Besides the wiring phase there is also phase shift from driver coil not on same plane. Then wire reverse can correct. When determining needed phase, only do one speaker. When you close your eyes and listen. You want all drives to sound as one. Out of phase with each other accusticly will sound like separate speakers. Swap the midrange wires and listen again you will hear the difference. You always want the bass cone to move forward on positive signals. Use a 'D' cell battery for cone direction test
Totally agree - would love to see more Chris Brunhaver, and Darren Myers as well. The lunch talk interview style format generally worked really well for me.
Very very nice explained , simple and should be easy to understand for all people . I think the problem with this is that most people can not wrap around in their head what is really going on when polarity is flipped . - Example - 3 ways system , bass , mid and tweeter with mid flipped polarity - - - Most misunderstandings is that they think - - '''' When the mid is flipped then it play opposite '''' When the bass and tweeter '' push out , then they think the mid is still pull in '' . This is because they doesn't get what the phase means . When the phase is reversed that mean the electrical polarity is reversed so what was the '' plus side '' now become the '' minus side '' Sound waves is made of waves going from plus to minus voltage and the time in between two cycles ( a plus and minus ) determines the frequency . ( the wave lenght in between two ) That is why two equally signals but 180 degree out of phase wll cancel each other out . There is a voltage cycle at 360 degree in a full cycle and at 270 degree it is fully reversed so what was max positive voltage is now max negative voltage . ( fully reversed when there is 180 degree in between them ) That mean a signal at 10 volt max positive at 90 degree and 10 volt max negative 270 degree will result in 0 voltage . If the two signals is 180 degree out of phase so the max positive and max negative is at exact same time . ( Thus in between the 90 degree and 270 degree is 180 degree difference ) The cycle start at 0 degree at 0 voltage and rice up to max voltage at 90 degree then it decrease to 0 at 180 degree and further decrease into negative voltage to max negative at 270 degree and rise up to 0 voltage at 360 degree . ( also zerro degree . 0 degree is start of cycle and 360 is end and start of new so also 0 ) This is for a normal symetrical sine wave there is waves with other characteristics . It is the same with Air preasure from moving speakers . When two equally speakers is placed front to front and phase is flipped 180 degree on one of them they will cancel each other out . When one push the other will pull and result is '' nothing '' . This is the principle used in noise cancelation systems ( eg headsets or in cars ect ) You take the noise and flip the phase and send it back out to calcel it out . You have to explain it like the red and black terminals on the speaker '' plus , minus '' so they can get the physical picture of what is going on . Even when say ' It is to get the phase right at listning position , they wont get what it means and still think '' the mid is kind of reversed '' But the mid is NOT reversed at all . ( Even if it look reversed ) When the phase is shifted then the electrical plus point of the sine wave is now become the minus side and vise versa . So to wire it up correctly with respect to '' plus and minus '' on the mid speaker terminals , you then have to reverse them . This simply mean that in the end all 3 speakers , Bass , mid and tweeter they Push and pull at the same time as if one speaker . They are all in phase as we say and the acoustical push / pull of the air in the room is also in phase ( They move the air in same direction at same time ) Another typically misunderstandings is that many people claim , '' when the speaker is reversed then you loose the tight transient's because the sound doesn't exactely start at once blah blah '' Yes believe or not , there is still many there claim that . It is simply NOT true at all , the phase ( voltage ) is reversed in the cross over filter and '' timing '' is still the same . so in the case of a 3 way speaker with mid reversed then the electrical '' polarity '' is just reversed and the mid still play at same timing . This is simply just a myth . There are other factors in play when it comes to transient response . One imortant factor is the acclerationof the speaker itself . A heavier cone is a bit slower to start and stop than a lighter one . Also the ' spider's ' construction tribute to how fast it can acclerate and the accleration is what determine how effective it is to replicate transients with good or bad response . The place ment of the speakers themself in the cabinet can also tribute to how well the speakers is ' lined up ' in phase with each other . A good rule of thumb is - When you look at the speaker cabinet from the side , then the speakers should NOT be lined up on a front baffle in one line . The mid and tweeter should be moved back so the voice coils are in line with each other . This to make shure that the wave lenghts ' line up ' so all the frequencies reach you at the same time . This will increase the ' transient's ' so it sound more ' fast ' . If they arrive a little out of phase ( Yes this is phase again ) then it will sound ' more muddy ' and '' clarity '' be less accurate . ( so to use HiFi words ) That is why you see speakers where the mid and tweeter is lined up longer to the back than the bass when look at the front of the speaker cabinet . Simply to get the wave lenght's lined up in phase . Reflections and the room itself is also a major player , tributing to change the acoustical phase . When a wave is hitting a wall and reflected back it get delayed and the delay itself when arrive at listning position can be out of phase too . A good example is a bass wave when it hit a back wall and reflect ( bounce back ) at certain frequencies depending of the room size , some frequencies will hit each other and cancel out when arrive and phase is going through negative . But other frequencies will hit when top of wave meet each other and ' amplify ' each other . If top of one half of wave with same polarity meet at same time the result is now a double size wave and this is what caused STANDING waves in the room . So you end up with some bass that simply is gone and you lack bass in certain frequencies while in others you end with way too much , resulting in this typically '' Boomy '' bass sound . On the other hand the higher frequencies they do the same and the result end up with ' harsh ' sound and too much and / or too little in those frequencies . But the higher frequencies also produce ' echo / delay ' too ... You can combat this by using bass absorbers in the corners and acoustic damping of the walls ect . But you can NEVER remove it 100 percent . You can only ' tame ' the problem so it be less ' visible to your ears ' A perfect room doesn't exist but you can get close enough that it will sound near perfect and really good ( even on a very cheap budget ) Yes there is many factors in play that can change the phase of the acoustical signal ( the air moved ) Being an Audiophile / musician can be a really challenge when it comes to get '' The Perfect Sound '' . Thanks making this video Paul . Short but very informative and easy to understand explanation . Im shure many people will understand it . Edit - Sorry for my typos , it is very late now here in the philippines and im sitting with my coffe .
Paul, great info as usual. A bit of clarification, if I may. 2nd and 4th order crossovers invert phase generally, as you say, to smooth dips at the crossover point. They do in fact rotate phase. 1st order crossovers are the only one's that do this at a constant phase of 90 degrees. Only a first-order crossover has a CONSTANT PHASE DIFFERENCE at every frequency above, at, and below the crossover point. That means there is Zero change in phase between them, so the signals out of the high- and low-pass sections are "in Phase", always, on every frequency. And all of that means, finally, that first-order circuits are the only ones that pass all their signals through with the SAME time delay, so all of those emerge in the same time-order in which they entered. This is called time-coherent behavior. So you get the original transient, one not smeared out in time. A speaker that is designed to deliver a time-coherent output is automatically "phase coherent". The converse is not true, as you may know: A phase-coherent speaker is not necessarily a time-coherent speaker. In fact, if you see the advertisement claim phase coherency, you can bet that the speaker is not time coherent. To say it differently, the combined output of the two drivers is always unity at zero phase, even though the two vectors are always 90 degrees apart. This is difficult to conceptualize, but the math behind it is relatively simple. The higher order the crossover, the more phase shift one incurs. A 4th order unit can be PHASE coherent, but not time coherent. As we know, music is not recorded this way and I would argue should not be played back this way. Thanks for the informed information on RU-vid.
Good explanation. That happened to me with a 2-way I designed, using a 1st order filter (plus Zobel) for the woofer and a 2nd order for the tweeter. Assembled, my frequency test showed a big dip at the crossover point. So I reversed the polarity on the tweeter and the result was a beautiful, flat response curve. That was a learning point for me.
It's almost like your painting a nyquist diagram in the air. Seeing someone explaining phases, filters and phase shifts like that makes you smile 😀 Great video.👍
This answer by Paul is pseudo-correct in one way, yet so entirely incorrect at the same time. An inverted phase midrange driver *never* delivers in-phase waveform at the listener’s ears at the listening position over its main passband. Only the blend at crossover regions is improved. The remainder of the primary passband of the midrange driver is inverted in phase, including at the listening position. This is one of the most shamefully overlooked issues regarding loudspeaker performance in the world of “high-end” audio. An inverted phase midrange driver can by definition never deliver an accurate reproduction of the original waveform of the recording to the listener’s ears.
Yes you are correct, and yes it's sad that it's not ever talked about more. I also believe that many of the people that love single driver systems may be particularly sensitive to this phaseyness through the midrange. Transient information simply never lines up amongst the multiple drivers.
@@Paulmcgowanpsaudio Hey Paul. Question: I use a 4th order linkweitz Riley crossover on a "four way"(40hz to 80 hz; 80 hz to 500 hz; 500 hz to 2.2 khz and 2.2khz out) I found with the 4th order, instead of 180 degrees, it works best by taking the 500hz to 2.2khz out of phase by 90 degrees. Do you think that has something to do with the difference between the 1st order roll off, high and low, vs. the 4th order roll off, high and low???? Thanks and have a good one!! Carl
@Douglas Blake Hey, thanks for the reply!! Actually, I'm a live sound engineer who can only afford consumer grade speakers. I needed to reach out further with my behringer 2 way powered speakers, to about 300 feet with good frequency response and even sound pressure levels. So, I got the idea from audiophile sites like Paul's that if I reduced how many frequencies the speaker and hence the built in amplifier had to deal with, I could get more power and hence more range to each of the speakers. B115d for 80 to 500hz, b112d for 500 to 2.2khz and b115d for 2.2khz (built in crossover) on out. I use three DBX venu 360's for the phase and crossover's(total of six speakers 3 left and 3 right with separate Subs). I adjusted the phase to the mid range, since it wasn't summing up close within five feet of the speakers. As you can see the midrange is covered by a 12 inch speaker. I was just curious about the effect a steeper slope on the crossover points had on phasing. BTW, it works great!! Have a good one. Carl
@Douglas Blake Thanks, point taken and understood. With the way I have them set up, they reach out clearly to 300 feet. With just one on the left and one on the right, I was only getting 150 feet maximum(outdoors). I have run them with the 3 way setup several times already including a four day event with 16 bands performing in 110 degree heat, in the sun, during the day(not uphill both ways in the snow LOL). Two of the six went into automatic shutdown on the third day (I have spares), but I ran them afterwards, and they sounded good. I think it's safe to say that these inexpensive Behringers are built pretty tough(one DJ actually dropped them down a stairwell, and they worked fine for years afterwards). But, I will keep an eye on them for degradation. This is why I like this site and the people on it. I think a lot of sound engineers know how to do something, but don't take the time to learn why they are doing it. Thanks for your help!! Carl
Right. But doing this way, the Kick response is crap. (Subject first Transient recognition / processing) Drums, dynamic played Guitar, Piano, all comes blurred due to this. Real time coherency is the key for naturalness!
It is not that you want all drivers to be in phase, as only a few cross-overs have that aim (such as Linkwitz-Riley), and you have to try to achieve that even as the drivers have their own phase. It is that you want a flat magnitude! And then there is a natural way to wire the drivers as parts of the transfer function may have negative sign in its numerator.
Thought I'd add that the phase and polarity stuff (outer box or inner crossovers) for a single speaker box doesn't matter until a stereo setup is needed. Then placement, polarity/phase becomes important.
It could also be that the midrange has an odd order crossover and the woofer and tweeter use even order crossover (or vice versa). Just as each stage of an amplifier inverts phase, so does each order of a crossover.
Yay!!!!!!!! Finally someone else is talking about this!!! Please help me!! Lol I am just a girl that knows good sound when it comes to car audio and let me tell you I am about ready to give up because I have yet to find someone that knows how to set the correct crossovers ... I have always been into audio and was wondering if you had any info on where I could start some kind of formal training??
Really depends on what you want to do with that training. If you want to work in the industry I might suggest considering a career as an electrical engineer or an acoustician.
@@stonefree1911 I think i am only mostly right, beacause of filters for full range speakers. Some of them are also reffered as "crossovers" in literature.
No laughing matter, voice coils heat up with more power which causes a shift in the impedance and that in return changes the filter characteristics and thus the sound characteristics of your speaker.
Absolutely right regarding wires connections. Anyway I had incidents where due to specific behavior of amplitudes no proper connecting sounded more proper. Tricky to decide what to do better. Of course we experience also much smaller phase problems - as it's known every resonance of enclosure or speaker produces adequate weaving of phases Midrange speaker actually works with (with proportional to frequency in it's band) change of phase. Characteristic called "waterfall" tells about many of them but complexity of view makes it hard to use. Tweeter must therefore work in opposite to low toner. there is no fading but sure it is not copy of original Every driver has it's geometry affecting phases It is a lot of mess from which can save us only apparent good luck and a lot of own experience . Old producers who judged outcome by ear possessed it. Unfortunately their job wasn't much appreciated because "their" exceptional drivers were then used almost always in bad final constructions. which suited only popular demand. Some of them looking average so so left me speechless. It could be a dream to make identical today as they could be solution for perfect sound without phase confuse.
Paul stresses that the goal is acoustic summation at the crossover point, but I'd argue that it's electrical too. A 2nd order LP on the woofer will usually rotate the phase by around 90 deg at the -3dB point (depending greatly on the filter alignment). A 2nd order HP on the tweeter will ALSO rotate the phase 90 deg in the other direction. That puts the 2 drivers 180 deg out of phase at the crossover point... both acoustically AND electrically. Forget about 3-ways.... even on 2-ways you will often find the tweeter wired in reverse.... It also depends on the physical time-alignment of the drivers, the baffle angle, and the intended listening angle. There's a LOT of variables, even for just a 2-way speaker.
Paul could you make a video and explain, what are 2 way, 3 way speakers and the pros and cons of each, and what music they are more suited for, and what is the difference in sound from FM radio classic rock station for example and that same music thru your turntable, what are the nuances. Cause I just set up a classic pioneer receiver in my room, and I heard pink floyd meddle album and I hear alot of stuff on it compared to the same coming out of the FM station. The Fm station music sounds kick arse, but the turntable music wasn't as in your face per say, but it seemed like i am there in the studio. Please ignore if this don't make any sense. I am retired and just set up the old classic sound system that we had mid 70's, to enjoy myself and bring back a special place in time where things were not crazy like they are now, and for kinda therapy. - Rey Lopez in California Bay Area.
I have a pair of B&O S 60. They have 4 drivers in them so they are phase coherent , the main drivers are coupled " right" +, -, it is equipped with a so called filler midrange driver. To fill in the gap in between the rotation of the crossover.
My Realistic Optimus 650 speakers had the mids wired out of phase, it uses 1st order X overs on mid and tweeter, none on woofer. The mid uses full bandpass, unlike just a cap on cheaper house brand speakers. I flipped the phase on the Realistic mids and they sound better and also removed the 10 ohm resistor for the fact the mids are already inefficient compared the the Foster and white van mids out there
I thought they wired it that way to prevent some frequency suck out.... Yet, the mid and woofer will be operating in reversed polarity, as well as the tweeter. It will not give a phase coherent sound. A first order crossover in a two way speaker will keep both drivers in phase. I have heard several first order crossover 2 way speakers. There is a cohesiveness preserved that way. Wiring the tweeter and mid/woofer out of phase may provide a flat frequency response.... But their will be something there that will sound out of focus. For not all frequencies will be moving forward on an impulse of the music, as it occurs in real life.
I know that filtering shifts phase. The midrange driver being the most filtered, it makes sense to flip its phase. Thanks Paul for making ideas like this easier to understand
The signals in the cables themselves propagate at a substantial fraction of the speed of light. This is only an issue at RF, where the frequency is so high that the signal cannot "see" the other end of the cable. Instead it just sees some characteristic impedance, regardless of what's on the other end, even if its a dead short. At audio, what goes in is what comes out as the frequencies are very low, and you would need literally miles of speaker cable for the group delay to become an issue.
In this case its done solely for acoustics, and to help control driver interactions. This is much the same as you might do with inverting the phase of a sub to make it sound better. The cancellations can help tone down peaking where the main speakers and subs are playing at the same time. Edit: There is also the issue of cancellations of the drivers in the vertical direction, and this is a substantial issue with 2-way and 3-way speakers. Outside of the listening window, there can be large gaps in the response where there is substantial cancellation. In this case they will choose the phase, among other things, to help optimize the on and off-axis response.
@@Lamizeharblu In general, other than the skin effect (which is mainly only a problem for power distribution) audio behaves identically to DC. If you feel comfortable conveying DC through a wire at the currents your speakers run at, then the audio will do fine. Only real worry is with the input side as the signals are high impedance, so you need either shielded cables, or balanced XLR to keep AM and mains noise out.
My IRS IIIa's Emits are out of phase with the rest of the speaker. When I built my Emit satellites I designed them off another Infinity speaker so they are 2 in and 2 out per box. I love the way they sound so it must work.
Sir, My name is Pallab from Kolkata INDIA.My question is about Damping factor of amplifier. How damping factor affects Low Mid and High frequency of audio?? Can very high damping factor of amplifier affects Mid and high frequency?
It effects every driver, but only at resonance. Usually only the (sub)woofer is used around resonance, so, in practice, only the woofer is effected. One of the big problems with electromagnetic damping is that it is dependent on the voice coil impedance, and that is heavily modulated by temperature, current and excursion. So, electromagnetic damping is imprecise and unstable.
@Douglas Blake The point is that you don’t need electromagnetic damping, there are better ways of damping, and you don’t want it, because electromagnetic damping is imprecise and unstable and will ruin the tuning of the loudspeakers and the frequency response. You’re better off without. I have here an amplifier with an output impedance of 1400 ohms: no power compression, stable damping, stable crossovers!!, no current distortion, no microphone effect and no intermodulation distortion by a modulating inductance… So, the damping factor is critical and it must be as low as possible. At least 1/100 or preferably smaller.
@Douglas Blake Actually I use 4 ohm speakers to lower it even further. So the damping factor is less than 0.003. Long ago, with low powered tube amplifiers they want a DF around 1 (optimal power transfer). With solid state amplifiers with high feedback and more power available, and the hype around DF, they wanted a DF of 100 or more (optimal voltage transfer). And if you’re trying to avoid electromagnetic damping like me, you want a DF of 1/100 or less (as I stated already earlier) and non-conducting voice-coil formers (to eliminate also the eddy current electromagnetic damping), for optimal current transfer, and elimination of all kind of impedance issues. A loudspeaker driver is after all a current driven device.
interesting topic, but a quick google search made things worse. and i need information since i am building a speaker crossover at the moment. you have fir filters that solves phase issues, an so many way to make a crossover, there is also active crossovers vs the normal passive that is normally talked about. at least as far as i know switching polarity on a driver lets you have 2 equal and opposite phases with no drawbacks.
@Douglas Blake i ditched the idea for making a passive crossover, now after decades of waiting there are digital options, so now i have a chance of designing a complete speaker, i can now just enter the frequency and chose slope and type, now i have a basic crossover :)
@Douglas Blake deducting stereo signal i need 4 channels each side, 1 for the tweeter, 1 for each midrange as its a MTM design, no need to wire them in parallel and i avoid low ohm and i can try different way to cross things over, like a 2.5 way speaker if 2 way dont work out well. then 1 channel for the subwoofer. all i needed was two hypex 3way plate amps and a stereo power amp(connected thru a dac) for subs. 8 channels total. all controlled by my minidsp shd studio.
@Douglas Blake i was originally intended it do be done with no plate amps, but for my first speaker with crossover it seems like a easier choice to go with plate amp. no room to mount plate amp anyways so that was not a selling point. i like to know how you would approach powering and dsp my system, think i got the right parts and it will work together. this way i can make more speakers and keep the amp setup, make all kinds of speakers. to clarify the chain here, pc as source, into minidsp shd studio which does volume and room correction and sub and main speaker crossover integration. the minidsp digital outs the dacs, the hypex fusionamp 123 does its thing with amps and dacs inside, and i put the 2 way crossover in there, with the appropriate delays and gain for each driver. two stereo subwoofers with external dac and NAD c 298 power amp. it sure is an experiment, figuring out the correct phase for every driver is the most difficult i think.
@Douglas Blake i just finished building the speakers, two main MTM with a tweeter and two 3" fullrange woofers. three binding posts in the back, then i built two subwoofers, sealed with 10" driver. 15hz was the low end extension goal, and it does it well in my small room, 10hz flat in measurement but not in real life. electronically i would think that usb from the pc is the best option, and i use balanced analog xlr from dac to poweramp, and my plan was balanced digital xlr (AES/EBU) into the 3 way plate amp. coaxial digital rca to dac from my mnidisp control unit. the last piece i am missing is the plate amps, some delivery time. the rest i got here. i dont know how a analog preamp would fit in there, i need a digital signal with volume adjustment into the plate amp, otherwise the plate amp needs to convert the analog input into digital for dsp and crossover then convert it back to analog again for the power section. i also need a dsp unit further up the chain for room correction, i have been using my pc for that before, but its inferius to the minidsp all digital solution, i also dont have a seperate channel for the subwoofers if i use my pc. i have all class d amplification, and i like icepower the best for low heat output, especially at low volume and idle. i have icepower module poweramp and purifi poweramp right now, the plate amp that is coming is hypex ncore. audio outputs from my PC i am ok with if its digital and bypass the analog section, but then i am building a new pc next year and getting new version of windows with more features related to audio, lets see what that does for me then.
Yes vandersteen do that. all drivers go out of phase in the operation frequent especially in the lower frequency, you should do that if you have subwoofers as well And don't forget there's 90 degrees and 180 degrees phase dissipation
