The midbass dilemma

October 19, 2018
 by Paul McGowan

Midbass is that all-important frequency range between 200Hz and 500Hz and covers the most important instruments in our musical libraries: voice, cello, viola, brass, tympani, woodwinds, bass, guitar. Just about everything we treasure has some element of midbass. In fact, it is the foundation of almost all music.

Yet, despite its essential presence loudspeaker designers have traditionally treated it as just another frequency, counseling that for best performance we have only the position our left and right speaker cabinets to affect its correctness: closer together and midbass increases, farther apart it thins out.

Infinity founder Arnie Nudell was obsessed with midbass rightness. In his view, there was little else of greater importance. "Get the midbass right and everything else falls into place".

Infinity and Genesis loudspeakers of his design had no more chance to get the midbass correct than any other speaker for the first 50 years of his design prowess impact on our industry. And then, a flash of insight. Why not create a 4-way loudspeaker that offers users the ability to adjust the all-important midbass with the turn of a knob?

And thus, the adjustable midbass coupler was born, one of the quietest revolutions in audio's long history.

His invention covers the narrow range of 100Hz to 300Hz with a specially designed dynamic driver blessed with its own high wattage power amplifier and volume control. And, it works. This innovative approach to musical perfection permits the user an unheard of luxury: the ability to place the speaker pair where they image best, then tune the midbass to achieve sonic bliss.

This feature will, of course, be in our upcoming line of speakers to launch in late 2019. But, in the meantime, it's good to noodle on the benefits of midbass.

Music finds it critical.

Subscribe to Paul's Posts

18 comments on “The midbass dilemma”

  1. A look at any of these sorts of charts show how the mid-bass (say 100-500Hz) is key to most of the orchestra, but vitally it covers most of the male and about half of the female normal vocal range.
    https://external-preview.redd.it/6Kdm6YblxU-NNAjEumwIRXJa8pE4BZzU5IL4RVYb9Hs.jpg?width=960&crop=smart&auto=webp&s=1e7d3a19e5ace87a68ec51ba16005bfb6c3599d9
    It is therefore no surprise that I've heard both Pete Thomas (PMC) and Alan Shaw (Harbeth) say that the basic test of a speaker is spoken vocals and you can tell within seconds whether it is basically a good or bad speaker. You don't need hours of listening.
    I would only buy speakers from a company that designs its own drivers because that is what it really is all about. The speaker cone or panel is what pushes the air move back and forth and makes the sound pressure wave and if the speaker manufacturer hasn't considered that from the very basics then they don't get my vote.
    My speakers use a single proprietary mid/bass driver that according to Stereophile is -5dB at 40Hz and has a crossover at just over 3khz, and by many accounts has the best mid and mid-bass out there, which is why they are so popular. They certainly have limitations, especially in the bass department, and if I was willing to spend 5 times the price I'd buy a pair of Focals (not least as they come in gloss white). Given the extraordinary level of driver design input into over decades, I do not like the idea of letting the customer tune the mid-bass. The speaker should have got that right in the first place. That's good design and they don't suffer from being close to back walls.
    I like this video that explains some of the basic good and bad engineering of a mid-bass cone.
    https://www.harbeth.co.uk/library/movies_hug/HV1014-02Cones_Part2L.mov

    1. Mid bass balance is also the subject to room placement, architecture and interior design. Because of the floor-bounce effect, it varies with height of the driver, distance from the listener, floor composition and covering, wall proximity, furniture and the ceiling height.

      I just heard an eleven channel demo system set up by the factory team from PMC, playing back eleven channel recordings made by variations of three different microphone arrays in reverberant environments. The rear pair where pulled out from the wall and the overhead quad were hanging in space as the ceiling was about 20'. There were gross timbral differences between the LCR and side speakers, which were a few inches off the walls, and the less well supported speakers, and this interfered with the desired effect of envelopment.

      PMC and Genelec are normally installed in rooms designed around the monitors. I find they have too many phase and off-axis issues. I prefer better time domain response like Dunlavy, Earthworks and Grimm and the only monitor that addresses off-axis frequency and phase well is the Kii III.

      I asked many of the presenters at the AES Convention if they tested on musicians. Most of them looked at me like I was crazy. The only affirmative answers were a Lexicon engineer and a PhD candidate who was doing his dissertation on guitar amp modeling with neural networks, who also is a guitar player.

      Can anyone tell me why they think they - OR THEIR FAVORITE DESIGNER - know what music sounds like better than the people who play music for hours every day?

      Can any so-called objectivist tell me why their machines know what music sounds like better than the human practitioners?

  2. I fully support the idea of (not only) having the midbass adjustable. At least equally important to me is having bass adjustable (as you also planned for the new speakers).
    But why the heck not making the speaker fully active then?

    To me the only reason would be for those optionally using pure tube amps for mids/highs. Everything else could have been built into the speaker as well and with great add. benefits.

    But this would break with a basic rule of high end audio: „we need more components (to sell)“ 😉 Just very few break it.

    1. I forgot to say:
      it’s not only the vendors, also most customers probably like to fiddle around with more components, even if an integration (most have probably never experienced before) would offer more benefits.

    2. I agree and know that low bass needs to be adjustable due to many variables in recording levels and switching from analog to digital and therefore so does mid bass...maybe to a lesser degree. I also must add that adjustable bass is essential especially at higher volumes when close to critical mass of room capacity.
      I suspect that PSA will eventually develop a totally active speaker...why wouldn't they. But it sure won't have a BHK amplifier activating it.
      For me at this time of my kit development the AN2 is going to be a great fit.

      1. The question is: does it need a BHK powerhouse for mid/treble activation?

        My experience is, that even amplification of 1/5th to 1/10th the effort of a passive one still gives more benefits when activating a speaker. The demands/benefits of active and passive speakers in amplification seem different in their effect.

  3. Discussing a single parameter as being most crucial for a serious loudspeaker design is not serious at all because this restricted focus neglects the fact that there are many more parameters to be precisely adjusted for an authentic sound. Concerning mid bass you have to address at least the fact that unfortunately there are still huge room mode effects in this frequecy range. Thus the speaker design cannot be discussed without taking into account the listening room and the relative position of each driver relatively to the boundaries and to the other drivers. Just imagine two big woofer membranes placed on the lower end of the baffle creating different floor reflections and mixing in the crossover region with the mid bass wave front emitted from the middle of the baffle. What a smeared sound wave.

    1. I must agree entirely and also with jazznut. I have a reasonably large room by UK standards, but do not consider a full range speaker a sensible option. My 8" mid-bass drivers are fine enough down to 40Hz. My system allows me to apply a high pass filter, should I choose to, but fortunately there is no need. The lower bass is provided by a subwoofer that is separately located and tuned to the room and the main speakers.
      Were I to get something like the Focal Utopia V2, which is a very modest sized full range speaker with sublime midrange (and looks to kill) and easy to locate, you still lack the ability to control the bass other than by using separate DSP components.
      I think Mr Nudell and now Paul assume that room issues are irrelevant, which may likely be the case at their price-points. However, for most people that is not the case and is why most speaker manufacturers make a range of speakers focused on a different room sizes as much as different budgets.
      I also agree with jazznut that in the long term the market is likely to move towards smaller active units with built-in DSP, driven mainly by the economics of our over-populated planet.

      1. But Steven, PSAudio with offering a DSP bass unit, an adjustable midbass coupler and mids/treble as „room acoustics friendly“ AMT drivers, they are providing the most dedication to room issues (except fully active speakers, which also have mids and treble adjustable).

        So my point was: while this is great, why not fully active?

        But I know the answer 😉 : people want to make use of their existing amps and PSAudio also have to sell theirs.

  4. Hello Jazznut. You probably have a valid point. That being said then why not a multichannel external dsp where people could use their existing amp(s) add more as necessary or upgrade to fully active as an integral part of the speaker system. As an interim all the bases are covered that way?

  5. In a previous post I mentioned that the first designer (as far as I remember) in using a mid-bass coupler as such, was A. Nudell in his QLS, which consists of a 100 mm Phillips driver that works in the range of frequencies between 200 and 600 Hz, already as far back as 1976.

    This small driver couples the extraordinary Watkins: two vc driver, with the line source of six 1 1/2" dome midranges.The diameter of the presumed piston was chosen to give it greater agility and to be able to fit in a better way with the fastest midranges. This makes sense.

    For some subscribers to this forum, they may not be interested in the technical specifications, but for the rest of them, it is important to remind them that these are spectacular:
    + - 2 dB from 18 Hz to 32,000 Hz, the upper end, through 48 "line source high frequency radiator (EMIT) reproducing signals from 4 through 32 KHz. An interesting parameter is the horizontal dispersion, impressive for the time to be + - 2 dB from 18 Hz to 20 KHz in 180 ° frontal hemisphere.

    The mid-bass coupler was not used in subsequent models like 4.5, RS1 etc., for me, inexplicably, hence the pre-eminence of the QLS.

    Although the idea of ​​using it was novel, it is not until now that it has been retaken by some designers, only that today it is used with a much larger Sd that allows reproducing in a more convicting way, instruments such as the counter-bassoon and others.

    The current DSP era allows more accurate use of drivers within the range in which the response is flatter, as well as making the necessary adjustments that contemplate the correction of the room, etc. (REW 5.1)

    With the help of a DSP, I have been able to use a mid-bass coupler consisting of three 8" drivers that cover the range from 62 Hz to 400 Hz, leaving true sub woofers (4th way) the lowest octaves of the instruments. It is idle to say that passive Xovers are not used, this is the big jump.

    1. "horizontal dispersion, impressive for the time to be + – 2 dB from 18 Hz to 20 KHz in 180 ° frontal hemisphere"

      This is not merely impressive for the time, it is impossible with a rectangular baffle. This spec is "smoothing", averaging pink noise over 1/3 octave bands or less resolution. This speaker also must have awful vertical dispersion, which enters your ears through floor and ceiling bounce.

      The only real machine test is a gated sine wave sweep in an anechoic chamber. To get polar response this way takes days, and produces results far less useful for marketing.

      1. I was just transcribing the Infinity promotional catalog of the time.

        It means that if this specification is impossible, it is already strewn entering the field or ignorance, or lack of integrity of the manufacturer, by means of which potential buyers are misled.

        Presumably, they have somehow had access to some laboratory test to make such an assertion.

        Regrettably, this is an example of how the high-end audio industry is managed, in that potential customers have no way of verifying that the assertions have any technical backing.

  6. Which audible frequency ranges are the most important? IMO all of them. Which inaudible frequencies are the most important? IMO none of them. In fact you're better off without them because the only thing they can do is cause problems especially in the audible range. At the lowest subsonic frequencies they waste amplifier power, overdrive woofers and subwoofers, and at the ultrasonic range they can cause audible distortion and burn out tweeters and even amplifiers.

    One of the beneficial uses a graphic equalizer can provide is to instruct you how changes in the relative loudness in different frequency ranges can affect tonal balance. Too much or too little in each range has a different subjective effect. Even if you don't use one as part of your system, you could learn to remember what these changes do to sound and make judgments about equipment and recordings if you have a good memory of what live music sounds like.

    Covering the full audible spectrum with only three ranges of drivers has always been a problem. With ten octaves and loudspeaker drivers being inherently resonant devices their range is usually limited to 2.5 to 3 octaves each. That's not enough to cover 10 full octaves. No matter what you do somethins gotta give. In AR's attempt to build full range 3 way systems their first major effort was AR3. The problem there was at the top of the woofer range which had a ragged response near its 1Khz crossover frequency. AR3A moved the problem down to the region of 525 khz where the woofer crossed over to the improved midrange. A super AR3a was the AR LST also marketed as the Cello Amati. It used 4 tweeters to increase treble dynamic range and 4 midrange domes to increase power handling capacity of the midrange domes at their low end of 525 hz. Further improvements were made when their next effort introduced their invention of ferrofluid cooling. Still not good enough they turned the system into a 4 way system With Teledyne AR9 installing two 12 inch side firing woofers crossed over to an 8" driver used as a woofer in AR4 series which handled frequencies from 200 hz to 1.5 khz. An upper midrange dome with an aluminum ring used as a semi horn above 4 khz crossed over to a 3/4" dome tweeter at 7 khz. The semihorn midrange and the recess of the dome slightly plus the addition of an acoustic blanket to reduce reflection off the baffle reduced dispersion giving in to the market demand for speakers that focused their sound to a greater degree than their prior speakers. So now they had a full range system that easily covered the entire audible range. The free front drivers have their own level controls each marked 0 db, -3 db, and -6 db. So the subwoofers and lower midrange look pretty much like the photographs of Arnie Nudell's speaker. The semi line array midrange an tweeter drivers are of course entirely different from the single midrange and tweeter domes in AR9.

    IRS Beta required a "mid bass coupler" because unlike IRS it did not have sufficient power handling capacity from one EMIM to cover the upper bass. When you see a two way system add a subwoofer it is a 3 way system. and when a 3 way system adds a subwoofer it becomes a 4 way system. Re-engineered original Bose 901 shows you can build a full range two way system but you have to jump through hoops to do it. I don't recommending trying it. BTW, the aggregate piston area of original and series II Bose 901 speaker is the equivalent of a single 14" acoustic suspension woofer but its maximum excursion is only a fraction of what you'd get out of a well designed 12 " acoustic suspension woofer so you'd need multiple units to achieve the same output and because of the system resonance frequency design above 180 hz you'd need a bucketload of amplifier power at very low frequencies, say 3 or 4 systems and at least 600 to 1000 watts at 2 ohms. You'll also need an additional boost of about 12 db at 30 hz in addition to the 18db the Bose equalizer provides. For your efforts you will get a lot of sound at 30 hz with about 10% THD which isn't bad at very low frequencies.

    So if you want to hear deep bass but also want to hear the sound of tubas, trombones, double basses, cellos, pipe organs, pianos, and trombones, not to mention drums, you really do need a 4 way system. AR's clever design combines the dual firing sided fired woofers with the 8" lower midrange that couples the deep bass well to the room and to the lower midrange without suckout. It incorporates principles developed by Roy Alison who originally worked for AR. Here is a piece written for Cello. If you have a friend who is a music student who wants to get a scholarship to a college as a music major, if he or she can learn to play this piece transposed for viola exactly the way it is played here the odds are very good of getting one. One of my sister's students did exactly that and started a bidding war for her. Violists are in high demand and there aren't many around. You can get sheet music on Virtual Sheet Music's website and they have a software package that allows you to transpose music into any key and for any instrument. So the winning school was Syracuse University and the scholarship was for $165,000 in a special program. There are two versions of Mischa Maisky playing this piece but this is the one you want to emulate. The other one isn't nearly as good. Every subtle inflection and nuance is critical so skipping steps to simplify learning it will not get you what you want.

    https://www.youtube.com/watch?v=mGQLXRTl3Z0

    BTW, the whole set of six cello suites is available as a DVD on Deutsche Gramaphone.

  7. Sadly, I have to admit I don't know what most instruments actually sound like when played by themselves acoustically in a room.

    I do know that when I have had a jazz combo play acoustically at a party in my home using our baby grand piano with saxophone, drums, etc. that the dynamic range and overall volume is not what I could listen to normally in the house.

    And I also know whether I like the sound of recorded music when reproduced in my listening room at levels I can tolerate and can only compare that to changes made to my system itself.

  8. Midbass is where most of the energy is concentrated in time averaged measurements. It takes substantial volume displacement to reproduce peak levels, and therefore substantial surface area to reduce anharmonic distortion. Remember that even though Doppler distortion is low at normal average levels, spectral contamination is audible to .01%. Experimentally I have found that 2x8" or 10" midrange speakers are necessary to balance live acoustic music with acceptably low distortion, and I use 12" speakers for mid-bass couplers from 100Hz-400Hz.

    Larger diaphragms increase efficiency too, which reduces distortion from power compression and increases effective amplifier headroom.

Leave a Reply

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

4865 Sterling Dr.
Boulder, CO 80301
1-800-PSAUDIO

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