Adding another way

August 3, 2018
 by Paul McGowan

When loudspeakers were first introduced they were essential 1-way designs: a single driver that carried the limited frequency range possible in those days. Soon we added a tweeter to augment the single driver's high-frequency response in what became known as a 2-way and there the state of the art sat for many years.

Hardly satisfied with a woofer's performance in the demanding midrange area a third way was later added in the form of a midrange driver that fit in between the woofer and tweeter and thus the 3-way was born.

It is certainly possible to add even more ways and speaker designers have by the addition of internal subwoofers for the lowest octaves and super tweeters for the ultra high harmonics.

What all these ways have in common is fundamental to their task of producing a full range musical performance. They divide up in ever smaller chunks the job of frequency reproduction. Regardless of the number of frequency divisions, not much has changed in the art of loudspeaker design for the last 50 years. That is until Infinity founder Arnie Nudell introduced the Variable Midbass Coupler.

Arnie always had a love of midbass (the range of frequencies between 100Hz and 600Hz) because this range is the basis of tonal balance. It's fundamental to voices and many primary instruments and, coincidentally, where the loudest peaks of music happen. When we set up a pair of speakers it is this midbass region we struggle with most. When the left and right speakers are too far apart the midbass sounds thin and we think of it as strident or anemic. Too close together and it's the opposite requiring a diet to remove unwanted fat. Yet, every time we move the speakers for best tonality we screw up the imaging. It is a never-ending battle.

Arnie's invention of the VMC changed everything. By building 4-way speakers with an internally amplified subwoofer and VMC, the difficult setup process suddenly vanished. Now it was possible to place the speaker pair where it imaged best and adjust low bass with the subwoofer controls and tonality with a turn of the VMC control. More than that, by separately amplifying the VMC it would be possible to achieve stunning amplitude levels in the very area nearly every speaker on the planet cannot come close to reaching.

When we launch the line of AN speakers in 2019, you will have your first chance at experiencing for yourself the power of Arnie's invention that we believe will fundamentally change our expectations of music reproduction in the home.

Stay tuned.

Subscribe to Paul's Posts

40 comments on “Adding another way”

  1. I’m sure this will be a major breakthrough for many!

    Ultimately someone cares for tonal configurability of speakers aside of fully active concepts and even extend it. With a separate mid bass and bass (DSP) adjustment option THE (more or less big) problem of most audiophiles get’s solved. It’s even better than my current option of leveling highs, mids, bass and sub separately.

  2. Reading these posts of the last days, I assume placement of these speakers in the next (2019) audio show is a breeze.
    What could possibly go wrong ?

    1. Always a cheerful experience reading your responses, jb ; )

      Maybe they’ll be smart enough to wheel whatever pair they’re showing around into different rooms at PS and beta tester’s homes and practice first.

    2. Regarding tonality aspects I would indeed expect a much easier good result.
      Certainly any speaker in a marble plated or with soft walls and carpet fitted room then still sounds nearly as bad as it should. And certainly this easier matching of bass to room and customer preference will be valid for a small sweet spot only, as bass nodes change a lot throughout a room.
      So one should expect improvements within physical realities.

      Remember: today most listeners have to change cabling and components each time they need to fit a change within the setup to the room and their preference again.

  3. So Paul - is the VMC a bit of Arnie’s work that has not been used in a speaker before? I searched it, but unclear. Sorry if I’m simply ignorant of Infinity design history and so forth. Or is that tomorrow’s post? ; )

    1. Not to my knowledge. I first learned of the idea about a decade ago in a pair of prototypes he developed. There it was just a one-off idea he had tried. But, when he built the last prototype, the IRS Killers we use as our reference, the VMC became the centerpiece and its power suddenly dawned on us all.

      Near the end, Arnie was experimenting with different amplifier options including slapping a vacuum tube in the front end of the VMC amplifiers. He became obsessed with the VMC and for good reason. It really works and, I believe, the separate amplification of the drivers and the ability of the user to adjust the level was pretty much the secret to the stunning power of those prototypes.

      I remember a session with the Mahler 3d symphony from the San Francisco Symphony. If you know the piece there's excitement when the double basses come in. Their low energy is the foundation of this opening bit and it wasn't quite right. Not enough "oomph" in those lower registers. Arnie stopped the piece, fiddled with the VMC, started it over and magic happened. I was floored at the difference.

      I don't know of any loudspeaker other than that prototype that had this control and midbass performance. It really is an extraordinary thing to hear when it's done right.

      1. Any active crossover system has this ability.

        I believe it only matters for signals that are correlated between left and right. Try listening in mono on one speaker, and I suspect the "magic" of lower midrange level adjustment will be less than using both speakers.

      2. That’s interesting, thanks Paul. Though it may generate more questions than it answers.

        Is the assumption that one will basically tweak it for the room/setup, and you’re done? Or every time you put on a certain piece of music you’ll be fiddling with it before you can listen to the fiddlers?

        1. At least I can tell you, yes, you could do it with every recording, but practically you do it once for the room and an average of recordings and maybe tweak for a few extreme deviations (a Hardbop recording has another frequency spectrum than a Roger Waters or Donald Fagen one, and although they should be mixed/mastered to a certain toality standard, they mostly are not). IMO one practiclly can't listen to whole genres of certain decades if the system is matched to todays audiophile recordings.

          As I stated before, in my case of active speakers I can level adjust the AMT tweeter, the mid and bass chassis as well as certainly the sub (parametric) as well as the Q of the bass chassis in the speaker and an optional tweeter EQ. Alltogether this is a lot to influence, but at the low end still not as optimal as Paul's add. tweakable midbass coupler and DSP bass.

          From that experience I can tell I would never ever buy a speaker anymore without at least various bass adjustment options to fit it to the room. I bet the option of Pauls speakers (which is even more than others with DSP'able active bass) will change High End life of all their owners...they will definitely ask themselves how the could ever use a non adjustable speaker and not only a few will recognize that they selected their upstream gear by the wrong criterias, as they can compensate tonality issues with the speakers then!

          AND I think equipment voiced with such a configurable speaker might sound not rich enough in quite somne other speaker combinations.

        2. The occasional use to match certain recordings with special tonality and to match to different listener positions in the room is also the reason why I’d vote for a few presets at least for the DSP.

  4. Paul,

    I admire your drive and ambition to push the state of art in sound reproduction. You couldn’t have taken on a bigger challenge had you gone from making piccolos to pianos.

    Will you be assembling the ANs in your new facility in Boulder? Are you able to keep all parts of USA origin?

  5. True innovation is going to happen when fully active speakers (with volume control) are mainstream with wireless connectivity (Source & speaker) done in a true audiophile way. Digital source straight to the speakers. That eliminates so many problematic areas. We are stuck. The big audiophile innovations are happening within a 60 year old “landscape box”. If Apple can make two small AirPods work in stereo (with room correction) sound as good as they do, I can only imagine how good it would sound if the audiophile industry embraced it. But, that would eliminate the need for wires, equipment racks, isolation devices, separate amps/preamps. No one wants to turn the industry on its ear like that. Whole hosts of companies would disappear.

    1. If you address only “innovation” and “HomePod sound quality” as reference, you might be right. But perhaps you never heard a true audiophile system? Digital components create many unwanted side effect degrading the sound. Just look at the LAN rover approach of PS Audio. I recently added a fitler device between my 10 k$ DAC and my amp and got a significant improvement of sound quality. Or just listen to a pure analog system with a master tape source. The true advantage of wireless active systems is user comfort as youmentioned already. And concerning roo acoustics: I am not aware of a room correction software that is based on a convincing psychacoustic idea working for every room geometry. And interesting idea is now launched by PSI, Switzerland offering an active absorber.

    2. Although I would welcome to get rid of cabling, it wouldn’t be that much of an innovation and importance to me in a fixed high end setup to be honest...I see it more important with mobile equipment.

      And I think the current level of sound quality will take a while to be transferred to wireless unfortunately. I have no clue but I think if HF and other radiation even has such strong impact to equipment, cabling and room, how much must it have to a wireless connection which can’t be protected from it in my imagination. And how much harm does that wireless connection itself.

      A small speaker making big sound would be my wish for innovation 😉

      1. Jazznut... You might want to consider the Audience 1+1V2+ speakers if you want small speakers with a big sound. But, you will still need a REL subwoofer (best match) to get there.

  6. Related to the new AN-1:

    It is curious that after several decades, A. Nudell has resumed the concept of using the mid-bass coupler that he used at the end of the 70s with the Infinity models: QLS and Quantum 2, which at that time (although the diameter: 4” of this unit was very small for the purpose) it served to set a precedent in the design of speakers.

    This fact alone constitutes, in itself, an improvement over the IRS V, so it is expected to be a commercial success within what many call Hi-End, given the experience of the team working on the project and that continues with the improvement of Nudell's conceptual principles from earlier times of his career.

    Personally I am convinced, and I have done so, (first with active analogue X Overs and today with DSP), that the correct approach in loudspeaker design involves the use of 4 ways, in which the fourth way must be implemented with a mid-bass coupler, which must have an adequate Sd and to which assign the frequencies corresponding to the bass notes of the instruments of a symphony orchestra, such as cello, bassoon, etc. so that they are convincingly reproduced.

    This does not mean the elimination of additional speakers that reproduce realistically the lowest notes of that orchestra, since the tonal balance is considered a categorical imperative.

    In addition, the use of passive X Overs is, given the current digital technique, a total anachronism and must use as many amplifiers as are necessary to adequately cover each way, thus eliminating the requirement that they are high power.

    All this, and much more, is possible nowadays, and with great solvency when incorporating a DSP in the system.

  7. I believe what you are talking about is changes in radiation resistance. There is a bump up to 6dB as two diaphragms move in phase from closer together than a wavelength, just like proximity to boundaries. When you have a mirrored pressure source, the compliance of the air halves, which increases the efficiency.

    If you are trying to compensate this frequency response variation, it seems you need a variable shelf frequency - which means another active crossover and amp pair, switched passive crossover or possibly an auto-transformer with a wiper like a Variac.

    Audiophiles and audio engineers develop an unnatural sensitivity to frequency response. When I hear a violin in a room or through a speaker, my mind is filling in any missing frequencies from my memory so the comb filtering of room acoustics and the vagaries of frequency domain reproduction are compensated.

    When listening acoustically, any frequencies that are attenuated or accentuated are information about the room boundaries. When there is cancellation or superposition, one of the waves always arrives first. This is a wealth of information that is scrambled by the two channel paradigm.

    The cancellation and superposition of waves from speakers in time domain identifies where the speakers are, which collapses the "imaging". Speakers have evolved to be locationally vague by scrambling phase response off-axis and stereo listeners learn to ignore the residual timing information.

    Time is also mangled by every knob or process on the signal including panning, which gives audio listeners another reason to ignore phase and stick to the frequency domain. One can hardly find a recording that has not been mixed and mastered.

    I further insist that 6" speakers need to be employed ABOVE 600Hz for full orchestral peak levels. I use 10" or dual 8" midranges, and my mid-bass couplers are 12". 6" with sufficient Dumax may be able to push the requisite air volume, but they will muddy from Doppler distortion.

    Your midrange drivers have over 600in2 of diaphragm, the equivalent of five 6" drivers. Does it not make sense to have more Sd for the mid-bass couplers?

    1. The sound quality degradation by multi way design loudspeakers can easily be detected when using a decent pair of crossoverless headphones! 😉

  8. I'm not sure who built the first 4 way speaker system. The problem of dividing the audible spectrum into segments is that loudspeaker drivers mostly have a usable range of about 2 1/2 to 3 octaves but they must cover 10 audible octaves as though they were a single non resonant device over that entire range. There are two competing mutually exclusive requirements for generating low frequencies and high frequencies. Low frequencies require moving a lot of air and to do that without any breakup of a moving membrane into harmonic modes requires something strong and heavy. It's going to have a fair amount of inertial mass. To reproduce high frequencies requires something with little weight or its inertial mass is too great. Additionally as frequency increases the propagation of sound becomes increasingly narrow. For a very small driver for a headphone it is possible to meet all of these requirements with a single driver. Maybe the best speaker is a million headphone drivers. But for loudspeaker systems there has to be this division of labor.

    The reason Infinity IRS could get away with being a 3 way system was that there were so many midrange drivers, EMIMs that each one only had to produce a small amount of sound at its lowest frequencies to match up with the woofer. Even so, multiway speakers have all kinds of other problems. For the less expensive IRS Beta, the economy on the number of drivers created the same problem everyone else faced, covering the entire audible range of 10 octaves with three drivers. The solution is the same one other manufacturers used, installing a fourth driver reducing the operating range of each one. Acoustic research struggled with this problem for decades building one variant of a 3 way system after another. Every one of them had a shortcoming one place or another. In AR3 it was at the top of the 12 inch woofer range at 1 khz. In AR2ax it was in the deepest bass. In AR3a it was at the bottom of the dome midrange frequencies. Improving the drivers by inventing ferrofluid to increase driver power capacity in AR 10 pi and AR11 or just adding more drivers as in AR LST helped but the problem was still there. Finally AR threw in the towel in 1975 adding a fourth 8" lower midrange and two side firing opposing 12" drivers as subwoofers closing the gap. I really don't know who was the first to do it.

    Designing a successful high fidelity sound systems requires much more than a speaker with a flat frequency response on axis. It's a far more complex problem as I suddenly understood it in 1974 and that entirely changed my view of the problem and the best methods to go about solving it.

    1. I always miss the second part of your revelation story, SM: why wasn’t it possible to convince at least 12 apostles audio engineers to follow your ground breaking concept - at least in the days of start-ups and crowd funding where they create such crazy things as a Maglev turntable.

      1. Paradigm change is always difficult, but in audio the dominant paradigm is entrenched by 80 years of ears adapted neurologically to audio, and bad science measuring this warped and stunted human hearing.

        This vicious circle also conformed to the Industrial Age paradigm of increasing convenience. I takes thousands of hours listening to real, live acoustic or electric music to learn what it really sound like, because you have to grow, wire and program the neurons to decode it. It is far easier to push a button and hear broadcast from the comfort of your favorite chair.

        Note that PA systems are just bigger, louder "Stereo" systems. Going to concerts where all the instruments and voices go through a speaker pair is still training ears to unreality.

        I have heard systems that capture more spatial information (Ambisonics) or convert recordings to fuller sound fields (Ambiophonics), but they are more finicky in playback equipment and setup, and don't work for most recordings.

        1. It's a lot like boating, where most boats are built to float on the surface of the water and a few are built to travel above or below the surface.

      2. People want to do the same thing they've always done and often if they haven't thought of a new idea themselves they reject it. Where venture capital is involved, those who put up the money want most of the profits.

        If you can even get someone like a large corporation interested in an idea, at first they will dismiss it, then they will investigate it on their own, then if it has any merit they will steal it, and then they will deny you had anything to do with it at all. If you sue them in court for patent infringement, even if you are right they can bankrupt you with legal fees in a court battle. In the 1990s I saw a product that infringed on my patent and consulted two patent attorneys. They'd have loved to have sued but they told me it would cost me over $100,000 in legal fees, take over ten years, and in the end I would lose anyway. I approached another patent attorney about 10 years ago with some novel upgrades and he told me some horror stories about patent infringement and intellectual property theft that sent me running for the hills. For one poor guy, by the time the US government offered him 20 million dollars for what they alone had stolen from him for his invention of the flat bed scanner he was already in an insane asylum.

        Theft of intellectual property is and always was far more common than you'd think. Rosalind Franklin discovered the structure of DNA, not Watson and Crick. Farnsworth invented electronic television, not Sarnoff. Bell infringed on other patents to get his telephone to work. Armstrong's invention of FM radio was rejected by the US patent office who said it wouldn't work. His widow had to fight just for recognition of his invention. Tesla was cheated by both Edison and Westinghouse. Tesla's inventions made the modern world as we know it possible but died broke. At one point he had to dig ditches for two years just to earn enough money to feed himself. Edison alone cheated Tesla out of $50,000 which was a lot of money at the time. Bill Gates did not invent Windows. Unless you are prepared to finance your own invention, don't even try. I think the guy who invented intermittent or variable speed windshield wipers eventually got some money out of GM. Intellectual property reflecting true innovation may become the most valuable commodity there is. Everything else can be bought for money.

        1. Why not convince some famous and rich artists by having them given a demo of your system? They might require a special recording according your ideas and could promote the special mix for the top end audiophile niche. Same concept was realized by Sony and Philips having convinced the famous conductor Karajan to promote the horrible sound quality of early CDs. Acuvox might have an explanation why Karajan could not hear the bad sound quality however listening daily to the most famous orchestras.

          1. First of all Paul, I completely lost interest in any possible commercial possibilities for my inventions a long time ago.

            My invention works with any recording, better with stereo recordings than mono. Not so well with phonograph records since a pop sounds like a canon shot.

            Adjustments are super critical. All recordings would have to be remastered and all systems would have to be custom designed and installed for each place it is to operate in so that they all sounded the same as each other. This is difficult because the room is an integral part of the system. It requires a more or less rectangular room. An L shaped living room dining room is its bane. So are large openings into other rooms since the system relies on reflections that have to be more or less left right symmetrical. It is not a very practical idea for commercialization. It would be a very expensive high risk capital expenditure to develop it. The prototype will likely remain a one of a kind system. Before I die, I intend to destroy it.

            1. " Before I die, I intend to destroy it."

              Seems extreme, afraid everyone will find out that the emperor had no clothes? Why not find someone doing research in the field, and leave it to them?

            1. Yet, the "conductor's sweet spot" can be replicated with certain records when using hefty rear speakers (the fronts can be smaller) with a quality digital time delay that is carefully set up. .

              I know this for a fact, because a conductor of the youth orchestra at Tanglewood heard my system (with time delay.) After listening, he surprised me. I was wanting him to hear a piece that he might consider performing. Yet, he commented on what he heard...'This sounds like what I hear from my podium." It can be done! But everything has to be set up precisely.

              1. Very interesting story, Gene. I stand enlightened!

                Of course, if one were to insist that to hear it like a conductor, one must be standing, and this would require some unusually high spikes for your speakers...and perhaps a very high ceiling. 😎

    2. There are a number of ways to increases linear driver bandwidth. One approach is reducing inductance, which is the limit to upper frequency and transient response. The most common methods are increasing flux with Neodymium magnets or a Faraday ring on the inner pole piece.

      Another is controlled decoupling of the motor from the diaphragm area. AFAIK this originated at Altec-Lansing with their classic curvilinear paper cone. Successful implementation of this technique shows a rising response of 3dB/octave from the smooth decrease in effective moving mass with frequency. This has the added benefit of decreasing the effective piston diameter with frequency, maintaining more consistent polar pattern.

      My favorite is the Air Motion Transformer of Dr. Oskar Heil. Squeezing the air instead of pushing it means higher efficiency and more volume for low frequencies and super low inductance (inherently cancelling!) for more high end and transient response.

      1. To get a 4" midwoofer to produce the same output at 30 hz as it does at 1 khz takes a 30 db boost, not an 18 db boost at 30 hz especially when the system resonant frequency has been deliberately pushed up to about 250 hz with an 8 db peak. This means that for the equivalent output from 1 watt at 1 khz, you need 1000 watts at 30 hz. You also need at least 30 to 40 of them to equal the SPL of two 12" subwoofers with a 60 watt amplifier. But no matter how much power you pour into it, you won't get much above 10 khz from a 4" midwoofer. It makes a terrible tweeter. Gordon Holt was right when he said they had too much inertial mass for high frequency drivers. These are among the reasons why the design failed. But even if you had a much stronger lighter material than paper or plastic, say graphene, you'd still have a problem with high frequency dispersion from a 4" driver.

        Both of the ideas you mentioned required multiway speaker systems. None of them were entirely successful. The Heil tweeter disappeared off the market for decades and only returned recently. You can buy them from Parts Express for about $350 each.

        One of the most interesting posts on this site for me was an explanation of the servo control amplifier for the Infinity IRS. I had always wondered about the relatively small size of the woofer tower for six 12" speakers. It turned out the open loop gain of the speaker/enclosure had a resonant frequency of 60 hz while AR1 achieved 42 hz and AR9 achieved 28 hz with a clever crossover trick. Those required 1.75 cubic feet per driver. But even the Tonegen 1259 which Ken Kantor helped develop based on the AR1W design required 3.3 cubic feet for one driver. These are all for acoustic suspension designs which IMO is by far the best design for a woofer/enclosure system. It took me 25 years to figure out how it really works and why it is best.

        1. Servo-woofers do not work near system resonances. The rapid phase shift will put them into oscillation. Notes that start and stop suddenly store energy at the resonant frequency in response to the modulation envelope transient. That radiates audibly anharmonic distortions, and servo control will go chaotic trying to cancel these out.

          1. If the loop gain of a servo system is one or greater when the phase shift of a servo system is 180 degrees , you have built an oscillator. This is called the criteria of stability. This is like driving an amusement park bumper car where steering response is slow. You want to make a left turn but you turn the wheel and nothing happens right away so you turn some more. Eventually when it does start to turn you have oversteered and you're not making a 90 degree turn, you're driving in a circle. This is also true of steering large ships.

            Most industrial control systems are servo driven using a PLC (programmable logic controller, a sort of industrial computer.) The speed of response is adjustable. When the 5G internet of things becomes ubiquitous I wonder how many systems will be driven to oscillation where there is no local control to override remote control.

      2. "My favorite is the Air Motion Transformer of Dr. Oskar Heil. Squeezing the air instead of pushing it means higher efficiency and more volume for low frequencies and super low inductance (inherently cancelling!) for more high end and transient response."

        Yes indeed, the AMT is very dynamic sounding although it's a ribbon, that's one of its strenghts!

  9. Paul, some years ago a friend and audio manufacturer (now retired) suggested to me the mid-point of acoustical energy from a symphony orchestra (obviously on average, not at a specific point in time due to the variations in the music) was something like 270 Hz. From that a decent system should be able to reproduce as much acoustical energy below that frequency as above it to present balanced sound. We don't typically think of music reproduction in terms of acoustical energy, but that concept seems to support Arnie's VMC perspective.

  10. I always thought it more important to remove the crossover from the high end (intelligibility) range rather than the midbass. Hence the Community M4 and Plasmatronics speakers.

  11. While the recognition of the midbass vs. imaging struggle is worthwhile to articulate, the VMC doesn't strike me as a major breakthrough any more than using an equalizer (digital or otherwise), or, more similarly, a separate level control on a driver in a speaker. The active amplification of the midbass driver may be unusual, but it's just an extension of the idea of a user-adjustable amplified and integrated subwoofer. In fact, this sounds like an argument for fully active speakers - but then we wouldn't see those beautiful PS amps! If cool initials are needed to make this concept sound innovative, instead of VMC, how about MVIDAAOSA (Midbass Vs. Imaging Dilemma And Amplitude Optimization Solution Adjustment)?

    (Sorry to always sound contrary in the comments - I enjoy the Posts!)

    1. What I guess is when comparing this separate amp‘ed driver with an EQ it is even more extreme than comparing the mid performance of a 2-way speaker with a 3-way. Because the midbass coupler is also separately amp‘ed and because bass control is even more sensible to being transduced by just one chassis with one amp or two chassis with two amps. For me the midbass coupler is only one of 7 of my favourite concepts implemented in this speaker.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram