Speaker phase confusion

August 27, 2021
 by Paul McGowan

11 comments on “Speaker phase confusion”

  1. It’s probably a second order(12 dB per octave) crossover where the mid range is wired electrically out of phase because second order rotates 180 degrees. First order and the very popular 4th order Linkwitz/Riley wire all 3 drivers electrically identically.

    1. Good evening Hahax!
      I didn’t know this about my Avantone Pro CLA10 studio monitors, until I both paid very close attention, and conducted a technical test on them.
      Before I tell you all what kind of a test I conducted on them, I will say this first to all teenagers that mite be reading this post.
      Please don’t try this at home.
      Not until you have enough experience to do this kind of testing with speakers.
      Get it wrong, you’ll end up regretting it.
      The kind of test that I conducted, is not for the faint of heart, by any means.
      I checked out my quod of Avantone Pro CLA10 studio monitors.
      I was wondering for a good while, how come I couldn’t hear the mid bass if there was any to be had by these monitors.
      First I worked the balance control knob to the left, then to the right.
      When each one was playing by itself, each one, gave me the mid bass.
      But just as soon as I turned that balance knob back to the middle, the mid bass went away.
      And so, I grabbed a 9 volt battery, and connected each pare in series.
      While one cone in one woofer moved itself outwards, the cone in the other woofer moved itself inwards.
      And this was when I fed each one 4.5 volts.
      This is the hard way of finding out that your speakers are out of faze.
      And so, I then took the monitors apart, and slitely changed the way they were wired up.
      And I done this just by switching the wires around on each driver.And I reconducted the test again.
      This time, the cones in both woofers moved themselves outwards.
      An after that, I reconnected them to my amplifier, and put on some Jessica Carson.
      Long story short, I heard the mid bass coming from both pares of monitors.
      Go figure this one!

      1. Makes sense. Both woofers should be in phase, move in and out at the same time. But the question was different. It referred to the midrange electrically out of phase with the woofer(and tweeter). The answer is that because of phase shift in the crossover(probably second order roll off) the woofer and mid range fell out of acoustic phase(the only phase relationship that matters to hearing). The fix was to put the woofer and mid out of electric phase which put them back into acoustic phase.

        1. Good morning Hahax!
          I get that.
          But truth be told, the technical relationship, is exactly the same.
          Because if you really stop to think about it, there is a such thing, as channel oriented speakers.
          Some amplifiers will have their terminals configured in this way.
          They will look like this, +–+.
          Or you mite see amplifiers that have their terminals configured this way.
          +- +-.
          If you’re not yoost to working with speakers and amps that are sat up like this, then they can confuse the living hell out of you.
          But like you said, reversing the polarity of the mid ranges, will fix the sonic problem.t

  2. This is true for steady state signals produced by e.g. guitar. In case of dynamic pulses the single chassis are not in phase, that’s the reason why broadband speakers reproduce drums more natural with better attack. Except 6 dB crossovers. The Holy Grail is to find the best compromise.

    1. Bud Fried, who I knew well, would agree with you about 1st order crossovers. Then again, although I know the peril of single samples, among the very best drum solo reproduction I’ve ever heard is my system and the second cut of ‘The Pentangle'(the 1st non classical album reviewed in Stereophile) and my system is a custom 3 way using 4th order Linkwitz/Riley crossovers. With my eyes closed I can almost imagine the movement of the drum skins.

  3. It’s important to distinguish between “out of phase” and “reversed polarity”. When the mid-range unit is connected as the original post describes, the mid-range signal is inverted. In other words, when the woofer and tweeter are moving outwards (compressing the air in the room) the mid is moving inwards (decompressing the air in the room). If we fed an asymmetric waveform to the mid, we would see that waveform inverted as an acoustic signal, not the original signal 180 degrees out of phase.

    As others have noted, the rationale for the reversed polarity mid is to provide a flat amplitude response in the transition between woofer and mid, and between mid and tweeter. All fairly standard stuff as far as I’m aware. The phase response is going to be all over the shop.

    What this does highlight is that most speaker designers are concentrating on providing flat amplitude response across the frequency range, and entirely prepared to sacrifice phase response to achieve this. The fact that many well-regarded loudspeakers use high-order crossover filters should encourage us to think that human hearing is not very sensitive to phase.

  4. Hi Mark,
    fully agreed with you. On top of that phase response is different even inside the chassis itself. By example using two chassis with almost same data ( except sensitivity ), one professional musician bass and the other high fidelity, both will sound very different. While the first one produces fast and punchy sound the other sounds more pleasant with less coloration and better deep bass. Similar result for AMT and calotte tweeters. A huge problem if your taste ranges from classic to metal. With most “high end” speakers Metallica sounds sleepy in contrast to studio monitors. That’s why my current speakers utilize two chassis, one hifi for deep bass and one musician for mid bass, the rest of is covered by output transformerless ESL. All coupled active by 24 dB crossover. IMHO a very good solution to enclose both worlds.

  5. Isn’t the discussion of phase-shift by the crossover most misleading? Shouldn’t it be all about time-shift? The mandatory requirement for a correct timing of the drivers is that in the crossover region the mid-bass driver shows the some movement as the tweeter in order to get fast transients – most characteristic for music, especially plucked strings and percussion – correctly reproduced. This would mean that tweeter and mid-bass should have identical time-shifts. Reducing the problem to phase-shifts assumes that there are no sharp transients and music would be represented by pure soft infinite sine-waves of a most limited frequency range. For getting a near perfect frequency response just flip the phase and the frequency response looks more acceptable – but timing is still incorrect! Flat frequency response doesn’t mean that the step response is correct!

    1. I don’t think that the discussion of phase shift is misleading although, as you point out, it might not be the whole of the story.

      The main objective for most loudspeaker designers is to create a flat amplitude response on axis. As we know, one of the challenges in meeting this objective is keeping response flat at the transition frequencies where adjacent drivers are radiating together. In the original post, the system alternates the polarity of drivers so that the superposition of woofer and mid, or mid and tweeter, is approximately flat. The designer will have taken into account the natural roll off of the drivers, the phase and amplitude response of the crossover filters, and the geometry of the drivers. Presumably the designer is prepared to sacrifice the phase response to achieve the goal of flat amplitude response. As you say, the step response is going to look very strange.

      It would be a very much harder design objective to create a system with flat frequency response on axis, together with linear phase response. That almost certainly involves making compromises elsewhere. What we have to decide is the extent to which phase distortion is audible, and if the compromises involved in controlling phase distortion provide a net benefit.

      1. Indeed! Finally it is all about compromises. But what is the best compromise? I remember that Wadia was following the goal of best timing in contrast to the majority of DAC-designers who simply went for a frequency response flat up to 20 kHz. And remember the discussion about pre-ringing depending on the type of digital filters. I am pretty sure that time-alignment is most crucial – not only for the design concept of Wilson Audio loudspeakers. 🙂

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