$ony Quality

[randycat99]

Maybe you're not understanding what they're saying. They're talking specifically about the tweeter and the reason for its flat response all the way up to 50kHz. They're not talking about distortion in midrange and woofer drivers.

I understand exactly what they are saying, and it is just flashy marketing hype other than having achieved response out to 50 kHz (which is admittedly an impressive feat). I'm saying distortion exists (specifically nonlinear) at varying levels on any tweeter whether it extends to only 16 kHz, 22 kHz, or even a whopping 50 kHz. The only distortion that is "shifted" by said response extension to 50 kHz is breakup distortion, which is arguable comparably addressed in one that only goes to 22 kHz or even one that is in calibrated controlled breakup at a mere 6 kHz as in a softdome tweeter.

Less distortion is better regardless of what kind of distortion it is.

True, but the kind these guys are talking about wasn't necessarily a problem to begin with unless you have a gross breakup distortion within the audible range.

For example?

See any "better" metal dome tweeter. If you don't believe me, fine, but this information is bordering on "commonly accepted". Just look around, I'm not lying.

If the driver is being distorted then it's transient performance will be compromised.

It is certainly a factor which can impact inherent transient performance. However, if true transient accuracy is not there, no amount of distortion reduction is going to make it suddenly have "exact transient performance". Having 50 kHz response certainly helps, but the primary problem of fundamental resonance of the tweeter moving system overshadows that entirely. If you ever get to compare the pulse response of this particular tweeter with a 50 kHz low-passed pulse (let alone a 20 kHz low-passed pulse), you will find that there is still some ways to go before they are truly the same or even match up in the time domain. You could increase the extension even higher, but that only gains you "inches" in benefit where the aforementioned mechanical resonance issue would buy you the "final mile".

That's what a crossover does. Why else would speakers need them?

The crossover not only mitigates the excitation of fundamental and upper resonances, but it also defines the operating range for a given driver such that high power operation is possible w/o over-excursion at too low a frequency.

For example?

For example, most if not all drivers (no matter what range they are designed for) will have true transient response within a few Hz of DC. That is, the waveform in will be identical to waveform out with no time lag and no phase delay. As you go up in frequency, phase delay will increase and vary with much association to the fundamental mechanical resonance. Once you have phase delay, that is automatically a deviation from true transient response. In a typical tweeter, the first resonance may be at 2 kHz at which point phase angle may be 90 deg or so (certainly not 0). The tweeter hasn't even reached a point where output has come on line, and the phase response already indicates a departure from perfect transient response. Getting to about 3 kHz the useable range of the tweeter is finally apparent, and by this time phase response has continued rising above 90 deg to top out at 180 at some upper frequency. By that point, your upper frequencies which help you define the edge of a transient are lagging by at least 180 deg (effectively the actuator is undergoing an in-stroke when it should be doing an out-stroke). Now factor in inductance of the actuator coil and acoustic transformations due to dome and baffle geometries, and you get all sorts of phase aberrations. A perfect transient response at this point would literally be a miracle. So all you really have is the fancy 50 kHz extension to enable great transient potential, but since the phase is fubar by that point and virtually everywhere below that, the point is moot.

By extending the frequency capability of the tweeter up to 50kHz, you get less audible distortion from the tweeter itself. Same principle why a midrange will distort if fed a high frequency signal.

That is only true if break-up distortion was a problem to begin with. Meanwhile, there are other equally audible forms of distortion that occur over the entire range of a given driver that have nothing to do with how extended your upper range is.

Well if people can hear a difference in dynamic range then a comparison is completely relevant.

People "think" they hear a lot of things that may or may not really be there, and therein lies the source of much debate to this day even though the technology is fairly well established and verifiable. You may think you are hearing something and attributing it to dynamic range, but it could be something else entirely that is responsible, up to and including a simple placebo or suggestion effect. In the case of LP vs. CD/SACD, one only needs to consider the technical limits of each respective media to realize that it isn't dynamic range where LP can prevail as superior. It simply is technically impossible (just as a 300 Mpoly/sec Xbox game is, even though it looks like it might be). If the dynamic range is there, it is absolutely attributable to the particular mastering of the program material.

Mind you, all of this is not to say that what you are hearing from your system is not good. It is every bit as good as when you first heard it and prior to reading this post. You would be surprised how badly mangled a sound wave can suffer and still sound great. Thus the process of associating what you are actually hearing to specific technical metrics can be quite deceiving. Many people "think" they have a concept of what a difference of 93 dB feels like (as humble a number that is by SOTA digital standards), but you would be astounded by how far off you were if you are ever exposed to it in a calibrated, controlled environment. Similarly, many people "think" they could hear out to 20 kHz easy, but the reality is most people would be lucky to get much past 14 kHz (a worse case may be down to 6 kHz). True, you might hear 20 kHz sinewave if it is loud enough, but the reality is you may be pumping 300-400 W into a tweeter to really start hearing it (and that would be for about 10 sec, after which the tweeter would be going up in a plume of smoke).

 

[PC-Engine]

The thing is we don't know how good the transient response of the tweeter is unless we test it with special equipment. We can't just say that this specific driver's transient response becomes moot as at X frequency the phase delay is greater than 90 degrees because we don't know it's mechanical resonance. Obviously every driver has it's limits, but since we don't have the data for this particular driver, we can't just assume. We don't even know what material it's made from.

 

[randycat99]

It's basic operating principle for an electrodynamic driver. All drivers of that nature inherently will have those properties (just as you can only have so many transistors on a given die size to constitute the desired computation units). To circumvent that on a fundamental level, you are talking about either a massless or infinite spring driver system.

The instruments required to measure it are really not that "special". I'm sure they have them (essentially a PC, some FFT analysis software, a high-quality reference mic, and an amplifier), tested their tweeters, and have the exact sort of measurements I alluded to earlier. Whether they post them or offer them up to public observation is entirely a different question. It would simply show that the transient response isn't any more ideal than any other better dome tweeter out there (but still "good" with regard to any properly operating tweeter). That won't stop them from marketing claims of perfect transient response, of course. They simply have much more ultrasonic extension than is typical of their competition, and 50 kHz being such a big number would seem to be enough to "substantiate" the claim with most readers. It's much like reading the marketing blurb for a new videocard vs. really understanding what the technical specs are indicating and intimately knowing what is possible on the hardware on a programming level. Some stuff will wash, while other stuff is most obviously fluff or nonsensical on a technical level.

 

[maskrider]

Cables make a difference, no matter high-end or not, adequate quality cables are indeed very necessary as they will not deterioate so much comparing to casual ones.

Well I suppose it depends on what you consider 'casual' but one can easily get cables that won't deteriorate without going to the esoteric brands such as Kimber, Transparent, Tara Labs, or the much more esoteric such as Elrods ($1500 for Mk III Signature Power Cords!!).

I consider casual ones to be like Cardas neutral reference or better. I have sold most of my cables collection but I still have a pair of Burmester silver and a pair of Transparent Reference XLR in their box, had tried some exotic Siltech, vdH and some others before.

Higher-end I will mean US$2000 or above in price with normal length.

But at the end of the day, it is the electronics at the signal outputs that determines how much cables will influence your system. I had been very active with high-end cables some years before, tried many brands and models. Still have a couple of them in their boxes, waiting for me to use them when I save eough money to buy a good quality pre-amp.

I'm enough of an Audiophile that I know about 'high-end' cables. Yet as I mentioned previously, I'm still quite skeptical about the claims made by the aforementioned Esoteric cable brands. I'm no RF engineer but I doubt problems such as 'skin effect' truly matter for the typical lengths of cable runs in a Home Audio setup. Personally I'd have to audition them in a controlled DBT before I would believe them.

I didn't doubt anything on your side nor questioned anything, just wrote what I had experienced.

And most of the the claimed technical advantages are marketing oriented, the truth (or the myth) is hidden behind the smoke screen, and that is closer related to the inherent inadequate design of the electronics that make them perform as they are in those systems.

 

[Ty]

I consider casual ones to be like Cardas neutral reference or better. I have sold most of my cables collection but I still have a pair of Burmester silver and a pair of Transparent Reference XLR in their box, had tried some exotic Siltech, vdH and some others before.

Higher-end I will mean US$2000 or above in price with normal length.

Have you actually tried to listen to any of these in a Double Blind fashion to see if you can 'hear' the difference? I haven't had that opportunity myself.

I didn't doubt anything on your side nor questioned anything, just wrote what I had experienced.

And most of the the claimed technical advantages are marketing oriented, the truth (or the myth) is hidden behind the smoke screen, and that is closer related to the inherent inadequate design of the electronics that make them perform as they are in those systems.

Apologies. That was poor phrasing on my part. I should have said, "I know OF the existence of High-end cables', not that I know how good or bad they are. I didn't mean to come across as a know-it-all or as touchy.

But I agree that with most products, one should do the research to cut through the marketing smoke & mirrors.

 

[maskrider]

Have you actually tried to listen to any of these in a Double Blind fashion to see if you can 'hear' the difference? I haven't had that opportunity myself.

Yes, tried once quite some years back. Some of the top or close to the top models have distinguishable brandname sonic signature when they are in tested in a system with enough resolving power.

Apologies. That was poor phrasing on my part. I should have said, "I know OF the existence of High-end cables', not that I know how good or bad they are. I didn't mean to come across as a know-it-all or as touchy.

But I agree that with most products, one should do the research to cut through the marketing smoke & mirrors.

I am no good at using the emoticons, neither, ha ha !

 

[PC-Engine]

It's basic operating principle for an electrodynamic driver. All drivers of that nature inherently will have those properties (just as you can only have so many transistors on a given die size to constitute the desired computation units). To circumvent that on a fundamental level, you are talking about either a massless or infinite spring driver system.

The instruments required to measure it are really not that "special". I'm sure they have them (essentially a PC, some FFT analysis software, a high-quality reference mic, and an amplifier), tested their tweeters, and have the exact sort of measurements I alluded to earlier. Whether they post them or offer them up to public observation is entirely a different question. It would simply show that the transient response isn't any more ideal than any other better dome tweeter out there (but still "good" with regard to any properly operating tweeter). That won't stop them from marketing claims of perfect transient response, of course. They simply have much more ultrasonic extension than is typical of their competition, and 50 kHz being such a big number would seem to be enough to "substantiate" the claim with most readers. It's much like reading the marketing blurb for a new videocard vs. really understanding what the technical specs are indicating and intimately knowing what is possible on the hardware on a programming level. Some stuff will wash, while other stuff is most obviously fluff or nonsensical on a technical level.

Actually you don't need a massless system if your magnet is strong enough to move the system assuming the material has a good Young Modulus. And once you get into electrostatics it's moot. You can even have servo control of the suspension mechanism. The fundatmental barrier is not 25kHz. That's where materials research comes in. There is no law that says that barrier exists. Also the damping factor also controls phase distortion so all of this talk about a theoretical wall at 20-25kHz is BS.

 

[randycat99]

Actually you don't need a massless system if your magnet is strong enough to move the system assuming the material has a good Young Modulus. And once you get into electrostatics it's moot. You can even have servo control of the suspension mechanism. The fundatmental barrier is not 25kHz. That's where materials research comes in. There is no law that says that barrier exists.

Ahem, mind you, you are arguing this with someone who has studied speaker operation since college and tried to make a career out of it.

A "strong enough magnet" has no bearing on top end or absolute transient response. Primarily, it effects system gain (sensitivity) and damping. Once you get the optimal damping (to balance the mass and spring of the system), further increases in motor strength will not improve transient response. It will lead to other operational issues, however.

Elastic modulus and mass are inherently tied together to determine the resonant modes of the system. Unfortunately, materials with high E typically have high m, as well. (It's called metal. ) You could also consider ceramics, but they tend to be even heavier and are not readily manufacturable to such micro-thin wall thicknesses that metal can be. So there is only so far materials research can take you, outside of inventing a new element altogether. There is some promise in composite materials, but don't expect miracles.

Electrostatics are still subject to mass-associated dynamics, despite what most audiophile electrostatic proponents would have you believe. They won't have a typical break-up resonance like conventional domes drivers, but they still have mass, suspension compliance, and thus finite transient responses.

Servo control of the suspension system is still subject to mass and suspension compliance, and thus finite transient response. You will find that servo-driven systems are even more plagued by "speed" issues than electrodynamic systems. That is the whole point of why electrodynamic systems are so prolific.

 

[randycat99]

Also the damping factor also controls phase distortion so all of this talk about a theoretical wall at 20-25kHz is BS.

Yes, damping ("damping factor" is not the correct term you should use for this discussion) affects phase response. If you overuse damping, you will annihilate high frequency response, thus making transient performance even worse.

To clarify, there is no assertion that damping has anything to do with a "theoretical wall at 20-25kHz". Material properties (those known to man and are physically possible in our physical universe) give the existence of this 25 kHz-ish limitation. It's not really a "wall", either. Damping of the material will determine how far beyond that frequency reasonable operation can be utilized (or not far at all). Either way, any hopes of near-perfect transient reponse past that point fly out the window once that material property is exceeded. TO REPEAT- it doesn't preclude operation at higher frequencies, it just dictates that such operation will have crossed into the non-ideal regime.

Aside from all of that, the real limiting factor to perfect phase response (and thus transient performance) occurs down in the 2 kHz area (for virtually any dome tweeter type of system), anyway, as indicated many posts ago. It is utterly inherent to the operation of a vibrating system with mass and compliance.

 

[PC-Engine]

Well you seem to know alot about speaker operation, but I still believe advances in material research can negate this theoretical wall. There are many composites using different forms of manufacturing that has better properties than metal. The primary reason why metal is used is because it's easier to form not necessarily the best

Also there mechanical damping and electrical damping. A high frequency driver relies more heavily on mechanical damping and like I said to have good transient response to overcome this high damping you use a stronger magnet.

 

[MfA]

I couldnt give a toss about high end audio for myself, but from a technical standpoint I wonder ... why arent there more speaker systems which solve this little conondrum "simply" by changing the input so the output becomes correct?

 

[randycat99]

Well you seem to know alot about speaker operation, but I still believe advances in material research can negate this theoretical wall. There are many composites using different forms of manufacturing that has better properties than metal. The primary reason why metal is used is because it's easier to form not necessarily the best

Uh, no. Metal is used because it is a better performing material. The advances in material research you speak of can lead to many "higher performing" materials, but that doesn't necessarily mean "better at being in a tweeter". It could mean greater ductility, less brittleness, higher strength before yielding, lower bulk weight for an equivalent strength, greater manufacturability, etc. None of this deals with sonic velocity, which directly affects the appearance of break-up modes. There's no way you can currently cheat sonic velocity. It comes down to simply density and elastic modulus. As either quality are strongly inherent to material elements, themselves, there isn't much hope for new developments there aside from inventing a new element. One possible exception is "foamed materials" (out of materials that were previously not thought to be foamable), but like I said before, don't expect miracles.

Also there mechanical damping and electrical damping. A high frequency driver relies more heavily on mechanical damping and like I said to have good transient response to overcome this high damping you use a stronger magnet.

That is unequivocally wrong! A stronger magnet leads to a stronger motor which leads to more damping, not less. There is no "overcome this high damping". Electrical and mechanical damping are cumulative. Use too much, and you compromise transient performance.

 

[PC-Engine]

That is unequivocally wrong! A stronger motor leads to a stronger motor which leads to more damping, not less. There is no "overcome this high damping". Electrical and mechanical damping are cumulative.

Yes it's cumulative, but what I was saying is that to get better transient response from a highly mechanically damped driver like a tweeter, a stronger magnet can over come this mechanical damping by moving the mass quicker. The electrical damping comes from the amp and is usually helps the lower frequency driver. The damping factor of the amp is for the low frequency driver. The high frequency driver depends more on mechanical while the low depends on eletrical.

Also regarding materials, Pioneer back in the 90's had a pure carbon dome tweeter that went up to 35kHz

 

[PC-Engine]

I couldnt give a toss about high end audio for myself, but from a technical standpoint I wonder ... why arent there more speaker systems which solve this little conondrum "simply" by changing the input so the output becomes correct?

If I understood what you just said I would offer my insight on your question.

 

[randycat99]

Yes it's cumulative, but what I was saying is that to get better transient response from a highly mechanically damped driver like a tweeter, a stronger magnet can over come this mechanical damping by moving the mass quicker.

It will move the mass "quicker" as far as gain and sensitivity are concerned. It won't move it faster as far as higher excursions at higher frequencies due to the negative impacts of overdamping. Contrary to your belief, tweeters rely on comparable amounts of both mechanical and electrical damping. A stronger motor does not "overcome" mechanical damping. It adds to it. They are cumulative, period. An overdamped system is not going to give you any sort transient performance to brag about, AT ALL. It doesn't matter how much more force you apply, the farther you swing into the overdamping regime, the slower your response will be.

The electrical damping comes from the amp and is usually helps the lower frequency driver.

A low frequency driver has some of both, but electrical damping is the dominant component. Nevertheless, the 2 damping components are cumulative. They do not counteract.

The damping factor of the amp is for the low frequency driver. The high frequency driver depends more on mechanical while the low depends on eletrical.

This is hogwash. Either system relies on the cumulative damping of the 2 damping components. The "damping factor" of an amp is something entirely different, and virtually always irrelevant for any speaker being driven by a modern solid-state amp (not a tube amp). You don't want to bring "damping factor" into this because not only is it a different term than you mean altogether, but it is not relevant to the improvement of transient performance on an electrodynamic speaker system, period.

Also regarding materials, Pioneer back in the 90's had a pure carbon dome tweeter that went up to 35kHz

Fair enough. You can tell how feasible it was by how prevalent it is in use today in current speakers, right?

 

[MfA]

Measure the response of the speaker, calculate a corresponding inverse filter and apply it to the audio before D/A conversion (dont forget to put some sensible bounds on amplification at given frequencies though, putting lots of power into trying to get a speaker to reproduce frequencies it just cant wont be healthy). To avoid cabling issues use active digital speakers too while you are at it.

From my point of view that seems the best solution if you want the sonic waveform to be as close to the one stored on the CD as possible. Trying to limit non-linear distortion by better speaker design seems a worthwhile goal, although in theory you could correct some of that with DSP too, but Ill agree with randycat that the extent to which you can limit phase distortion and other forms of linear distortion that way is rather limited. AFAICS it is also a hell of a lot more expensive to try to lessen linear distortion that way than using DSP.

 

[randycat99]

I couldnt give a toss about high end audio for myself, but from a technical standpoint I wonder ... why arent there more speaker systems which solve this little conondrum "simply" by changing the input so the output becomes correct?

There are a few examples (very exotic, esoteric models) that use a corrective feedback strategy. The kinds of correction are still limited by physics, however. You can correct for things such as amplitude at a certain frequency, overexcursion due to overdriving signals, even phase if the correction is done in a digital domain. There's more, of course. It can be pricey, and it certainly is complicated. Some things are just not worth "correcting" if you can just design the system to work at its best just naturally. Also there is a feasible limit to how far you can correct things before you are just spending inordinate amounts of power in correction, not making sound. It's still a design philosophy in its infancy with the potential to do some incredible things. Things like cost and complexity may keep it from the mainstream indefinitely.

OTOH, you would be surprised how difficult it is to convince the most purest audiophile to listen to anything digital and be impressed. It just doesn't have the same "credibility" as analog. Given the potential for digital manipulation to warp the original signal to virtually anything you wish, you can understand how hesitant they are to something that could twist and contort the original sound to such a degree even if the end result from the speaker could be a substantial improvement. Being a purist, means you inherently do NOT want something doing that to your "pristine" signal.

 

[maskrider]

There are a few examples (very exotic, esoteric models) that use a corrective feedback strategy. The kinds of correction are still limited by physics, however. You can correct for things such as amplitude at a certain frequency, overexcursion due to overdriving signals, even phase if the correction is done in a digital domain. There's more, of course. It can be pricey, and it certainly is complicated. Some things are just not worth "correcting" if you can just design the system to work at its best just naturally. Also there is a feasible limit to how far you can correct things before you are just spending inordinate amounts of power in correction, not making sound. It's still a design philosophy in its infancy with the potential to do some incredible things. Things like cost and complexity may keep it from the mainstream indefinitely.

TACT Audio has several models of RCS (Room Correction System) which will correct system and the room related frequency response, reflection (up to 1st order) and phase error after setting up in your system, for sure up to the limit of your system.

They are not really cheap but not expensive neither.

 

[MfA]

Feedback is another option, but it is solving the problem a little late ... if you know the response of the system "feedforward" correction seems easier (what I propose cant really be called feedforward, but you know what I mean I hope).

 

[maskrider]

Feedback is another option, but it is solving the problem a little late ... if you know the response of the system "feedforward" correction seems easier (what I propose cant really be called feedforward, but you know what I mean I hope).

There will always be problems when a system is put into a room, they interact and affect the performance.

Another manufacturer (don't know if it does that now) has a library of frequency response data of many speakers that will fit to their digital audio processor to adept the signal of the source to the speaker.