean "time coherent" nor "no time-delay differences", an important distinction. See my 2nd-to-last post. Phase is not time. Phase is relative; time is absolute.

Chances are that Mr. Bau was among the first to combine the 4th-order filter with the supposed "perfect" second-order mechanical rolloff of that soft-coned woofer, plus combine a 1st-order electrical filter with the natural second-order "sealed-box" rolloff of the tweeter:
That would be 6 x (-45 degrees) shift on the woofer (-180 degrees), and 3 x (+45 degrees) shift on the tweeter, (+135 degrees). Which means they are 315 degrees out of phase- only 45 degrees away from 360 degrees- near the same as from two 4th-order acoustic-rolloff filters. That remaining 45 degree discrepency? If one considers that the woofer's cone will not be breaking up in a "perfect manner", which is true, then that 45 degree differential can be easily "added back in" from the woofer's non-perfect cone breakup. So you come up to a total of ~360 degrees phase shift- which is indeed "no phase anomalies", as they are "in phase". But not in sync.

The main reason Spicas were so spacious has to do with
a) being a two-way with good drivers,
b) the felt-work around the tweeter,
c) the shape of the cabinet,
d) the slant-back of the cabinet,
e) the lack of resonance/echo inside the cabinet,
f) there was only one cap on the tweeter, in the days of poor capacitors.
g) he was among the first not to let the woofer-cone breakups interfere with the tweeter.

Smart guy. An inspiration.

Best regards,
Roy

Karls, Phasecorrect, no speaker is without time delay in its lowest three octaves, because a moving mass on a spring has 90 degrees of wave-period time delay at its primary resonant frequency.

Also, any port's output is time delayed, from the interior time-of-travel, and from the exterior extra time-of-travel from that opening. Port outputs are also polarity inverted.

The specific reasons a properly-engineered t-line (which are few) has low distortion bass:
--The port opening is reproducing the lowest bass and so the cone is not moving very much at that resonance point- this is part of the definition of a "ported" speaker.
--The upper impedance peak always seen in a ported speaker is mostly absent- a peak due to the port's air mass bouncing off the compliance of the air in the enclosure. This, Karls, is what you are referring to. Why is it not there? From the proper application of the wool stuffing and the shape of the small enclosure right behind the woofer.
--The t-line cabinet CAN be shaped so that its rear wall generates less echo directly behind the cone- but not via the usual tapered short horn leading into the line. A smooth taper only efficiently loads the returning third harmonic (of the t-line's fundamental resonance) back into the rear of the cone, causing a serious dip in the cone's output at 3X the fundamental t-line resonance.

Also, resonance is not always accompanied by an impedance peak- there's always resonance at an impedance minima, which even a t-line has. That's the frequency where the cone is driving the port or t-line most efficiently. So a t-line is a resonator, and no more well-damped at THAT resonant frequency than a ported speaker. The ported speaker has trouble at the next resonance- its upper impedance peak, as noted above.

A t-line also often uses a very low resonant-frequency woofer. In combination with the actual cubic volume contained in that t-line, this leads to a really high impedance peak at a very low frequency, usually well below 20Hz, which can be hard on an amplifier's power supply when excited.

T-lines are less efficient, ONLY because the woofer chosen has a longer voice coil, for more stroke to reach down to that impeance minima. A longer VC means greater moving mass. It is not because they are "more well-damped behind the cone", "which sucks energy from the cone". Utter nonsense, if the t-line is properly designed, as shown 35 years ago in the AES papers, available from Old Colony Sound Labs.

Wool is used in a t-line A) to make it an acoustically longer line (saves floor space), and B) to suppress upper-bass resonances. Wool is transparent to the lowest bass- it offers very little attenuation, which means that the low bass is no better damped. This is in the AES papers as well.

The best way to think of a t-line is as a very small enclosure with a very long port, needed to tune that enclosure to resonate at a low frequency.

A ported speaker is a medium-size cabinet with a modest port length, but without much acoustic stuffing, which would close off the volume of air needed to drive the port. So, with less stuffing, the ported enclosure is "noisier". A t-line enclosure is usually much quieter in the upper bass than a ported speaker's enclosure, and often much quieter than poorly-designed sealed boxes.

From a properly-done t-line (like the old IMF's), you hear extended, low distortion bass. Which arrives so much later than the upper bass, it sounds like it came from another part of the house. And because it took a while to get up to full amplitude, it takes the same amount of time to stop. Which means this resonance puts its signature on different recordings. Which is why sealed-box woofers offer better sound- still putting their own signature out there to hear, just less of one.

A transmission line, by definition means "transmitting energy without reflection". Except that "t-lines" in speakers reflect energy back to the cone, taking several cycles to build to full resonance at the impedance minima. So a t-line speaker is not a transmission line, as the energy came back to the cone.

The only true transmission lines for speakers would be A) an infinite horn (energy goes one way w/o reflection), and B) a muffler (energy goes away and cannot return). A t-line is neither.

Best,
Roy

Phasecorrect-
If you use pink noise or MLS or swept sine waves (common for testing dispersion), most of what we see on the printout does not explain what we hear. Dispersion, as heard on music, depends on some very definite factors:
--the diameter of the driver (not so much its shape) vs. the wavelength.
--does that driver remain a rigid piston in its operating band? Most do not.
--how the reflections from the enclosure's front and sides, and reflections off the other drivers' surfaces smear transients.
--how the crossover disturbs those transients.

The real problem is that we listen to music. Look at a musical waveform on a `scope- what do you see? Do you see any sine waves, or square waves, or sharp, stand-alone impulses? No. You see an ever-changing wave "form" that has more dynamic range than the face of the scope can reveal. It REPRESENTS how the mic diaphragm moved in and out, and how our ear drum is supposed to.

The music we hear- all its tones, rhythmic interplays, harmonies, imaging- our minds interpret from that complex "wave envelope". It is this unpredictable envelope's shape that counts. When a designer focuses on the theoretical "sine wave" components only, then the shape of the envelope has become immaterial to him.

Except to the ear. Which is why time coherence, and lack of cabinet problems, and linear drivers, and fewer crappy crossover parts, and proper crossover points are all important. Those all affect dispersion AS HEARD ON MUSIC.

Phasecorrect- you asked, "if time/phase accuracy is indeed retained...why do all time/phase coherent speakers sound different?" Because they are basing their claims of accuracy upon flawed measurements. The measurements don't pick up on all that we hear.

Ever wonder why we can't often play poor recordings? Everyone blames the studios, but it's the speaker's time-domain problems that are further distorting that distortion, contributing to unlistenabilty. Test: play a poor recording on phase-coherent headphones (Grado, Stax, others) then play it on a high-order crossover speaker just as loud.

Music is about time as much as tone and loudness. If you only test for two out of three, you won't be designing- only shoving parts into a box.

Best,
Roy

Eldartford-
Have a look at this link- let it load all its photos-

http://melhuish.org/audio/response.htm

You'll see how bad many drivers are at impulse response. You'll see ringing- just count the time period for each "cycle"- the ring is 1/period. You will see corresponding ripples in the impedance curve for each of those rings, and also see cabinet/floor reflections arrive, depending on the test and the enclosure.

Best regards,
Roy

Just because it can do a square wave does not guarantee it will sound excellent on music. A nice square-wave response does not reveal many other problems that affect what we hear. However, it is a GIANT step in the right direction.

Roy

Bigtree,
Vandersteen 2's and 3's have an unusual arrangement where the rear driver is both active AND passively driven. It is an active woofer below ~100HZ and is also a passive radiator reacting to the cubic air volume driven by the front woofer- which is ingenious. The 5's woofer is in an enclosure with an amplifier and a lot of EQ to make up for being mounted so far away from the upper woofer, and to compensate for its small enclosure.

Vandersteen has reduced the baffle size greatly, but that does not mean that these speakers are free of baffle reflections. There are still large amounts of reflected sound:
from the tweeter impinging on the mid (a little),
from the mid onto the front woofer and onto the entire cabinet,
from the front woofer reflecting off the entire cabinet.

Why? The felt applied to the face does not absorb much below 800Hz. And below that frequency is the range where the mid and front woofer also both want to be fully omni, thus reflecting off the entire cabinet and each other, and the felt does not prevent that. If you would like to see some of the math behind this, read my latter postings at this site in Europe, called The Vinyl Engine:

http://www.nakedresource.com//yabb/cgi-bin/yabb/YaBB.pl?board=general;action=display;num=1038342561;start=

I respect what Mr. Vandersteen has accomplished- his engineering is far better in many obvious ways, and in many subtle ways, than virtually all other speakers, which is why he has so many satisfied customers who can play just about any kind of music with them. But he has not banished reflections, just reduced them. Those still cause abberations in the total sound output. This is one reason the crossover circuit is complex- to make the final measurements read OK in spite of the reflections.

Best,
Roy

Hi Karls-
Much of what you say I agree with. And to answer you, I need to qoute your statements below. However, there is a lot of peer-reviewed published research which disputes a couple of points you have taken as fact:

You say, "(1)The commonly bandied-about equation describing speed of sound changes based on stuffing density is patently wrong on its face, and almost no one seems to notice."
Yes, the EQUATION is probably wrong, but there are many research articles which have directly measured the dramatic DECREASE in the speed of sound with decreasing frequency. I have never seen an experiment that shows the opposite, I'm sorry. Some are in the AES Anthology reprints from Old Colony Sound Labs. I have studied them for many years and see no mistakes in the many methods used to make that measurement. I'd be interested in seeing any research that demonstrates the speed of sound does NOT change dramatically. But I can believe that a particular equation would be wrong.

"(2)There are all these theories about how the stuffing works, from the air causing movement in the stuffing to adiabatic/isothermal changes to who knows what else. From what I have seen, these are 90+% BS. Viscous damping due to air movement past the fibers is almost all you need to understand stuffing."
These theories differ, along with their perceived BS content, because we don't really know how the fibers behave under all types of signals- transient or continuous, loud or soft...

For example, fibers can couple as a unified mass at certain frequencies, depending on the type of fiber, its packing density, the orientation of its fibers, the length of each line segment, the loudness of the sound and its duration. If the fibers do lock together under certain signals, then quite simply there must be less frictional loss as the fibers cannot rub against each other (because they are locked together). And that means little attenuation. Furthermore, they would then behave as a mass/spring system on that signal- which means resonance. So, viscous damping is not all we need to know, as sound in fibers does not always encounter viscous damping.

As far as adiabatic/isothermal arguments- The pressure throughout the t-line (or any box) is subtantially constant. When it does change, it does not last long enough to initiate a temperature change. Those are two important numerical values to know, so one can use them to come up with a theory (and a decent equation) for why the speed of sound does indeed change- a theory and equation that fit the experimental data and fit all preceding theories and equations which other experimental data validated.

So, why is the pressure substantially constant at all points in the t-line or any box? As some air molecules collide with the fibers, that scrubs off some of their velocity with each impact- some kinetic energy is lost to frictional heat as the fibers are made to rub together.
Fiberglass is rough, microscopically.
Wool is lubricated by its lanolin, and also smoother. Experiments show that fiberglass absorbs some low bass and a fair amount of midbass, and that wool absorbs very little low bass and just a little midbass. So the connection to roughness/smoothness seems obvious. But the real kicker comes after calculating the actual pressure differentials (which are responsible for velocities) at any point in the line:

The speed of sound, even if lowered inside the line, is still really fast- if the woofer starts to compress that air at 50Hz (taking 1/200 second, 5 millseconds, to reach its max stroke), upon reaching that max 1/4" stroke, the initial sound pressure is already 4+ feet distant! That means the pressure EVERYWHERE in the line changes nearly instantaneously- there is no/little air flow! It also means the woofer cone is moving far slower than the speed of sound down at 50Hz, ~2.7mph down there (1/4" in 1/200th second).

And since (not IF) the pressure everywhere rises and falls at once, there is little pressure differential to be found across any one region, so any local velocities cannot be high- there's little molecular velocity to scrub off. Which is exactly why materials of all sorts fail to absorb anywhere near 100% in the lowest bass- the individual air molecules just aren't moving much, so there's little velocity to scrub off.

In fact, with the pressure high, the molecules are mostly colliding with each other, as they're orders of magnitude closer together than the distance between the fibers- experiencing far more lossless collisions with their neighbors than lossy ones with the fibers. With the pressure lower, they do travel at an average higher velocity before hitting their neighbors, which are still `way closer than the distance over to a fiber.

You say "[Viscous damping] is why the impedance curves come out so flat."
Yes, by supressing the upper resonance peak that the undamped line would normally generate, as explained in my previous post.

"In an undamped line, you have a whole series of sharp impedance peaks at n/4 for all odd n. (Note that these are pipe resonances just like in an organ, and that this is very different from a ported box, which has only two peaks which are compliance/mass resonances.) The stuffing removes these peaks entirely at even midbass frequencies,"
Well, not entirely removed, but otherwise I agree with you on most every point. Except that a ported box has three resonances- at the two peaks and at the minima between them. And so does a transmission line, but the upper one is usually pretty-well damped, and we still have both the impedance minima resonance and an ultra-low frequency resonance (record-warp range) which most tests don't bother to measure, so the impedance curve looks flat. But here, flat does not mean non-resonant.

"The stuffing... damps them extremely effectively at lower frequencies (including at the lowest 1/4 lambda resonance)."
I have never seen any experimental evidence of that at the lowest 1/4 wave resonance, unless the line is stuffed like a bell pepper.

"On the other hand, ported design is specifically intended to function without damping, for all practical purposes."
Yes. The trick is to have the damping "turn on" at the higher bass frequencies, so it doesn't sound like a box... and Phasecorrect, that is my answer to your first paragraph in your last post- that I found a way to keep the box really quiet above resonance, whether sealed or ported, after years of building t-lines and all other designs. The only energy storage is at resonance, an amount which comes with a critically-damped system. It is not a lot, nor is it "stored" for longer than a half-cycle of the LF resonance (= critically damped).

"In addition, although this isn't discussed much, T-line woofers have their fundamental resonance frequency dramatically reduced when placed in the line... the entire mass of the air in the line becomes coupled almost 1:1 to the cone... a very substantial increase in effective mass."
Yes on all points Karls, and that will also reduce efficiency. This coupled mass can disconnect though, but only at VERY soft SPLs and VERY high ones, for viscosity reasons beyond the scope of this thread, but in any fluid dynamics text.

"An argument could be made, however, that due to compressibility, the initial attack at higher frequencies is much faster than if an equivalent real mass were added."
Yes that argument could be made, as this would be the de-coupling I mentioned above, which does not happen at our "ordinary" SPL's. But if that "less mass" effect was indeed true at the woofer's higher frequencies, then all that would do is allow the same input voltage to come out LOUDER in the higher tone range below the crossover point- because there's less mass to push for the same voltage. But the rise time would not be any different- that was already limited by the crossover. Which means the woofer is not "faster".

"Speed" is determined by the woofer quality, and also how sloppy the designer was in allowing any fibers to come too close to its cone, as there is an invisible layer of air that remains attached to the cone several inches deep. If the fibers reach into that zone, they drag down the cone's velocity.

"Contrast this to a sealed box, where the only way to drive the resonance down is to add real mass,"
Not the only way. One could increase the woofer's compliance only, or increase compliance AND mass (plus change the box' size).
"[added mass] hurts the transient response at higher frequencies."
Intuitive, but wrong. Added mass hurts only efficiency, not the transient response (which is the same as high-frequency response). Transient response is determined/limited by the crossover point. This was in shown the AES Journals- clearly, nearly 70 years ago. Added mass would hurt the transient response (HF respopnse) if we were allowing that woofer to go up to say, 1kHz or higher. If the added mass is a high %, that would tilt the woofer's tone balance downwards from 100Hz on up, but that can be balanced back to "flat" by designing the woofer to have higher compliance (also clearly explained in the AES Journals). So considering the woofer is probably crossed over below 300Hz, then its rise time is not affected by additional mass, just its loudness and maybe the tone balance of the woofer on the way to the mid.

"And in addition, an argument could be made that decay at all frequencies occurs much faster as well, due to the high level of damping the stuffing provides."
That is a very good way of describing any quieter box. If quieter, then there's less energy returning to the cone, so the decay time on an impulse would indeed be less. But we cannot measure that directly- only indirectly, and by listening via before/after comparisons.

"This increase in effective mass at low frequencies is very nearly "something for nothing", and is probably why T-lines seem to have both "speed" and "weight".
It's not for nothing if you have to slave over a far more complicated and heavy cabinet- especially in production! However, t-lines sound like they have more "speed" because the designer probably picked a better, more linear-motored woofer that has a vented voice coil and spider, and also has kept the stuffing away from the cone's vicinity. "Speed" is also a result of a properly-built box, and t-lines are always strong cabinets. Sealed and ported boxes are usually weak, as their designers don't know woodworking well enough to make the strongest joints using less wood. And neither do their cabinetmakers, as they are not mechanical engineers.

The sensation of "weight" comes from the lower distortion of the better woofer and from the woofer stroking less at the impedance minima, and from extra low-bass output from the port. And from the port's LF time delay, as those delayed LF's also linger on longer, longer than the rest of the music- making them audible on their own. Which is not entirely amusical!

You wrote, "I cannot disagree about the [time] delay of the back wave [out the t-line port opening], but I question whether it is an audible effect at the very lowest frequencies (because, again, a properly stuffed line will absorb everything from the lower midbass on up).."
Karls, I don't understand your comparision here between 'lower midbass' and the 'very lowest frequencies', but...

"The question becomes whether an 8-ft delay is audible at 35 Hz. I can't say because I don't honestly know. It could well be."
It is, especially when you can compare it to a sealed system which goes that low without the t-line resonance. Remember- it is not just an 8' delay from the t-line output- the woofer has its own delay in getting moving, often equivalent to ~another 8' delay. So the t-line output is actually ~16' behind the midband.

"I am not trying to disparage the quality of a low-Q sealed box in any way, as I too think it is often the best real-world solution, but I think that there is a lot more going on in a "T-line" than is commonly appreciated, and worse, a lot of plain misinformation floating around."
You're right on all points. To learn more, refer to those AES papers, and others in overseas journals whose titles escape me, but for which I have copies on file if you want them- experiments done by experienced scientists who had nothing to gain from seeing the results come out one way or another- just performing basic research, then trying to come up with theories that fit that experimental data- which any theory must, or it is just conjecture.

Thanks for your thoughts, Karls- you make some very good points. It sounds like you have done a lot of reading and made many speakers.

Bigtee- thanks for your thoughts. You are right about what have just said about Vandersteen.

And Phasecorrect, you are generally right about everything in your last paragraph! With regards to the question in your first paragraph- I cross over high enough to the mid that the woofer has stopped changing phase due to its own mechanical/acoustic rolloff down at ~40Hz, and has not started changing phase due to its HF mechanical rolloff. I also figured out how to put an aperiodic damping on the back of our mids to keep their 70-100Hz impedance peak from "turning off" my simple first-order electrical crossover up at the 300-400Hz crossover points, and to keep its own LF phase shift from adding to the desired x-over phase shift. And all of that holds for what I did for our mids/tweeters at their 2.8-3khz crossover points. And then I minimized our cabinets (but not too much), separated the drivers so they reflect much less off each other, developed the cast marble recipe we use, and figured out how to make strong, yet slender woofer cabinets, as their walls are not 2-3" thick. Go hear them.

Too many speakers jam a crossover point onto the woofer just an octave or two above its LF resonance, so the unavoidable woofer mechanical/acoustic phase shift adds to that crossover's phase shift. Thus, nothing comes out right, and we also hear room positioning become critical. And to make up for their losses due to the phase cancellation between woofer and mid, we see those woofers measure `way too loud, which makes John Atkinson scratch his head, because "it doesn't sound like it measures!" Right. Because time is being left out of that measurement.

Please read carefully my previous post and look into the link I gave in it to understand more, because I don't know how much more I can explain about time coherence. The link I gave is much more about the mechanical limitations of the transducers- no reason to duplicate that here!

Mr. Bischoff, this is all your fault.

Signing off in the big snow,
Roy
Green Mountain Audio

Gmood1-
I am glad to know you do not generalize based upon the design of the speakers. Ears should lead the way- So play an extremely wide variety of music and recordings (old/new/audiophile/distorted) until you hear what the speaker cannot do, as if you don't find those faults in the store, you will find them in the home at some point.

And when you find a flaw- such as "too peaky sounding on bluegrass", that means not only can you not play bluegrass, you'll find you cannot stand the sound of the massed, slightly dissonant strings that a 20th-century composer such as Samuel Barber or Morton Gould used to great effect, or soprano voices, or a Vienna Boys Choir disc, or realize the effect which comes over you hearing a Rachmaninoff piano concerto at full tilt, or appreciate more fully the genius of Hendrix, or the delicacy of touch required for ragtime piano, or Dixieland, or the inflections of Billie Holiday, or Janis Joplin, or Creedence Clearwater, or Chris Whitley, or King Crimson, or No Doubt, or Massive Attack, or Metallica, or Screamin' Jay Hawkins, or appreciate the real differences between...

So you play only the 'approved' audiophile recordings, of rather bland music.

You are wrong however, when you say there is no "best way to design a speaker" I could assume you are talking about basic decisions like woofer size, port or transmission line, six tweeters or one, but actually I really don't know what you mean with that statement.

There is a best WAY to design a speaker, which I'm sure you hadn't known, nor would I expect anyone to. It's the scientific methodology used to think through and then test and build and test and... And that method is for the designer to always start with the listener's location and the room around him and the SPL required and the bandwidth desired and the coverage angles. Those are exactly the parameters any professional concert-sound designer starts with. Then he's paid to work backwards to the drivers which will deliver that desired sound. Time coherence is only part of the equation, an important part.

And this approach to design is contrary to the way most all home speakers are designed- most of their designers got a wild hair and said something like, "the d'Appolito configuration is the way to go!" and never went beyond that, into understanding what happens because of that decision out at the listener's location. They began their designs at the cabinets instead of at your ears. This explains why so many high-end speakers are bought and then sold- the dissatisfaction.

So again, I'm not sure what you meant- maybe it was "don't trust any designer". Fine- in fact an excellent idea! But as you use your ears, don't do yourself a major disservice by listening to only audiophile recordings to find the best speakers. Happy listening!

Karls, thanks for the links- I've had a look, but will not respond here, as this is not the thread for that, and I probably have not the time to say anything useful. I do see, at first glance, what appear to be some wrong assumptions about what the impedance curve peaks mean vs. the 1/4-wave line lengths. But I'm likely wrong- their measurements do not go low enough below 20Hz to reveal the errors.

Phasecorrect- you make some good points about design execution, and bear in mind most speaker designers are nowhere near fully trained. Fortunately, no permanent harm comes from bad speakers, so those designers can "get away with it". You just wouldn't want them to engineer your car or medicines or food-handling machinery or house.

If I seem mean-spirited or overly critical- I am sorry, but what I've said about poor design methods is true- I have spoken to `way too many designers, while great guys, well-intentioned, smart and hard-working, simply never slogged through the graduate calculus and fluid dynamics, thermodynamics and the mechanical engineering it takes to make a speaker that performs well on most all music, in most rooms, with most amplifiers. And reviewers support those halfway design decisions saying, "These speakers really need tubes." or "They really can't play a distorted recording." While those are accurate statements, they put the blame on something else in the chain, and not the speaker. What a disservice to you, the listener! But then reviewers are usually not technically trained, so it's only natural. I would hope that anyone reading these submissions of mine here and on that European link I gave will see how basic physics applies to speakers and how that explains what we hear and also the discrepencies between measurement and hearing.

Best regards,
Roy

Karls- you got it right.

If you wanted to know the actual % modulation distortion, you'd have to know the stroke and frequency of the mid's vibrations that are affecting the tweeter's sound. Which are random, as far as the tweeter is concerned. Which means the modulations are unpredictable on music- so all we can say is that they should probably make the sound hazier or dirtier.

Jeff's crossover is probably the only one that could make a co-ax design work well. Even then, the tweeter dome would require a modest horn around it- a waveguide to keep the tweeter's sound from bouncing off the mid's cone. Of course, the mid's sound will bounce off that waveguide's exterior...

The "horn-loading" coloration is a common term- what we're hearing are the quasi-transverse reflections from the sidewalls of the horns, and the reflection from the mouth of the horn back to the throat.

At the mouth of the horn, the sound pressure goes from travelling in a high acoustic impedance to a low acoustic impedance. Thus, from the un-equal impedances, a reflection/standing wave takes place inside the horn. Then there is the matter of a horn's possible throat-compression ratio that boosts efficiency and distortion (love them PA horns, don't you!). You can't compress/rarify the air more than 1% or you get harmonic distortion from the air itself.

Roy

This might help: a link to my postings on "The Vinyl Engine".

http://www.nakedresource.com//yabb/cgi-bin/yabb/YaBB.pl?board=general;action=display;num=1038342561

Best,
Roy

Good question! Active x-overs/built-in amps are great!

Of course, the cost of those parts and the extra labor multiply into a higher retail. Existing powered-speaker companies offset some of that by using far less than the best woofer, tweeter, cabinet, wire, terminals, amplifier design and electronic parts, and via quick-assembly techniques- like push-on connectors to the drivers.

The powered monitors sound better than their competition, but this only reflects on how bad their competition is, and not on what can really be achieved with proper drivers, cabinets and passive crossovers for the same total cost, or less, including the outboard amplifier- which can be changed that day if it goes bad, as opposed to sending the powered speakers in for repair. Talk to a repair tech about the bright idea of building a VCR into a TV set. Makes them curse...

Home users would have to scrap their existing amplifier to use powered speakers- not likely.

Whose amplifier design do we put in, assuming they would even sell their "best design" to a speaker company? Rowland, Edge, Audio Note, 47Labs, Bryston, Wavelength, Manley... none are perfect. And they always get better (almost always) every year or two.

Those are the reasons which have kept us from putting in active electronics here.

Except for the compromised powered monitors out there, or cost-no-object reference speakers, we would seem to be stuck with passive crossovers.

Best,
Roy

Cdc- Thiels use a complex crossover circuit to force first-order acoustic rolloff responses from the drivers, because of their driver choices and cabinetry. Dunlavy and to some extent, Vandersteen employ similar approaches.

Your question about a single driver- if there was one with a perfect cone, free of breakup, the end result would be little bass response and very narrow dispersion in the highs. To widen dispersion in the highs and extend high-frequency response, a whizzer cone is attached, driven through a mechanical crossover (the adhesives used) between it, the voice coil, and the main cone. Those type of crossovers at best are 2nd-order, which means there's 1/2 wave-period of time delay at the crossover point. There are cone breakups usually still present. There are standing waves in the whizzer, reflected from its un-terminated outer rim. Those all are responsible for less than pinpoint imaging. At least their designers get to leave off any electrical crossover, which is a good thing, considering how much information is lost in most of circuits. Whizzer-cone drivers are not time coherent, but just less imperfect that most speakers that do use a crossover.

Phasecorrect- just remember that "phase correct" and "phase coherent" do not mean "time coherent". The converse does. See my first 02/12/03 post above.

Best,
Roy

The simplest electrical crossover on a speaker is an inductor placed in series to the woofer, and a capacitor to the tweeter. The amplifier drives into both simultaneously. If they are perfectly equal and opposite in "reactance", then they cancel out, as far as the amplifier is concerned. This cancellation is what makes this the only dividing network without time-delay distortion.

This is a first-order network. Its two components can be used only when the drivers and cabinet designs are "perfect".

Not bloody likely.

More complex circuits are usually required, whether using two or ten more parts. The result can still be a "measured" first-order acoustic rolloff. The extra parts "modify" driver non-linearities and "make up" for cabinet reflections. Which they cannot- but they can fool the microphone. Of course, extra circuit-parts cannot be perfect either. Reductions in transparency and dynamics are givens.

To keep the number of crossover parts to the barest minimum, one has to use the most linear drivers- which are relatively few. However, not that few: specific examples include tweeters from Morel, Dynaudio, Foster, Stage Accompany, Pioneer TAD, and Scanspeak. Certain mids from Audax, Eton, Davis Acoustics, Bandor, Jordan, Foster, Peerless and Aurasound. Specific woofers from Scanspeak, Davis, PHL Audio, Volt, Audax, Peerless, Pioneer and Aurasound. There are others.

And every one of them is far more expensive than the drivers used in most designs.

For a commercial designer who wants to use the simplest crossover, it's hard to find the best drivers under deadline conditions. But by using the most linear drivers, within proper cabinetwork and correct bandwidths, the crossover circuit can be reduced to just a handful of parts, for clarity and for time coherence. The converse is entirely true.

Best,
Roy

Yeah, the original Ohm had a lot going for it. Magazines replaced square-wave tests with the computerized MLS test, which can interpolate the phase response and any ringing from the MLS psuedo-noise (but not in great detail- as most of what you see are the averages of 20Hz-wide frequency bins).

Some of the first-order speakers currently marketed do well on square waves, but manufacturers see no reason to publish the test, for marketing reasons, so the competition cannot find out easily, and because this test is not the only one to be passed for good sound, as I'm sure you suspect.

What you heard, good or bad, in the Ohm lays far deeper than what the square wave can reveal- for two reasons:
--the square wave is composed of only odd-order harmonics plus the fundamental (any even-order content seen on the `scope is distortion). Thus it only tests certain tones, not all tones.
--the square wave's dynamic range is far different than music- it does not stress the drivers enough, nor last long enough to excite the woofer.

A square wave is a guide- if you can find out where the little departures from a flat-topped characteristic come from, and then fix them, great! However, there are better tests for the problems behind those squiggles, ones which a smart manufacturer is not going to reveal, nor a poor one reveal that they don't perform!

You raise valid points- not a very professional industry is it? Becoming a better listener and gaining some technological understanding seems to be the only way to find something decent!

Best,
Roy

A ported or transmission line speaker can never be time coherent, nor dipole, bipole or omni, OR SEALED. See my 1/16/03 post, about why any moving system has a natural time delay down at its resonance. The ported or "transmission line" (which is a port variation and NOT a transmission line) designs have the SAME phase shift as a sealed box for the sound leaving the front of the cone, and the port opening's output has additional time delay AND polarity inversion.

For living room use and mix monitoring, I believe point-source design techniques are the best way to achieve fidelity- primarily to avoid hearing time-delayed output from more distant drivers, or from more distant panel regions.

Amplitude linearity is most important when the speakers are of minimum-phase design, as then one can hear small deviations from amplitude linearity. But if the speaker has lots of phase shift, that skews the harmonic structure of the music, which puts those harmonics out of phase and thus alters the perceived timbre of the instrument or voice. Which makes it harder to "accept" what your test microphone is saying is "flat" amplitude response. The warped phase response keeps amplitude deviations from being noticed as much. In those designs, phase and amplitude are not independant parameters. When you remove phase nonlinearities, then amplitude response IS an independant parameter.

Amplifiers have problems- if a speaker has phase shift, it will change the sound of those problems, usually making them worse. It is distorting distortion. The amplifier has its lowest distortion working into a flat impedance curve from the speaker, but many speaker designers try for that by adding extra parts in the crossover- which act to flatten the impedance curve, but reduce clarity. And some of those impedance-flattening techniques do more harm than good- especially the ones for the woofer resonance- what a mistake!

Good questions, Mr. Unsound.

Best,
Roy

And your point would be... not to discuss the many approachs to design, and the pros and cons THAT EACH MAN MADE DEVICE FACES?

Or would it be that every speaker is good? Then why are you reading...

Oh well,
Roy