WAJ on AUDIO - for truth in hifi / stereo / high-end audio

Home

.

BASSic ISSUES; Pt. 2: The Best Speaker-Crossover Is NO Damned Passive-Crossover!

by W.A.J.

.

In part 1 of this piece we examined the speaker-components, or parts, traditionally blamed for poor speaker-performance, at midbass in particular. Those parts are; drivers and enclosures. Of course, the probable cause of sub-standard performance would likely be a combination of 'wrongs' - not just one.  But, what if the fore-mentioned drivers and enclosures, which traditionally bear the brunt of the blame, are really the 'lesser' offenders, by far?

What if someone were to suggest that the ever-popular passive crossover network is quite possibly the very worst contraption ever to corrupt an audio-signal? What would be your reaction to that? What if it were demonstrated that no other component/part of an audio system does as much damage to the quality of the sound - especially at midbass? For those seeking the best sound; what if one were to suggest that the best thing one could do with the typical passive crossover is to avoid it like the proverbial plague?

Someone needs to say it - in no uncertain terms - and it's about time!

Indeed, why should the unsuspecting audiophile continue to seek the very best components; amp, pre-amp, front-end, etc., which one's money can buy, only to top it all off with what is well known and confirmed to be THE sonically-inferior second-best type of crossover system? Doesn't seem to make a whole lot of sense, does it?

Don't blame me, I'm only the messenger. I'm not the one who organized the reality. I'm only asking; What if...? Don't get me wrong, I'm no great advocate of the passive crossover, but I do believe there's a place for it. For instance, it's perfect for casual users who are content with modest systems. However, for those who seek the very ultimate in sound-quality, or for those seeking to make their modest systems the very best that they can be, there's arguably no place for the passive speaker-level crossover network. This is because the sklillful exclusion of said passive crossover system will always ensure a better performance - ALWAYS - especially at midbass, and at related transients.

Neither am I a great advocate of active systems, per se. (I'm not extremely enthusiastic about contaminating the performance of a carefully selected pre-amp, for instance, with the addition of an active electronic crossover-unit, or any other 'active' component, to the chain). But enduring the vagaries of the typical speaker-level passive X-O system because of this doesn't make much sense, either. And this is why, for now, I prefer experimenting with; passive line-level x-overs, at the front-end of the audio-system - PLLXO. But that's just my personal choice, for now. All that aside, it could be strongly argued, as irrefutable and demonstrable fact, that any good electronic X-O WILL also be better than the best speaker-level passive X-O, overall.

This is no well-kept secret. It is a well-known fact (to those who're interested) that; with all else being equal, an active system, or one bereft of a passive speaker-level crossover, will ALWAYS provide the better performance of the two. That is; again, with all else being equal, and with components of reasonable quality. Disregard those who disingenuously seek to distort the facts by comparing high-quality passively-driven audio-systems with low-quality actively-driven examples, by the way, and for obvious reasons.

Incidentally, there's very good reason why the very best speaker-systems of this world are active. Here are a few reminders of some examples which exist today; Magico Ultimate, AvantGarde Trio w BassHorn, Martion Orgon, etc., and etc. And, it should be interesting to note that some of these are offered, in passive configuration, as cheaper compromised options. What does that tell you, really? Of course, there are also excellent strictly-passively-driven aspirants to the very top-echelons, but... well..., refer to the-above!

As with most audiophiles, however, I've been guilty of not paying enough attention to the fact of an active system's intrinsic superiority, over the years. [Incidentally, this would be a cardinal-sin for those of us who may have previously owned/operated actively-driven mobile pro-sound systems, for instance. And excuses about not comparing apples to oranges, then, would be grossly inadequate. (No further comment on that).] As with most, I'd allowed myself to be misled by industry-propaganda purporting that passive systems can be 'just as good'.

Hogwash!

Fact is that active speaker-systems can be relatively expensive. And, mainly for this reason, certain industry-players (speaker manufacturers and the mainstream audio-press, for instance) have a vested interest in promoting the false notion that (cheaper to build, and easier to sell - with a higher profit-margin) passively-driven speaker-systems can and do provide top-quality performance, at all levels, up to and including the very ultimate level

This is simply not true.

Truth is; regardless of how good a passive speaker-system may be, it's an unmitigated fact (supported by the laws of physics) that it WILL be significantly improved by its conversion to active operation - with necessary adjustments, of course. (These 'necessary adjustments', incidentally, will have been caused by the need to correct measures implemented to compensate for the ills of the passive crossover, itself. Ironic, isn't it?) The prudent audiophile will alway be concerned that a passive's performance may always be improved - most times; drastically improved. Crossover performance is NEVER the very best that it can be, with a speaker-level passive system.

With an active system, on the other hand, one can be confident that the very ultimate in performance is being elicited from one's system. Nothing can be improved on that front. Any improvement must be sought in other aspects of performance since crossover performance (even with the least-costly reputable unit) is already THE very best that it can ever be - within the bass-oriented context of this article. Such inexpensive electronic crossovers will suit the vast majority of users, even as they easily outperform the best passives.

I'm not sure I could ever be much more explicit than that.

But for the fussy few (like me) for whom the preservation of very minor nuances is a very major issue, there are the options of; either expensive electronic crossover-units like those from Audio-Research or Marchand, or there's also the very inexpensive non-invasive option of the fore-mentioned PLLX-O type arrangement. (Incidentally, my own concerns about the nuances relate mainly to the low-mids, which most popular modern conventional speakers do not, and cannot, reproduce anyway - see elsewhere, for expansion on this point - rendering such concerns redundant, in the majority of instances. Outside of those concerns - which, as I said, are outside the scope of most conventional modern systems - it's quite possible that even an inexpensive active electronic x-o would have suited me fine. I suppose each individual will have to decide for one's self on the suitablity of the different options).

Throughout this piece, I'll be showcasing my voyage of rediscovery, so to speak. I believe it may be more effective and illustrative if I were to reprint snippets of previous articles and even my emails to an audiophile-friend highlighting my moments of epiphany as I rediscovered how much better speakers and drivers actually perform without the encumbrance of a passive crossover network.

Note that I will not be addressing the nitty-gritty, nuts & bolts, or general mechanics of this subject. For that I'll refer you to one of the excellent articles which specifically (if more restrained, politically-correct, and tactfully) addresses those issues: See here; "The Benefits of an Active Speaker Lifestyle", for one example! [Unfortunately, that article seems to limit consideration to active, all-inclusive amp/speaker combinations, which limit flexibility - i.e those with the amps, etc., built into the speaker box. (Refer to the segment they cite as a disadvantage for active systems; "Whatever You've Got, You're Stuck With It") This article does not. From our perspective, we consider anything; from that type of 'all-inclusive' active system, to a 'single-driver' speaker-system with separate amp, to a customized multi-way 'active' system similar to my own (2-way + subs system) in which all the components are separate (amps, PLLX/Os, and speakers/drivers) in addition to being easily-accessible and readily conducive to adjustment and change, where necessary - i.e. even more flexible than the typical passive system, in fact.]

Another similar article is linked here.

And a user-review of an active crossover is here. Note the user's comparison with his previous 'top of the line' outboard passive crossover-units, which came with his expensive Martin-Logan speaker-system. Most telling is his comment at the very end, after getting used to the tight, realistic, and much-superior performance of his newly-acquired active crossover unit: "PS- I tried the passives again for the heck of it, and almost gagged. Yuck, they were all wrong. Hard to believe the factory sends them out into the world this way".   

While generally eschewing the 'nuts & bolts', I will say at the outset, though, that I believe the inductor of the passive crossover is what does the most damage, and the most recognizable damage, to the sound. And I will also say that bass - specifically; midbass transients, especially - suffers the most damage from the inductors of passive crossovers. The inductor compromises an amplifier's ability to control, or damp, a driver's movements - they all do, to varying degrees, and without exception. The speaker's cone becomes a semi-controlled 'loose-cannon'. Even the world's greatest amp is undermined, in this way, by the inductor(s) of the world's 'best' passive speaker-level crossover - it's an inevitable fact of life. And even though air-core inductors are arguably better than the typical ferrite-core types, they still compromise the sound, significantly, compared to the rigid, accurate, and uncompromised control afforded by the active alternative.

Indeed, this whole situation could be likened to that of a driver and his car: The driver, with his hands and feet, is in direct control of the vehicle - just as an amp is in direct control of an actively (directly) driven speaker-unit.  Imagine how erratic his control of that vehicle would be if loose coil-springs or very thick and soft sponge-pads, for instance, where inserted between his points of contact with the controls of the vehicle. (Let's say the palms of the driver's hands are now strapped to 10"-thick sponge-pads, and so are the soles of his feet). The driver (or amplifier) would not be able to control the vehicle (speaker) in a precise manner, now, would he? The result would be; very erratic/imprecise steering, especially. This is analogous to the loss of control caused by the decoupling of the direct drive of an amp, to the speaker, by inserting a lossy imprecise 'sponge-pad' of an inductor between said amp and speaker-unit. 

Yet, with all that being said, some of the best and most expensive passives can be very good, for sure. But still...  

...And even the very best passive will always be second-best - without question - moreover, there'll always be that nagging element of a 'sponge-pad' between amp and speaker-unit(s)!

I should point-out also that, while it's true that a satellite-system with self-powered subs (operating below 150 hz, or there-about) could be deemed 'partially active', the irony is that it's the frequency-set handled by the typical mid-woofer which amounts to THE most critical, in my experience, and in the context of this piece - re; part 1.

Bearing in mind that this is a continuation of the article; 'BASSic ISSUES; Pt 1....' I'm, therefore, assuming the reader already has a basic concept of my own system, and continuing from there: Very briefly; the mid-woofer segment of my system combines dis-similar drivers (i.e. two KLH 12"ers, and one of Yamaha's 7" NS10s, per side) operating over exactly the same frequency-range (each type of driver excelling at its own unique traits - and complementing the other - over this range). These mid-woofers are augmented by an Altec mid/tweeter horn, per side, and Goodmans 18" subs in large cabs. 

Moments of Discovery: As my first example of what crossover-less operation can accomplish, here's a reprint of my comments regarding the crossover-less operation of the NS10 mid-woofers of my own system, years ago, from the article, "DIY Speaker-Systems....":

A recent experiment seeking the benefits of crossover-less design has yielded extra-ordinary results. The little white-coned Yamaha that operates as a mid-woofer in each channel was divested of its crossover network and, because of this, believe it or not, the sound of the whole system has improved – throughout the whole spectrum, excepting deep bass and treble (the Altec tweeter was already 'x-overless' with only a capacitor limiting lower frequencies). What most manufacturers, and many in the press, will not tell you is that; despite some benefits, ALL passive crossover networks diminish detail, dull transients, obscure clarity, add distortion, and strangle dynamics to varying degrees. I may elaborate in another article, but suffice it to say that all the claims made for crossover-less designs are now proven to be true – for me, at least.

This simple operation has taken the system's performance to another level. The gains in transient-response, dynamics, and overall clarity and detail, are truly exceptional....  .... – it’s like having the best of both worlds as the x-overs on the KLH’s do contribute to shaping the exquisitely natural sound of these drivers. (Or so I thought, at the time - Ed). Nevertheless, one never knows how much music is lost in conventional designs until one decides to go CROSSOVER-LESS, in some sense.

Up until recently, I'd been quite critical of the midbass performance of my own KLH 12"ers (though full of praise for their low-mids prowess). Here's an example of my disparaging remarks from the same article:

As it is, one NS10 operates purely as a woofer, and the other operates as a mid/woofer in each channel. The outstandingly accurate mid-bass from these does complement, and at the same time qualitatively override, the decidedly ordinary, slightly woolly mid-bass of the KLH's.  (KLH's have the sort of mid-bass most hifi speakers have - soft, round, and very pleasant, or pretty, like that of a highly-rated Spendor BC1 system, the best domestic speaker I'd previously owned - totally inaccurate).

Here's another example from Pt,1 of this piece: 

The cheap KLHs' only reason for existence in my system is for their unparalleled low-mids performance, for instance - nothing else - other strengths are garnered from other drivers and strategies. The low-mids of the KLH are more realistic than that of ANY other driver or  speaker-system I've EVER encountered, regardless of cost - and I've encountered quite a few. Midbass of the KLH is crap, tho, so it's purposely subdued in order to be over-rided and handled outstandingly in my system by the fore-mentioned NS10/Goodmans combo.

More recently, I'd said this, in Pt1 of this article, by which time I was coming to the realization that the crossover may have a part in the KLHs' deplorable midbass performance:

 For instance; a JBL 4430 studio-monitor, which I tested against my cheap KLH, displayed very similar bass/midbass to that of the unmodified KLH, with its cheap crossover 'mucking-up' the response in a monumental manner, as I've recently discovered. And the handicapped KLH, itself, displayed very similar bass/midbass to the Spendor (similarly 'soft') which epitomizes the midbass of most popular speakers.

The following is part of a recent email to my audiophile-friend, mentioned above - sent right after I discovered just how much the passive crossovers really undermined the performance of the KLH drivers:

I believe that, whenever or where-ever possible, it's always best to operate without a passive crossover, especially where bass is concerned. Oh sure, there are benefits in other areas (e.g. dynamism, distortion, an general transient response) but it's in the upper to mid bass where the benefits of crossover-less operation are most blatantly and immediately obvious, and most beneficial to overall realism. Drums, for instance, suddenly become startlingly realistic.

Short story (I promise). My practice of using the NS10 drivers in my system, crossover-less, came to an end some time after I incorporated the 811B horns, which now went low enough to interact with the NS10's un-attenuated higher frequencies. This was untenable since, though the anomaly was only apparent on certain tracks at certain frequencies, the aberration was intolerable. So a passive crossover is now connected to tame this effect. It's also tamed the impact and realism of the fore-mentioned drums (at the leading edges, in particular) for instance, though still above average.


However (and cutting a long story shorter) I also have a second 'system' connected to a TV in another room. This week, I connected (crossover-less) the Magnavox/Jensen 12s I'd gotten in the deal with the 93-Series amp, recently. (They sounded awful, btw - very efficient, but awful - at least; up to now). But then, I also tried a spare KLH driver, identical to the four in the main system. Crossover-less, I've now discovered this driver to be awesome and encouraging. Encouraging; because, unlike the NS10 (which is naturally midrange-prominent) the KLH 12 is naturally midrange-reticent. This means that there's much less of a likelihood that it would interact with the 811 in the way the NS10s did. Bearing this in mind, I can't wait to experiment with removing the crossovers from the 4 KLH 12s in the main system. If they can be made to operate in this way (in harmony with the horns) I can just imagine how much more awesome the midbass, in particular, would be.

And here are my emailed comments to my friend right after I'd experimented with removing the crossovers from the 4 KLHs of the main system, as promised:

Regarding the removal of the crossover-networks from the 4 KLH 12s, I'd say the experiment is a resounding success - should have tried this long ago. Only a very slight increase in middle-mids to adjust for - this should also mean a very slight increase in efficiency too. (Later - much later - I'll probably measure all this with a Real-Time Analyzer, just to confirm that all's well). The outstanding benefits of this configuration (as against the detriments of a crossover) causes me to wonder why KLH saw fit to contaminate the sound of these drivers with crossovers at all. These drivers really don't need them since their natural roll-off presents a near-perfect slope to complement a properly implemented mid/high driver. It's almost impossible to over-state how good these under-rated drivers are (used in multiples, for efficiency) in my humble opinion, and how much BETTER they are WITHOUT CROSSOVERS.

I'll leave it there, for now. Suffice it to say; crossovers will never again be connected to the KLH 12s. Middle-mids and (upper) midbass are now even more outstanding. (The excellent NS10s which previously excelled in these regions, in this system, are now arguably redundant). Perhaps I should elaborate in my next hifi article.  I won't say the difference is 'night & day', especially compared to when the system operated with crossover-less NS10s and 'active' 18" woofers. But the benefits are absolutely worthwhile, even in clarifying those low-mid notes (piano, etc) from the KLH 12s - formerly slightly blurred and obscured by the crossovers. (I'd previously accepted this as a characteristic of those drivers, mistakenly).

Needless to say, the most obvious benefit is in midbass transient-response
....

Hopefully, the fore-going renders a clear illustration of the very vast improvement wrought by the simple exclusion of passive speaker-level crossover networks.

'Night & Day' Differences? As I'd mentioned before, the gains aren't 'night & day', compared to what I'd had before (i.e. with the NS10s operating crossover-less, and augmented by the active 18" woofers). But, compared to the KLHs own performance at midbass, previously, with their crossovers, the current crossover-less performance does, indeed, present a 'night & day' difference. Absolutely. With their crossovers, the KLHs' performance at midbass was sluggish, to say the least. The leading edges at midbass were severely dulled, softened, rounded, slow, mushy and woolly. Pleasant to some, but nothing, at all, resembling real-world midbass and, especially, real-world midbass-transients. In this regard, the KLHs were very similar to most other speakers, at midbass - not even remotely close to the real thing, by any stretch of the imagination.

Yet, by just the simple change to operating without their degrading passive crossover networks, the performance of these drivers has been transformed - the very opposite of all those disparaging remarks is now true - 'quick', 'fast', tight, and 'sharp', by themselves, are grossly inadequate adjectives, really. If I tell you that the KLHs (in conjunction with the 18" subs, as usual) now very closely emulate the sound and characteristics of a real bass-drum, for instance, then there's not much more I can say to emphasize the quality they now display - as only a pitiful handful of speakers and speaker-systems can actually do this. OK, I'll just single-out; exemplary transient-response along with near-lifelike impact, and I'll leave it at that. Audiophiles routinely pay many thousands of dollars for the equivalent of a mere fraction of the benefits gained from simply excluding a passive crossover, as I've illustrated, here.

Incidentally, the difference with the NS10 driver was very similar. And I do prefer its extra measure of leading-edge detail and realistic upper-midbass tone (though the KLHs remain unchallenged at lo-mids - with or without passives, here). [I should add that, regardless of how universally acknowledged NS10s are in their acuracy, as comprehensive and exhaustive tests have confirmed, this accuracy is extremely enhanced whenever their mid-woofers are operated without passive crossovers - no other driver I've ever heard comes remotely close. And they harbor enormous potential to seriously and significantly influence the overall quality of very much larger systems than their 7" size suggests - truly extra-ordinary drivers, really.] But, then, it's a very similar, if somewhat less spectacular, story with every single driver I've experimented with since my first discoveries - they're ALL more realistic without speaker-level X-Os, by far.

Perhaps this is what the late great J.G. Holt was alluding to when he suggested that (crossover-less) table-top radios, and the like, somehow render more believable facsimiles of real-world sounds, to a certain degree. Here's a link to that article; "Getting the Notes Right" - refer to the 4th-from-last paragraph at page 2. In that piece, he rightly stresses the importance of the midrange. But I also believe that his allusion to the believability of table-top radios necessarily extends from the midrange to include their better depiction of upper-bass and upper-midbass, along with; transients, detail, and even tone, in many instances. Listen and you'll 'see'!

.
The 'Attack' & 'Steady-State' Phases; Their Relevance to the 'Leading-Edges' and to HiFi, in general, and to Crossovers, in particular:  Ah, yes! The leading-edges. How important are they, really?

I'd suggest they're extremely and/or critically important. But first, what are they? I suspect most will say the 'leading-edges' manifest the sound of the very first instance of 'attack' registered by each note of a musical instrument being reproduced by an audio-system.

And I'd perhaps register my agreement with those who'd suggest that the leading-edges are reproduced by the tweeters of our audio systems, generally. Indeed, even if we were to disconnect all other drivers excepting the tweeters, elements of the initial strike of a bass-drum, for instance (the leading-edges) would be clearly evident thru said tweeters - assuming typical tweeter-crossover points.

But the $64 question is; do the leading-edges stop there at the very instance of 'attack' (as reproduced by the tweeters) or do they extend past there to include somewhat lower frequencies and more of the substance of the transient, timbre, and tone of the instrument being reproduced? I'd suggest that they do. If not, then they should.

I'm no musical expert, but I'd suggest that the audiophile term; 'leading-edges' equates exactly to the musical term 'attack-phase'. It is my understanding that the 'attack-phase' stems directly from the initial action of the musician in elicting sound from his instrument. After the 'attack-phase' (leading-edges) comes the next stage; the 'steady-state' phase. This is the phase where the (acoustic) instrument resonates after the initial attack, and it's where timbre, harmonics, and tone, etc., are most evident.

My current contention is that; if our definition of 'leading edges' doesn't already include the initial traces of the 'steady-state' phase (i.e. initial traces of tone, timbre, harmonics, etc) then it really should.

I say this because of what I've discovered in my crossover-less experiments with the NS10 drivers (and the KLHs, to a lesser degree). Concentrating on just the midbass transients of the ubiquitous kick-drum, as is my recent custom for this topic, passive crossovers on these drivers resulted in not only the significant 'dulling' of the leading edges but also the dulling of the initial traces of the steady-state timbre and tone. I'm sure of this because whenever the crossovers were eliminated these regions suddenly seemed outstandingly vibrant and even more accurate. So very detailed and accurate were they that very minor differences in instruments could now be identified. Especially with the NS10s, one now felt it possible to identify even the different types of mallets and kick-drums being used. And this vast improvement is wrought merely by clarification of leading-edges AND the initial traces of steady-state timbre and tone - absolutely - brought about by the elimination of the passive crossover. 

I repeat; the vast increase in accuracy resulted from the clarification of the leading-edges AND the initial traces of the steady-state stage (timbre and tone). I reiterate; that is a combination of BOTH the 'leading-edges' AND the initial traces of the 'steady-state' phase.

I really doubt this point can ever be repeated often enough. So, here we go again...

...For this term 'leading-edges' to mean anything really worthwhile, then the definition of 'leading edges' MUST include reference to the initial traces of 'steady-state' tone generated at the very beginning of each musical note.

And it should be clearly and properly defined if only because this is the  important and critical region which literally details & defines the identity and character of whichever  instrument/note is being played/reproduced - significant compromise, here, should never be tolerated by the serious audiophile. 

In addition to the points made, here, and from the evidence of a paraphrased excerpt provided below, we'll recognize that any compromise in this region significantly compromises the whole of the overall reproduction - every single note is detrimentally affected. However, since this article concentrates on bass/midbass reproduction, I'll simply point out, here, that this is exactly how and where passive speaker-level crossovers do the most damage - they dull the leading-edges of bass notes and severely undermine the initial traces of steady-state tone, timbre, harmonics, and etcetera. 

At one stage of Pt-1, I'd pointed-out that the omission of the outstanding crossover-less NS10 driver (leaving the other woofers; the KLH mid-woofer, with its, then, passive crossover, and the active 18" subs, similar in concept to most ) resulted in very 'ordinary', 'detail-less', and 'tone-deprived' midbass - much like that of most speaker-systems. Bass-transients, especially, became  blunt generic 'blobs' of sound, devoid of distinguishing characteristics. In other words, omitting the outstanding transient and tonal detail displayed by the crossover-less NS10 at the very beginning of bass notes severely compromised the quality of the overall bass-reproduction.  

In bolstering this particular argument, I'll paraphrase a segment of a book my fore-mentioned audiophile-friend sent me; the popular best-seller: "This Is Your Brain On Music", by the award-winning neuroscientist, author, record-producer, and musician; Dr. Daniel J. Levitin. Refer to  the bottom of page 50, and continue from there. According to this account, the elimination of the attack-phase (leading-edges) rendered instrumental sounds virtually unrecognizable. (Similar to the effect with speakers/crossovers, previously described). This is irrefutable proof as to how very important this region really is.

[In agreeing with this, I'm merely and similarly adding (to our popular concept of 'the leading-edges') the initial traces of the steady-state phase - i.e. including timbre/tone - to the equation, as a factor compromised by passive crossovers and impacting the performance of audioplile loudspeaker-systems. I'd also point to the likelihood of a fact; that the 'attack-phase', in music, similarly consists of initial tone/timbre (not just the nude 'attack' itself) based on the implications and results of certain experiments highlighted below. Refer to the allusion to timbre, specifically, in the first sentence in italic, below, and to the the results after cutting away the beginnings of musical notes.]

The bold-faced first and last paragraphs of this excerpt are the most relevant. Dr. Levitin tells us..:

.
.
.
The avant-garde composer Pierre Schaeffer performed some critical
experiments in the 1950s that demonstrated an important attribute of
timbre in his famous “cut bell” experiments. He recorded a number
of orchestral instruments on tape. Then, using a razor blade, he cut
the beginnings off of these sounds. This first part of a musical instrument
sound is called the attack; this is the sound of the initial hit,
strum, bowing, or blowing that causes the instrument to make sound.

The action our body takes in order to create sound from an instrument
has an important influence on the sound the instrument makes. But
most of that dies away after the first few seconds. Nearly all of the actions
we take to produce a sound are impulsive—they involve short,
punctuated bursts of activity. In percussion instruments, the musician
typically does not remain in contact with the instrument after this initial
burst. In wind instruments and bowed instruments, though,
the musician continues to be in contact with the instrument after the initial
impulsive contact—the moment when the air burst first leaves his
mouth or the bow first contacts the string; the continued blowing and
bowing has a smooth, continuous, and less impulsive quality.

The application of energy to an instrument—the attack phase—
usually creates energy at many different frequencies that are not related
to one another by simple integer multiples. Put another way; for the brief
period after we strike, blow into, pluck, or otherwise cause an instrument
to start making sound, the impact itself has a rather noisy quality
that is not particularly musical—more like the sound of a hammer hitting
a piece of wood, say, than like a hammer hitting a bell or a piano string,
or like the sound of wind rushing through a tube. Coming after the attack is
a more stable phase in which the musical tone takes on the orderly pattern
of overtone frequencies as the metal or wood (or other material)
that the instrument is made out of starts to resonate. This middle part of
a musical tone is referred to as the steady state. In most cases the
overtone profile is relatively stable while the sound emanates from the
instrument during this time.

After Schaeffer edited out the attack of orchestral instrument recordings,
he played back the tape and found that it was nearly impossible for
most people to identify the instrument. Without the at-
tack, pianos and bells sounded remarkably unlike pianos and bells, and
remarkably similar to one another. If one splices the attack of one instrument
onto the steady state, or body, from another, one gets varied results:
In some cases, one hears an ambiguous hybrid instrument that sounds
more like the instrument that the attack came from than the one the
steady state came from. Michelle Castellengo and others have discovered
that one can create entirely new instruments this way; for instance, splicing
a violin bow sound onto a flute tone creates a sound that strongly resembles
a street organ. These experiments showed the
importance of the attack.

.

.

.
Obviously, I agree with all that is said in this paraphrased excerpt from the book. Note that it coincides exactly with all I've been saying about the effects of passive crossovers. Indeed, the terms; 'razor-blade' and 'passive-crossover' are virtually inter-changable, based on the effects of both, in this context. Apart from all their other ills, regarding dynamics, distortion, and the like, passive high-level crossovers typically compromise exactly the regions discussed in this excerpt, in exactly the same ways. And my contention is that this degradation is MOST obvious at midbass, in HiFi - caused mainly and specifically by the effects of the inductors of said passive crossovers. At least; this is what my own experiments indicate. Absolutely!

 

Conclusion: Competent readers who fancy similar experiments with spare speaker-systems should go right ahead - simply bypass the speaker-system's passive crossover to the mid-woofer and I guarantee the bass will tighten-up significantly, among other things. (Do not include the tweeter in this experiment, please - unless one intends to; 'put a cap on it'). And though you'll have to revert to using the crossover, due to the likelihood of the system now being grossly out of balance, sonically, at least the point will have been proven regarding the significant benefits of crossover-less operation, especially at midbass. And I'm also willing to bet that even my old Spendor BC-1 could have been drastically improved in a similar manner. I'm sure of it!

In fact, based on what I've learned, I'm confident that ANY speaker-system can and will be improved, at least to some degree, by simply eliminating the passive crossover network (necessary fine-tuning notwithstanding). Not many will be able to operate satisfactorily without a crossover (my own drivers also need to be guided, ultimately, despite first appearances) but the use of even modest but reputable electronic crossover units, instead of speaker-level passives, WILL elicit significant benefits in performance, as I've suggested throughout this piece - all else being equal.

The mystery alluded to in part 1 of this piece is now solved: The overwhelmingly popular speaker-level passive crossover is THE main reason for the pathetic and deplorable midbass (and midbass-transient) performance of the very vast majority of the world's hifi speaker-systems - now; without a hint or shadow of doubt!

I'm not suggesting that people go digging the passive crossover-networks out of their treasured and expensive speaker-systems, unless they know exactly what they're doing. But, for those contemplating a DIY speaker-build, for instance, my advice would be; not to even consider a passive crossover. 

Others contemplating the piecemeal purchase of a complete audio system may well consider investigating the 'active' option, at the very outset. Many passive speaker-systems may also be easily converted to active operation - re; the link to that Martin-Logan owner, for an example. And the addition of an electronic crossover-unit and amp(s) need not be prohibitively expensive, if this is the route one opts for. Nevertheless, one should always be aware that the 'passive' route always involves compromise, especially at midbass, and related transients, as we've seen - very significant and debilitating compromise, in the very vast majority of cases.

I'll go even further to say that a lot of the blame heaped on speaker-types, cabinets, and drivers, for a lack of quality in performance, should really be heaped on passive speaker-level crossover-networks.

Finally, and most importantly, consider also how many thousands of dollars have been  wasted (and are still being wasted - even as we 'speak') on upgrading front-ends, pre-amps, amps, wire, whole speaker-systems, and etcetera, in futile attempts to improve on the problems presented by the mere presence and effects of the notorious passive crossover, in otherwise-excellent audio-systems.

I'll leave it at that!

Home

Copyright 2013