The sonic benefits of an active crossover. A discussion.
- Thread starter treitz3
- Start date
Wanna revive the wager Tim? Blind test, right up your alley. 
Wether you are splitting up a large or a small signal there are passive parts involved. Active stages all have signatures. If you are talking only about efficiency, I would agree with you. If you are talking about levels of fidelity, you and I both know that depends on implementation. Active XOs will not turn a poor driver into a stellar one just as a passive one won't. I will emphasize one point. If one decides to go active, one had better be damned sure he's chosen the right drivers and knows EXACTLY where it's flat and where it rings and that means the spider and enclosure too, especially for woofers.
Using a poorly designed and implemented active crossover will give you the same level of signal degradation as an equally poor source. Now if you say sure that's analog active XOs but not digital, think again. Digital active XOs face the same problems. The engineering has yet to achieve the theoretical potentials as our very own WBF engineers all point out. Again, it is an implementation issue.
Wether you are splitting up a large or a small signal there are passive parts involved. Active stages all have signatures. If you are talking only about efficiency, I would agree with you. If you are talking about levels of fidelity, you and I both know that depends on implementation. Active XOs will not turn a poor driver into a stellar one just as a passive one won't. I will emphasize one point. If one decides to go active, one had better be damned sure he's chosen the right drivers and knows EXACTLY where it's flat and where it rings and that means the spider and enclosure too, especially for woofers.
Using a poorly designed and implemented active crossover will give you the same level of signal degradation as an equally poor source. Now if you say sure that's analog active XOs but not digital, think again. Digital active XOs face the same problems. The engineering has yet to achieve the theoretical potentials as our very own WBF engineers all point out. Again, it is an implementation issue.
Just the first two technical cites from one of your earlier posts (there are others):
Headroom
Driver control (damping)
Wanna revive the wager Tim? Blind test, right up your alley.
Wether you are splitting up a large or a small signal there are passive parts involved. Active stages all have signatures. If you are talking only about efficiency, I would agree with you. If you are talking about levels of fidelity, you and I both know that depends on implementation. Active XOs will not turn a poor driver into a stellar one just as a passive one won't. I will emphasize one point. If one decides to go active, one had better be damned sure he's chosen the right drivers and knows EXACTLY where it's flat and where it rings and that means the spider and enclosure too, especially for woofers.
Using a poorly designed and implemented active crossover will give you the same level of signal degradation as an equally poor source. Now if you say sure that's analog active XOs but not digital, think again. Digital active XOs face the same problems. The engineering has yet to achieve the theoretical potentials as our very own WBF engineers all point out. Again, it is an implementation issue.
Nothing to wager, jack. Of course it's about implementation. Of course a poorly designed and executed active speaker can sound bad and a well-designed and executed passive can sound good...great, even. The topic is the sonic benefits if active crossovers. I think "benefits" assumes good design and implementation. Do you think there are none? Do you think if a speaker designer took his own very well designed and implemented speakers and created an active of equal quality with the same acoustic (cabinet, drivers...) components, that there would be no sonic benefit? Because I have heard several designers, a couple on these pages, comment otherwise. We'll probably never have the opportunity to do the comparison so it's pretty theoretical, but do you really think there would be no benefit? Would there be a sonic benefit to the passive version, then? What would that be?
Tim
I reviewed and didn't find anything particularly technical in my posts, but I'm flattered, even if they were technically flawed.
Tim
Those are big assumptions Tim. Assumptions not in line with the language you've been using to extoll the virtues which I am not refuting. If anything, my objection is the way the approach is being bandied about as a silver bullet. It is NOT. There's no such thing passive OR active.
Where we differ is my perspective is more on external active systems and yours solutions where all the user has by ways of control are a gain control or two (monitors). In the latter, you've handed the designer the reigns. Not the case with the former where there are way more variables to be studied yourself. What driver to use, what is it's optimal operating range, how is it mounted, what are it's mechanical roll off characteristics, what amp should I use, what slope, what slope(s) for the driver next to it depending on the same things above? All this to only fall apart because the active crossover is imprinting itself. Make no mistake, they all do it's just a matter of degrees. If you can't hear it or get used to it, that's fine. Just don't make it look like because a system is active (direct connection from amp to driver) suddenly straight wire with gain has become a reality. You may have removed the passive XO but in it's place you've added gain stages in the signal path. An active XO for stereo will have a minimum of 4 gain stages.
Where we differ is my perspective is more on external active systems and yours solutions where all the user has by ways of control are a gain control or two (monitors). In the latter, you've handed the designer the reigns. Not the case with the former where there are way more variables to be studied yourself. What driver to use, what is it's optimal operating range, how is it mounted, what are it's mechanical roll off characteristics, what amp should I use, what slope, what slope(s) for the driver next to it depending on the same things above? All this to only fall apart because the active crossover is imprinting itself. Make no mistake, they all do it's just a matter of degrees. If you can't hear it or get used to it, that's fine. Just don't make it look like because a system is active (direct connection from amp to driver) suddenly straight wire with gain has become a reality. You may have removed the passive XO but in it's place you've added gain stages in the signal path. An active XO for stereo will have a minimum of 4 gain stages.
First paragraph, dead wrong. Those weren't assumptions, they were (as of yet unanswered) questions. Second paragraph, spot on. I absolutely believe in finding a really good designer and handing him the reigns. One of the greatest lessons one can learn in life is one's own limitations. I'd guess that for every 100 audiophiles playing the synergy game - active or passive - there is maybe one who has the chops to do more than thrash about. And those numbers probably give a huge benefit of the doubt.
Tim
Tim,
You will always be comparing apples with oranges. When you have a passive speaker you add an amplifier that optimizes its subjective performance in a given system. The main problem is that when you have an active speaker most people will erroneous consider that the only important parameter for its amplifiers design is power, distortion as low as possible and a very high damping factor. Some fancy people as me will think a digital amplifier for the bass, a tube for the middle frequencies and a low power class A for the treble
- I helped a friend doing it in the 80's in a Mark Levinson HQD system. And most probably it will result in a complete disaster, it will lack coherence. I have no doubt that competent people will be able to design a great active speaker. But as far as I have listened to and read about, the best existing commercial systems are still using passive speakers with great speakers. I heard about great active systems that are on par or may be better than those commercial designs, but they are unique designs, custom tailored to great rooms by experts who perfected them during many years. They can not be considered for this debate, as they are not easily reproduced and depend excessively one the expertise or their single owners.
As Jack already suggested, if you want to design an active speaker you should start from zero, not taking the crossover of an existing passive speaker and implementing the passive in active form.
Component count is all too often a red herring. There are plenty of active line-level crossovers that add much less distortion than a simple passive speaker crossover. And vice-versa. You also have to consider the system; amp tied directly to driver vs. passing through a number of passive crossover components first. There are always trades...
It's a tough comparison, no doubt. And I completely agree with Jack that an active speaker is best designed from the ground up, not converted from a passive design. One thing always seems to be missing from these discussions, though -- the inherent advantages of passive design. I'm not talking about the flexibility to upgrade/change/choose electronic components, or the fact that there are really good-sounding passive systems. There's no denying that. I'm talking about a performance gain, a comparative advantage that comes from the passive approach. I've been in a few of these discussions; can't say I've ever seen anyone come up with one.
Tim
Tim
Hello Tim
I will give it a shot. In an analog world it's easier and much cheaper to do it in the speaker. Below is an example of the voltage drives for a horn system. All CD type horns need passive/active EQ to make the response flat in the passband. You can end up with a complex crossover.
Now in a digital world it's all a piece of cake with the DSP available.
Rob
Attachments
I don't quite follow. In that system, what is the performance advantage of using a passive speaker crossover instead of an active or passive line-level crossover (analog, for the sake of argument)?
I would like to know how that could be?? Please explain how the signal going to a tweeter, typically through one cap and a resistor could generate higher distortion than multiple gain stages and literally 10X the number of passive parts.
It just doesn't make sense to me.
Rob
Hello Don
The complexity of the midrange drivers voltage drive curve. Any out of the box analog could not replicate that curve. Typically they are all limited to simple slopes, some adjustable in the crossover region over one or more crossover points. You can't do notch filters or any complex curve.
Now digital you can do it all.
Rob
Last edited:
Perhaps this wasn't your point but I didn't find significant technical flaws in what you wrote.
I take the view that correctly implemented, a line level XO prior to separate amps for each driver would be better, but it does in practice need an additional driver/buffer stage than the speaker level passive XO - provided that circuit element is impeccable then active has the edge. Or if the source is digital then no contest at all, a passive line level filter directly after the DAC is superior to having one later in the signal chain, at high level.
That's easy. Cost.
BTW you said "I think "benefits" assumes good design and implementation.". I don't see how suddenly I can be dead wrong man. Face it. Active or passive, the RIGHT answer is "It Depends".
Cost is arguable, but out of scope; we're talking about sonic benefits.
You misunderstand. The only thing you were dead wrong about was my assumptions of your position. They were all questions, not assumptions. And "It Depends" is always the right answer, and rarely any kind of answer at all. It avoids saying, or even exploring whether or not one approach, when done properly, can be demonstrably better than the other, falling back on "you can screw anything up in execution." Of course you can. But like "it's all subjective" it leaves us with no standards and blocks our progress toward higher fidelity.
Tim
Hi
I must say that my position about Active Crossover was that they were always preferable. I may have to revise this position but let me try to get some points out:
A crossover is a representation of mathematical functions. Filters are maths equations. They can be simple or complicated function. My questions:
From an implementation viewpoint, don't active filters provide better approximation of said filter (crossovers) functions?
Aren't passive crossover very wasteful? (For the most part anyway)?
Can we compare the distortion induced by passive crossover to those of active? IOW which of the two distort more?
I will grant the point that the system becomes more complex, complicated, expensive,etc but in a world where people go the route of taking cables of the floor ...
The filter I favor Active Digital filters do away with circuit complexity At the output the DACs needs to be well though-out of course but we do away with both circuit complexity and signal corruption if the filtering is done in the digital domain. Those who absolutely want it to be analog may just drop this solution, for the others that are espousing digital more and more I believe this to be the best solution if active is in their agenda. I have been looking into DRC and I think the DeQX, TaCT and Lyngdorf provide such solutions... Not sure about the Lyngdorf entirely maybe only two ways. There are many others one very sophisticated and very well-thought of is the Behold from Germany a brand that has made some noise in the USA but not much these past few years ...
I must say that my position about Active Crossover was that they were always preferable. I may have to revise this position but let me try to get some points out:
A crossover is a representation of mathematical functions. Filters are maths equations. They can be simple or complicated function. My questions:
From an implementation viewpoint, don't active filters provide better approximation of said filter (crossovers) functions?
Aren't passive crossover very wasteful? (For the most part anyway)?
Can we compare the distortion induced by passive crossover to those of active? IOW which of the two distort more?
I will grant the point that the system becomes more complex, complicated, expensive,etc but in a world where people go the route of taking cables of the floor ...
The filter I favor Active Digital filters do away with circuit complexity At the output the DACs needs to be well though-out of course but we do away with both circuit complexity and signal corruption if the filtering is done in the digital domain. Those who absolutely want it to be analog may just drop this solution, for the others that are espousing digital more and more I believe this to be the best solution if active is in their agenda. I have been looking into DRC and I think the DeQX, TaCT and Lyngdorf provide such solutions... Not sure about the Lyngdorf entirely maybe only two ways. There are many others one very sophisticated and very well-thought of is the Behold from Germany a brand that has made some noise in the USA but not much these past few years ...
I would like to know how that could be?? Please explain how the signal going to a tweeter, typically through one cap and a resistor could generate higher distortion than multiple gain stages and literally 10X the number of passive parts.
It just doesn't make sense to me.
Rob
Rare but happens. Cheap parts (capacitor, in this case) add hysteresis (distortion) higher than the level added by the active components in a well-designed active crossover. More likely in other drivers, natch, where power (current) is higher. And don't forget you are now hooking the amplifier directly to the driver. I do not recall all the details, it was decades ago, but we ripped the crossover out of a big Infinity speaker (not the IRS, one step down) and the measured distortion was lower after we went all-active (I think it was a four-way, actually, but I am not quite sure).
I know there were a couple of other cases but I don't really recall any details. Some middle-of-the-line speakers then (1980's) used non-polarized electrolytics in the crossovers (cheaper, smaller, etc.) and they were not always bypassed well. They were nonlinear with voltage and current, adding distortion.
N.B. The distortion from the drivers still swamped any crossover component distortion, active, passive, line or speaker level. And of course if you overdrive the active crossover it will distort...
One other point: inductors can exhibit hysteresis much worse than capacitors when driven hard, and those were usually the largest source of distortion in high-power crossovers.
Similar threads
