@@QoraxAudio That word actually has many meanings, including usage in British slang, which is where I know it from. It can mean one who punts a boat, one who bets against a bookkeeper or gambles, one who is a prostitute, or who watches pornography regularly, one who is the customer of a business, and one who marks points in the game of Basset. The usage I intended as the context I had heard it used (and which may or may not be entirely correct), is that of the 'arm-chair know-it-all', also sometimes referred to as "the peanut gallery" (which, in turn, has its own origin and meaning, coming from the vaudeville era): People who hold forth on a topic or subject they know little about as if they were the World's foremost authority, often relentlessly heckling the actual authority on the subject. To be clear, I meant it in a half-joking, light-hearted way. There are actually some very knowledgeable people who frequent here. : ) And that, my dear sir, is your etymology and vocabulary lesson for today. 😆
Good start out of the gate. The difficulty I see in presenting this topic are the problems with trying to present highly technical information to a mostly non-technical audience. At least one assumes that. Speakers push air around. How well they do that, by following the original signal with the same intensity (volume) as well as accuracy is the rub. Frequency dividers, crossovers are inherently problematic. They introduce delays in the signals, and especially so with each other. As an example (wish I had a white board or overhead), the signal path to the bass speaker (woofer), generally has a large coil of wire tightly wound. Its through an increased magnetic field when a signal applied, and collapsing magnetic when the signal is reduced, creates sn opposition to rapidly changing signals. It therefore makes a good electrical device to block ever higher frequencies. It may very well pass 100 changes per second (Hertz or older reference CPS), but increases resistance to change, say at perhaps 2000 Hz. The two way design employs a high frequency speaker. By way of it's physical design, it's tinier and generally rigid radiating surface, responds well to more rapid signal changes. Since it can't physically handle lower frequencies, another electrical component is used to block or attenuate lower frequencies, while passing higher frequencies, all the way up to our hearing limits. At my age those limits come sooner. I'm no longer able to hear 18,000 to 20,000 Hz. Still like most people who have been listening through decent equipment for decades, can usually tell a nice sounding tweeter from one that can't keep up. This most basic description of connection to a simple crossover, is only adequate to illustrate another problem with time alignment. This is referred to as a 1st order crossover. More complicated crossovers, which use additional inductors to pass low signals, and capacitors to pass high signals in a combination, create further intrusions into the phase alignment of a reproduction of sound. The electrical & magnetic properties of an inductor cause the changing voltage to lead current flow. [Quick overview: voltage is the applied pressure in the completed circuit, whereas current is the actual flow of the electrical energy.] Capacitors on the other hand, allow the current to pass before the voltage catches up. Again an over simplification. When the woofer receives a changing signal, it's movement will be slightly delayed by the arrival of current. The tweeter on the other hand receives the current ahead of the voltage. If you're following up to this point, you can at least tell, the diaphragm of each speaker will move at slightly different rates because of the necessary frequency dividing electrical components of capacitors and inductors. So no matter tge actual physical distance to your ears, the electrical signals will be controlling how soon each speaker begins and ends movement. Now that I said all that, in order to make sense of it all, it will be drawing time. Thank you for your patience if you read all that.
Both the title and description of this video are a bit misleading. Speaker designs by Jim Thief, Richard Vandersteen, John Dunlavy, or Earthworks' Sigma 6.2 use passive 6dB crossovers (AFAIK) and could validly be said to be demanding on the drivers, but that caveat aside, they are phase and time correct and therefore do function as their manufacturers claim. John Bau's Spicas used 6dB tweeter networks, but 24dB Bessel networks on the woofers. Further, the publications of Samuel Harsch, Steen Duelund, or Kido Yamanaka make it clear that all 6dB networks are not a requirement for time-aligned speakers - multiple mathematical solutions can be made to work. It is also possible to design a crossover network with multiple slopes, so that the initial rolloff is at 6dB but transitions to 12dB so that the drivers are better protected (and the bandwidth over which the drivers overlap is reduced). An example of this design philosophy is Jeff Bagby's series-crossover Tributes (available as a quite reasonably priced kit from Meniscus Audio, and Jeff's technical writeup can be downloaded gratis from Meniscus' website). Yes, it is true that more than a few speaker designs have sloped or stepped baffles but aren't actually time aligned. But not all speakers are like this. A less misleading title for this video could have been "Why many speakers that appear to be time aligned, aren't".
Chris, you should call it: "Why time alignment doesn't work with me?" You mentioned Dunlavy Vandersteen and Thiel - it worked worked them all right. It takes some effort to make a proper time aligned speaker with the 1st order crossovers.
Hi contradicts himself several times..time alignment, is difficult to get right, but when employed properly, there’s no going back to non time-aligned monitors
@@KCNYC I own time and phase aligned speakers with first order cross over (Lumen White Lumyere) and to me it's one of the things that make otherwise great speakers sound very authentic and natural (granted, this is a difficult task and comes with its challenges since the drivers are taxed more, drivers need to be aligned perfectly etc). There are many aspects to a great sounding speaker but speaker brands which heavily focus on getting psycho accoustics right and see it as a fundamental part of their design (to which time and phase alignment belongs as one part of the whole equation) just seem to sound very natural compared to brands which leave that more aside. Good examples would be Lumen White, Kaiser Accoustics and Marten speakers. Whether you personally like them or not, they definitely belong to very natural sounding ones and it has to do with giving psycho accoustics a high priority in speaker design, e.g. resonance patterns, acoustically more neutral sounding drivers etc
I was fortunate enough to meet Chris when I visited ps audio last fall. He was a gentleman, and enjoyed sitting in listening room 2 and discuss speakers and the company. Thank you Chris!
Hey, Chris! It’s nice to finally meet you, sir. That was a top notch description and pretty easily understandable explanation, which is much appreciated. Loved this vid, and I’d love to see more of you talking HiFi, but obviously and especially speakers more in future vids. Thanks so much, Chris! 🎶🔊😁👍
Hey, it is nice to see you and this was number one video on PS Audio. It is great to hear you pass on your knowledge to everyone that has interest in high end audio. It gets better the more do to it. thank you!
Great to meet ya Chris. Welcome to the world of critics! Instead of being a critic today, (I'm a recovering one) I'll throw some advice. Engineers like to use physical examples to explain and show how one thing effects another like USING YOUR OWN SPEAKERS YOU HAVE BEEN DESIGNING!! WHERE ARE YOUR SPEAKERS! (Whew.....) Kinda still in recovery. Thanks again and see ya soon.
It has been my consistent experience with analog crossovers that filter slope and Q, and path length alignment are important. At least near the crossover point, where both drivers are contributing significantly to the output, it is desirable for them to be in phase. At the very least, if they are not, then frequency response aberrations which are often not easy to correct will occur. I find that perceived clarity and coherence improve when this condition is met as well. Of course, this requires that the path lengths of the drivers to the listener’s ears and the crossover filters are optimized as a complete system. With few exceptions, I agree that first-order filters require too many compromises in other areas. An exception that comes to mind right away is the Quad ESL.
As best as I could tell, what that Mr. Brunhaver is saying is that even if physical alignment of drivers is used to achieve phase coherency in the overlap region of two drivers that interface, this won't eliminate the overall rotation in phase that occurs in a speaker from low frequency to high frequency. So what? Does he mean to imply that since the overall rotation in phase cannot be avoided through physical alignment of the drivers, that there isn't any benefit to making the two drivers phase coherent in the overlap? The impression I'm left with is that he doesn't understand the reason that with most speakers the tweeter is vertically aligned with the rear of the woofer cone. This is done for reasons that have to do with the proper summation of the two individual wavefronts within the overlap frequencies, such that the main lobe, for any frequency within the overlap, will extend along a horizontal plane, rather than tilt up or down. If the lobe tilts up or down, then the tonality of the speaker will be different at different distances from the speaker. In order to avoid tilting of the lobe, it is necessary for the two individual wavefronts to be phase-coherent, not merely at the exact crossover frequency, but throughout the overlap region. Vertical alignment of the two drivers is one aspect of achieving this goal. It might not be essential to do this in order to achieve this goal, but when it is done, it is done toward achieving this goal, and it makes perfect sense, irrespective of the fact that there will still be an overall phase rotation from low to high frequency.
I really enjoyd the depth of knowledge Chris share in this video. Really great work. Looking forward to more videos from Chris and the rest o the team!
Dirac sounds great, was really impressed when I had a chance to try it. I really do think most people in a casual listening setting aren't highly sensitive to phase, but when it comes to active mixing/engineering it definitely makes a difference. Interesting to hear your perspectives on crossovers and I'm glad you touched on DSP as well.
The vowels and consonants from the voice sound are coming from a foot apart with normal 2 way speakers but with coaxials and FR drivers the vowels and consonants are coming from the same point. I can hear the difference and you get a way better sound image as well.
Interesting point of view Mr.Brunhaver. I listened to the Eikon Image 1 DSP speakers at the 2019 Capital Audio Fest with Gayle Sanders sitting next to me answering all of my questions. This was of course before the horrific COVID. I was truly impressed with what I heard and if I had $25K laying about for a second system I would be tempted to purchase.
Glad to hear from and see Chris in this video series. Kind hard to say no to the boss. :-) but Paul generally knows what he's doing... 99% of the time any way.
There are compromises with every facet of speaker design, and you pick and choose the compromises that matter to you. For me, time and phase alignment make absolute sense. Having drivers purposely firing at different times, make no sense to me.
Interesting subject. I too have recently started using DSP as I have added a sub to my set up to complement the Magnepans. So I'm using the DSP for cross over and time alignment between the speakers and the sub. I just thought it was worth mentioning something that you touched on regarding the DSP approach adding additional stages of A to D and D to A conversion. This isn't true if you are using an all digital, pre output signal path like I am. So I run the DSP software on the same computer that I stream from. I just wanted to get that out there incase your comment regarding conversion misled some people in regards to using DSP. Obviously I totally appreciate that you would be referring to DSP in the context of speaker design so this would be handled inside the speaker after the analogue input and would therefore be subject to the additional conversion which you speak of. Otherwise, great video and I look forward to hearing more on the subject of speaker design.
Great presentation Chris. As with most products where one design characteristic is touted as the big game changer, the most accurate response is “Actually, it’s a bit more complicated than that”.
I have 3 BBE balanced pro-audio processing in between my Outlaw pre-pro and Yamaha amplifiers in a 5.2 surround system. It sounds more open and 3 dimensional compared to them switched out of the signal path. I love it, never getting rid of em.
This is a great topic to address. The concern involves only those frequencies that are shared by two drivers (the crossover region), and most importantly, the mid driver and the tweeter (not so much the bass and mid driver). Important to understand that the length of waves at @ 2-3 kHz (the common crossover region) is between 4.5 and 7 inches in length. So if the tweeter was out of phase by @90 degrees or @270 degrees (a 180 degree out of phase condition is easily managed by reversing the polarity of the tweeter), then the driver position would have to be shifted by 1-2 inches, creating a step and can create as much difficulty as the out of alignment condition.
This is kind of old but still. Using Audiolense for time alignment is audible. My experience, and others, is that it improves imaging especially the centering of the vocals. Correcting (EQ) frequency response is much more audible. Compare the same FR correction with or w/o time alignment is definitely audible. This is done using active design i.e. separate amplification for each driver and digital xover using FIR filters.
Dirac does work well. It was smart of MiniDSP to put the Dirac button on the middle of the remote so user's can A/B test it instantly. It clearly works. I imagine in PS Audio will ultimately produce a system that has a mic calibrated streamer/crossover/fir filter/preamp that feeds multiple amps that are perfectly matched to each driver. At least I hope they do. The difficulty is that it would require a lot of wires and stacks. So the preamp would need to be wireless and the amps mounted in the speakers.
Hi Chris, my name is Hans, and I was told that you have answers to some questions that I have. I am thrilled with the opportunity to communicate with you!
I purchased a FreeDSP-Aurora, and will be tri or quad amping. I'm in the process of re-capping, so don't have everything in place. The Aurora allows fairly long FIR filters to be programmed. So much to do, so little time!
Fantastic information Chris. I really enjoyed listening to you explain time alignment. Looking forward to your videos in the future. Nice to finally see you in person. Paul has been telling us great things about you. I cant wait to buy a set of speakers from you folks!
I've always been amazed at the difference it makes in clarity and depth when you reverse the phase on a tweeter. It doesnt always work with every build, but when it does, that tiny mod is a life saver when you're trying to get a speaker to sound just right. It's useful when you are running the tweeter on a 12db slope and your mid or woofer on a 6db slope.
I think it was an answer to my question regarding Cabasse Galion V speakers. Thanks therefore. For me, as a non native english person, it was heavy to understand, but I Look forward to the upcomming videos. Thx
Hey Chris, that was great. I've just built a new tall 3-way with my old 20+ years old drivers (by intention!) to see what I can achieve with them with a 8ch MiniDSP processor and a UMIK-1. So far the results are mind blowing, I think I'll keep these drivers in these boxes as is although they represent absolutely 0 value by now. I'm just playing around to see the difference between a heavily corrected system and a slightly corrected one and I think here one has to find her/his own taste, which kind of overall sound and behavior to choose. Now I seem to love a moderate bass correction (always the most problematic part due to room modes) and just minimal midrange/tweeter adjustment, leaving some naturally occuring smaller anomalies in the upper range.. not much, just a bit. That combined with the somewhat more corrected bass provides a great blend of the "old world" and the DSP'd world and the result was also for my wife and stepdaughter like "wooow". Great to see DSP not being the devil of audiophile societies but an option for further improvement - if applied well, with care (!).
You need to read anything written by Dr. Richard Heyser, Dr. Floyd Toole, Dr. Christien Heil or Thompson Ambrose and then learn how to use a TEF or Easera. Correct phase at crossover point is paramount.
You can measure the phase with holmimpulse. At crossovers invert polarities on speakers to align the phase curves. You can verify this on CSD with ARTA : Red colours all the way. At the end you move back the tweeters and mediums to have the best 0° phase. Holmimpulse and ARTA are free.
I feel so lucky to have purchased my Medowlark Kestrels a true time coherent speaker. Yes it makes a difference. Just watched another video about the other big debate - do cables especially speaker cables make a difference. What is was looking at is the higher end cables seem to "time coherent" than the cheaper cables. It isn't about the frequency responses or anything else. In the end after many many auditioning times with many many speakers the Meadowlarks just fit perfectly with me. That is the key in all this. You find what fits you and moves you. Been over 20 years with my speakers and have never wanted for anything else. I feel sorry for those audiophiles that are in constant change searching for that "perfect"" sound. Time to stop listening for the problems and enjoying the beauty of the music.
Good info Chris.. Here are some things many need to know,yet evidently,do not.First of all,we have a more diverse set of listeners than ever before.WE have people like me,who listen to music on UA-cam,on my aligned system,on my tablet,from records,from turntables,and on anything from MP3/MiniDisc to a Richard Long system.I have heard expensive gear,and boy oh boy,some of it was amazing,in 'proper' listening rooms.The problem is,I never really have time and interest to sit in a specific spot,in an acoustically treated room,and spend my whole listening experience there.It's just not practical,and I personally do not know anyone who does. Now we have way more Class D amps,more digital listening/recording devices,and a host of converters,both ways,to suit all prices and preferences.I recall going to someone's place in Harlem,and in his studio,he has Tannoys,and yes,I was impressed.I have been equally impressed by systems with EAWs,Mcintosh,PSBs,while others have come to hear me spin records on Alesis Monitor 2s and Kenwoods,with a Parasound power amp,etc,etc.Like others here have said,there is no magic bullet,no magic speaker,and what is revered by few,can be hated by many.This is yet another subject like which cables,what tonearm,what plinth,what type of wall treatment,and the FACT is,rarely,if all ,can you ever tell the differences blindly. I do have an old time delay box from way back,and it works for me,and I found it in a thrift shop for $20.I poked around,and tested over and over,until I heard real improvements,yet then again,I love to mix and match products anyway.Point being,today,people want it loud,quick,easy,and uncomplicated,and everything does not have to be scientific,which works for us all,because diversity brought us here,to the Internet,with gaming PCs,Mac Minis,Iphones,Notes,Wacom tablets,VR headsets,and advanced AI.Those who get stuck on anything,get left behind..
You ask 10 different ‘experts’ the same question, you’ll likely get 10 opposing opinions. They can’t all be right. I think Chris was making every effort to be true to his beliefs!
The recorded event is captured by the mics. Loudspeaker drivers reproduce the recording and launch it into our listening room. The goal is to recreate the recorded event for enjoyment. Putting it all back together as it's launched and propagates into our room is no trivial task... even before the room enters the equation. What we the listener experiences while enjoying our systems is the direct energy combined with a great deal of reflected room energy. That said, the image recreation benefits from the reflected energy being phase accurate uniformly with the corresponding energy from the other channel (ie., Left to Right, or Right to Left). So, overcoming the challenges of phase shift from coils and caps, physical discordant drive units, etc, present significant obstacles to imaging and soundstaging. However physical and acoustic ASYMMETRY within the listening room can shred the image and soundstage apart.
Yes the Vandersteen speakers are quite awesome and they do have a special time alignment thing going on with the sound. Personally I don't care what has and what has not, I just want the damn thing to sound good, and I agree with and without it can sound pretty good.
My speakers has the front of the cabinet sloped where the tweeter is 2 inches behind the woofer. My tweeter will sound great & 3 dimensional, but n0t the woofer. When I move the woofer toward me 2 inches its 3 dimensional sounding , not the tweeter. I can tilt the speaker forward, where the tweeter is even with the woofer, but this makes the woofer point down
Try with inside boxes just like I recommend in my videos, it isolates all the unit sounds of your speaker, bringing the sound quality up to an entirely different level. Also having you speakers relatively close to the back and sidewalls will maybe also help.
I have a pair of Theil tweeters left over from when I worked at Audio Gallery and they are sweet tweeters as long as you use them in their range. They sounded awful with any kind of lows entering them. We had the 3.6's and the 5's.
Fat Rat Pretty much everything he says I agree with. Most simple designs don’t require a whole lot of time alignment built into the crossover but even with some simple 2 ways it can be an important part to the final tweak of the speaker. Always exceptions.
Fat Rat Indeed. Many of the designer type love to chat at shows and the like. I’ve chatted with Andrew Jones before, such a nice guy. Sandy Gross I got to talk to at Cedia last year. He sold Golden Ear tho just lately.
I got my Dahlquist DQ-10's . . . staggered baffles for tweeter to mid-range . . . I just re-capped the complex crossovers to handle five drivers per channel . . . time and phase aligned. Jon Dahlquist and Saul Marantz. Alignment to the woofer is less critical than hearing range 150Hz and up.
Excellent video. Just so you know, Finite impulse response technology definitely does work. KS Audio, a German company, perfected the technology probably 20 years ago. Other companies are doing it now. I looked for your links below but did not see them.
Highly informative! You should mention active speakers. These designs can better control both crossover and phase. Nevertheless your point is correct - we don’t need absolute phase accuracy - “smooth” is really what is important and sharp changes in phase do cause audible issues. MQA and minimum phase filters mess up phase and I find this to be audible with careful listening if the phase delays occur at audible frequencies (high frequencies are delayed). My preference is linear phase. Most speakers have phase issues in the bass but this is much less audible then phase delays at mid range or higher frequencies. The manifestation of phase delays from minimum phase filters is usually less accurate imaging. Some minimum phase filters like MQA also cause mild compression (apodizing type filters). Active speakers aren’t perfect but if well designed they possess several advantages over passive designs.
Good one Chris! Excellent debut, really enjoyed the critical point of view and it'll be fun trolling around here in the out of phase waves you've caused.
I have switched between minimum phase and linear phase crossovers digitally, and there certainly is a massive difference that I heard anyway. Linear sounded so much natural, but I do think you need the speaker itself to be perfectly time aligned first!
Thank you for the video. I've never thought about something this difficult as I'm a very casual audiophile (if such thing even can exist), I just enjoy listening to my music and trying to get the best performance out of the hardware at hand. But now when I hear "time aligned" I will know what to ask about it.
Great to see others from PS Audio share their knowledge, hope we see more of the technical / engineering team have guess slots on Pauls UA-cam show. Hope we see more of your insights soon
Thank you Chris for sharing your knowledge of speakers. I suggested paul put you into some videos since he has mentioned you in previous videos. Hope to see more! Feel free in future videos maybe talking about your background and how all you designed for and perhaps speakers you had a hand in designing. Thank for the video! Perhaps scott can get in more videos as well. Haven't seen many from him. Enjoyed the few he did. Thanks!
OMG can’t wait for the new book from Paul’s series: “0% truth: how I moved from getting skinnier to actually becoming younger!” ROTFL Just joking, excellent explanation. Well done Chris!
Well then there is surround systems and car audio in which as far as I can see there is no way to get all sound waves to your ear correctly, or am I a bit dumb?
I also discovered (as you mentioned) some other variables that effect phase irrespective of passive crossover components and driver position, and I postulated that these variables are related to the great difference between a mid driver and a tweeter in the mass of the two driver elements, and the physical resistance to motion exerted by the size of the cone anchoring material (just a surround for the tweeter, and both a surround and spider for the mid driver).
Ya and I would think in a live setting the high frequency sound would be traveling faster then the Lows. All so very confusing, my Harbeth’s sing so I’m happy. 😎
Some people be going out buying time aligned speakers with no acoustic treatment in their room. Unless it's in a live setting with multiple speakers in the same range in an array I'd say time alignment between tweeter and woofer matter a hell of a lot less than phase issues from first reflection points
The offset alignment of drivers often allows the crossover to sum properly at the listening position, so time alignment in a way can be misused (or assumed) and is not the same as phase coherent. I can never understand why audio enthusiasts when given information instinctively run the wrong way with it. :) Music is based on time --- phase, time, and Q always important. :)
In terms of geting the phase right, comb filtering probably has more the most noticeably effect, since it will result in multiple peaks and dips in the ampitude response of the loudspeaker
And this is why you should just listen for yourself and buy what sounds best to you. There are a lot of theories on speaker design but a few are commonly offered as the holy grail and time alignment is one of them. It's completely misunderstood by most audiophiles and likely, as Chris points out, by most designers. Acoustic centers are not determined by where the voice coils of a speaker lie so very often the stepped baffle or raked speakers are not actually time aligned regardless of the crossover design. Crossover design, however, does play a critical role and there are plenty of ways to design a crossover to approximate time and phase alignment. Even so this will not guarantee good sound. There is no singular part of a design that if you get it right then, shazaam, audio nirvana materializes. And that's why you can design speakers many different ways and arrive at a great end result. Or, conversely, you can execute one design theory to perfection and still end up with something that sounds like a first time DIY'er built it.
That's not entirely correct: absolute phase is what's difficult to hear. Relative phase (that all drivers should be moving in-phase) isn't. Also, what's being time-aligned, if done correctly, are the so-called acoustic centers of the drivers, which isn't necessarily their membrane surface as claimed in this video, but where energy conversion is taking place in each one of the drivers (usually closer to the centre axis of the voice coil, but really, no educated guess will replace an exact measurement). I've designed loudspeakers of the kind discussed here, and can vouch for their ability to produce significantly more realistic sound, but yes, they're a major PITA to design and build, and it's also true one needs wide-band drivers to begin with. I'm convinced time- and phase-coherent speakers are worth the effort. And no, DSP-based systems, while easy to implement, sound unnatural by comparison - those just a short-cut for the lazy if you ask me. A properly designed system will always be superior to one using makeshift solutions to problems one wouldn't have in the first place if said system were designed from scratch.
I would add to things. 1) If there is a vertical offset between the drivers, and one is closer to ear level than the other, then this will force a difference in their relative positions. 2) Nice name. Leon Fleisher died too young.
@@jonathanknight8251 Referring to the legendary pianist and teacher born in San Francisco in 1928, he's still alive (at least I'm hoping so?). And yes, what you're referring to vertical offset is the same I meant when I say all acoustic centers of multiple drivers must ideally be placed at an equal distance from the listening position. It can't really be done without measuring equipment, but it's worth the effort, and so are filters that get the drivers to move in phase.
I can't listen to music on conventional multi-way speakers. I hear only non-logical sounds. Phase and time alignment are the most critical thing about a speaker. That's why I'm listening to an open baffle with linear phase crossover and the necessary delays to different drivers. The only one better speaker than that is full-range driver. :)
What I get from your description is that time alignment works and from my reading of professional speaker builders Wilson,B&W,PSB,Magico,YG,I have read some of there reasons for there builds learning of sound and how it transmitted and time alignment is very necessary,so all I can surmise is that I feel sorry for Paul hiring you
Time alignment is important with multi channel (3+) systems, 2 channel systems you can have a dipole with a crude out of phase delay channel bouncing off the front wall so who cares. "Folds" in phase hide details (information) just like peaks in linear response. Don't forget amplifiers often have significant phase rotation. When the system is slopped together phase doesn't matter (ergo some of the academic research on the audibility of phase never tested the source or the amplifiers for phase distortion). As the audio system's obvious flaws are continually diminished smaller flaws suddenly become "a problem". Passive loudspeakers are obsolete because of what you described, but the inherent flaws of 2 speaker playback are greater than the flaw of phase distortion. So you have to correct listening to only 2 speakers first. :)
Aka time alignment doesn't actually "align" the sound waves, but it sounds different, and therefore we think that the difference is the sound of aligned sound waves?
Time aligning drivers (physically aligning them) can help with phase summation in the crossover. But if I had to choose, I'd rather have the highs and mids hitting my ear a smidge before the bass. Your brain hearing the first consonant or sibilance sound of a word, before the bass arrives, gives the impression of clarity. I often like to place the subwoofer in front of the speakers but then switch the phase 180 degrees. It sounds quite clear for vocals, plus the sub hits you harder and doesn't interact with the back wall as much.
To the point of the video, they are physically aligned (because of the tweeter set back in the waveguide) but are not using a time/phase coherent crossover (not that they claim to) That was sort of the point of my video. Again, I think that that theirs is the correct engineering decision to make but they are non minimum phase etc.
@@cbrunhaver Hi Chris, I understand your point and it's very informative. I was just happy to see my speakers in the PS Audio office, it's a good look (both aesthetically and for Mads at Buchardt :) Thanks for the video!
supergiddy3 yes, I believe that our companies are rather friendly with each other and I have found his engineering work and designs to be very well implemented.
The only time alignment I've had contact with was room correction set up that the amp did you positioned a special mic at the listening position and let the amp do its thing and you could really hear the difference but isn't that really a form of DSP ? (Sorry it had nothing to do with the speakers themselves)
Howdy Chris.... you've met your Aussie match here btw ..! I'm an electro acoustic engineer and a recording engineer at Sony ... I go along with what you say here but the trick with speakers is to select the right ones that go well together in the first place .. after that it's a matter of phase and amplitude control using XO networks ... first orders are excellent for lower frequencies yes but for HF drivers 2nd or 3rd order networks are essential . I have software that measures both group delay and plots decay responses shown in the familiar waterfall plot . I use these extensively to achieve the right result . Do you at PS Audio use component valuation software in your XO designs ? This software works like pspice or ORCAD and changes the values of XO components to fit a particular response envelope with minimum group delay .. I guess you'll be handling all things speaker from now on as Paul does refer to you as the speaker guru ..!
I used to work at Sony for a few years and helped support Kaku and Yuki with the SS-AR1 and SS-AR2 product. Yes, I have a couple of crossover optimizer/simulation programs that we use. We also design all of our own drivers and work heavily to optimize those for the application as well.
cbrunhaver ~ how interesting.. you and I could chat and argue for hours on speaker designs I'm sure lol 😂 No two engineers always agree ..! Yes these component simulators only optimise as you said .. the final test is in open space using LMS software as I do to perfect the correct coherent phase and amplitude response . I keep an open mind about making your own brew of in house drivers .... my first reaction is why reinvent the wheel ? Big names in electrodynamic transducers ( Dynaudio comes to mind ..) have done nothing else BUT design high end and very predictable drivers with amazing consistency in characteristics. I know ... a very large number of high end speaker manufacturers seem to build their own drivers but in reality they contract out the designs . Whether this is good I'm still not really sure .. new designs require extensive testing and honing in order to narrow down any 'production spread '. I'd like your comments on any readily available software for T/L enclosure designs ... as these things do not use any lumped constants and this is what makes them so hard to design ... Cheers for now ...
@@janinapalmer8368 Well, I'm always glad to chat! My email is chris b (at) psaudio.com. LMS doesn't actually measure acoustic phase (it can derive it through a hilbert transform but that assumes a minimum phase system). A shame about Chris' death, as I was looking forward to his new system and was a huge fan of his work. Well, the likes of Focal and Dynaudio don't sell their drivers anymore (except to car audio), as it competes with their box speaker sales. JBL came to much the same approach in the pro world. For us, it's a value add - when buying an off the shelf driver, you're paying for their engineering staff and supply chain etc. Some guys have incredible robotic assembly lines and smart purchasing and are hard to beat, as compared to a lot of high-end driver manufacturers, we can get more for our money and something uniquely tailored to our application by doing it ourselves.
Good job Chris. It is very difficult for engineers to talk and you are doing very well. Most engineers I know are happier in their lab with a computer and lots of test instruments, not so much talking in public. We would like to know you better as there is a lot of knowledge in that young head of yours.
I'd love some answers on how to get into speaker design! I'm a software engineer with a past passion in electronics and with all the free time I've supposedly got on my hands thanks to lockdown I'd love to get a more in depth experience working on audio and electronics engineering projects. Any resources you would suggest?
My speakers has the front of the cabinet sloped where the tweeter is 2 inches behind the woofer. My tweeter will sound great & 3 dimensional, but n0t the woofer. When I move the woofer toward me 2 inches its 3 dimensional sounding , not the tweeter. I can tilt the speaker forward, where the tweeter is even with the woofer, but this makes the woofer point down.
It never ceases to amaze me that companies, engineers, and MANY others; go to the end of eternity to design "speakers" and all of its entities; to get the BEST sound possible; and then turn around and set it so that nothing comes out of the speakers, but "distortion", and "noise", etc, etc and Etceterrrrra, called "Music"! What in the pluperfect hallelujah hath mankind wrought? Oh well
thanx for this. yes, there is math/measurements. i find multi ch vids try to measure sound quality. in the end, i think it is about experience. 2, 4 or 20 channels, is perfect, if u can master sound n acoustics of room which is purely subjective.
Phase alignment is only important for recording. In the home the only thing that matters is how good it sounds. That said, the human ear cannot detect phase alignment above 5KHz. The ultimate phase alignment is between your front and rear hearing perception (talk about sound stage!) -- .006 seconds
@Fat Rat -- The Buchardt is good, but certainly not unbeatable, especially given its price point. www.audiosciencereview.com/forum/index.php?threads/buchardt-s400-speaker-review.12844/
Wow, Paul’s not the only star ⭐️
Well done Chris, always something
new to learn in this hobby.
Thanks Guys. 😎
Always something new to learn, even if it's your profession.
Hi Chris!
You're a brave Soul to come here and face these punters, ha-ha!
Thanks for the info.
Interesting stuff! : )
Thank you! I look forward to doing more.
What's a punter?
@@QoraxAudio
That word actually has many meanings, including usage in British slang, which is where I know it from.
It can mean one who punts a boat, one who bets against a bookkeeper or gambles, one who is a prostitute, or who watches pornography regularly, one who is the customer of a business, and one who marks points in the game of Basset.
The usage I intended as the context I had heard it used (and which may or may not be entirely correct), is that of the 'arm-chair know-it-all', also sometimes referred to as "the peanut gallery" (which, in turn, has its own origin and meaning, coming from the vaudeville era):
People who hold forth on a topic or subject they know little about as if they were the World's foremost authority, often relentlessly heckling the actual authority on the subject.
To be clear, I meant it in a half-joking, light-hearted way.
There are actually some very knowledgeable people who frequent here. : )
And that, my dear sir, is your etymology and vocabulary lesson for today. 😆
Good start out of the gate. The difficulty I see in presenting this topic are the problems with trying to present highly technical information to a mostly non-technical audience. At least one assumes that.
Speakers push air around. How well they do that, by following the original signal with the same intensity (volume) as well as accuracy is the rub.
Frequency dividers, crossovers are inherently problematic. They introduce delays in the signals, and especially so with each other.
As an example (wish I had a white board or overhead), the signal path to the bass speaker (woofer), generally has a large coil of wire tightly wound. Its through an increased magnetic field when a signal applied, and collapsing magnetic when the signal is reduced, creates sn opposition to rapidly changing signals. It therefore makes a good electrical device to block ever higher frequencies. It may very well pass 100 changes per second (Hertz or older reference CPS), but increases resistance to change, say at perhaps 2000 Hz.
The two way design employs a high frequency speaker. By way of it's physical design, it's tinier and generally rigid radiating surface, responds well to more rapid signal changes. Since it can't physically handle lower frequencies, another electrical component is used to block or attenuate lower frequencies, while passing higher frequencies, all the way up to our hearing limits. At my age those limits come sooner. I'm no longer able to hear 18,000 to 20,000 Hz. Still like most people who have been listening through decent equipment for decades, can usually tell a nice sounding tweeter from one that can't keep up.
This most basic description of connection to a simple crossover, is only adequate to illustrate another problem with time alignment. This is referred to as a 1st order crossover. More complicated crossovers, which use additional inductors to pass low signals, and capacitors to pass high signals in a combination, create further intrusions into the phase alignment of a reproduction of sound.
The electrical & magnetic properties of an inductor cause the changing voltage to lead current flow. [Quick overview: voltage is the applied pressure in the completed circuit, whereas current is the actual flow of the electrical energy.]
Capacitors on the other hand, allow the current to pass before the voltage catches up.
Again an over simplification.
When the woofer receives a changing signal, it's movement will be slightly delayed by the arrival of current. The tweeter on the other hand receives the current ahead of the voltage.
If you're following up to this point, you can at least tell, the diaphragm of each speaker will move at slightly different rates because of the necessary frequency dividing electrical components of capacitors and inductors. So no matter tge actual physical distance to your ears, the electrical signals will be controlling how soon each speaker begins and ends movement.
Now that I said all that, in order to make sense of it all, it will be drawing time. Thank you for your patience if you read all that.
@Fat Rat
...uh, huh.
Should I call you Captain Coffee, then? 😆
Love the crossover remark! “So your tweeter doesn’t woof and your woofer doesn’t tweet” ✌️
Both the title and description of this video are a bit misleading.
Speaker designs by Jim Thief, Richard Vandersteen, John Dunlavy, or Earthworks' Sigma 6.2 use passive 6dB crossovers (AFAIK) and could validly be said to be demanding on the drivers, but that caveat aside, they are phase and time correct and therefore do function as their manufacturers claim.
John Bau's Spicas used 6dB tweeter networks, but 24dB Bessel networks on the woofers. Further, the publications of Samuel Harsch, Steen Duelund, or Kido Yamanaka make it clear that all 6dB networks are not a requirement for time-aligned speakers - multiple mathematical solutions can be made to work.
It is also possible to design a crossover network with multiple slopes, so that the initial rolloff is at 6dB but transitions to 12dB so that the drivers are better protected (and the bandwidth over which the drivers overlap is reduced). An example of this design philosophy is Jeff Bagby's series-crossover Tributes (available as a quite reasonably priced kit from Meniscus Audio, and Jeff's technical writeup can be downloaded gratis from Meniscus' website).
Yes, it is true that more than a few speaker designs have sloped or stepped baffles but aren't actually time aligned.
But not all speakers are like this.
A less misleading title for this video could have been "Why many speakers that appear to be time aligned, aren't".
Preach brother
Nice debut Chris ! Do some more and Darren too !
Chris, you should call it: "Why time alignment doesn't work with me?" You mentioned Dunlavy Vandersteen and Thiel - it worked worked them all right. It takes some effort to make a proper time aligned speaker with the 1st order crossovers.
Hi contradicts himself several times..time alignment, is difficult to get right, but when employed properly, there’s no going back to non time-aligned monitors
@@KCNYC I own time and phase aligned speakers with first order cross over (Lumen White Lumyere) and to me it's one of the things that make otherwise great speakers sound very authentic and natural (granted, this is a difficult task and comes with its challenges since the drivers are taxed more, drivers need to be aligned perfectly etc). There are many aspects to a great sounding speaker but speaker brands which heavily focus on getting psycho accoustics right and see it as a fundamental part of their design (to which time and phase alignment belongs as one part of the whole equation) just seem to sound very natural compared to brands which leave that more aside. Good examples would be Lumen White, Kaiser Accoustics and Marten speakers. Whether you personally like them or not, they definitely belong to very natural sounding ones and it has to do with giving psycho accoustics a high priority in speaker design, e.g. resonance patterns, acoustically more neutral sounding drivers etc
I was fortunate enough to meet Chris when I visited ps audio last fall. He was a gentleman, and enjoyed sitting in listening room 2 and discuss speakers and the company. Thank you Chris!
Hey, Chris! It’s nice to finally meet you, sir. That was a top notch description and pretty easily understandable explanation, which is much appreciated. Loved this vid, and I’d love to see more of you talking HiFi, but obviously and especially speakers more in future vids.
Thanks so much, Chris! 🎶🔊😁👍
Hey, it is nice to see you and this was number one video on PS Audio. It is great to hear you pass on your knowledge to everyone that has interest in high end audio. It gets better the more do to it. thank you!
Great to meet ya Chris. Welcome to the world of critics! Instead of being a critic today, (I'm a recovering one) I'll throw some advice. Engineers like to use physical examples to explain and show how one thing effects another like USING YOUR OWN SPEAKERS YOU HAVE BEEN DESIGNING!! WHERE ARE YOUR SPEAKERS! (Whew.....) Kinda still in recovery. Thanks again and see ya soon.
It has been my consistent experience with analog crossovers that filter slope and Q, and path length alignment are important. At least near the crossover point, where both drivers are contributing significantly to the output, it is desirable for them to be in phase. At the very least, if they are not, then frequency response aberrations which are often not easy to correct will occur. I find that perceived clarity and coherence improve when this condition is met as well. Of course, this requires that the path lengths of the drivers to the listener’s ears and the crossover filters are optimized as a complete system. With few exceptions, I agree that first-order filters require too many compromises in other areas. An exception that comes to mind right away is the Quad ESL.
Thank you for introducing me to dunlavy duntech. Changed my life.
As best as I could tell, what that Mr. Brunhaver is saying is that even if physical alignment of drivers is used to achieve phase coherency in the overlap region of two drivers that interface, this won't eliminate the overall rotation in phase that occurs in a speaker from low frequency to high frequency. So what? Does he mean to imply that since the overall rotation in phase cannot be avoided through physical alignment of the drivers, that there isn't any benefit to making the two drivers phase coherent in the overlap? The impression I'm left with is that he doesn't understand the reason that with most speakers the tweeter is vertically aligned with the rear of the woofer cone. This is done for reasons that have to do with the proper summation of the two individual wavefronts within the overlap frequencies, such that the main lobe, for any frequency within the overlap, will extend along a horizontal plane, rather than tilt up or down. If the lobe tilts up or down, then the tonality of the speaker will be different at different distances from the speaker. In order to avoid tilting of the lobe, it is necessary for the two individual wavefronts to be phase-coherent, not merely at the exact crossover frequency, but throughout the overlap region. Vertical alignment of the two drivers is one aspect of achieving this goal. It might not be essential to do this in order to achieve this goal, but when it is done, it is done toward achieving this goal, and it makes perfect sense, irrespective of the fact that there will still be an overall phase rotation from low to high frequency.
Good discussion. Very nice to hear a well tempered discussion of the subjective trade offs of time alignment approaches from someone with experience.
I really enjoyd the depth of knowledge Chris share in this video. Really great work. Looking forward to more videos from Chris and the rest o the team!
Dirac sounds great, was really impressed when I had a chance to try it.
I really do think most people in a casual listening setting aren't highly sensitive to phase, but when it comes to active mixing/engineering it definitely makes a difference. Interesting to hear your perspectives on crossovers and I'm glad you touched on DSP as well.
The vowels and consonants from the voice sound are coming from a foot apart with normal 2 way speakers but with coaxials and FR drivers the vowels and consonants are coming from the same point. I can hear the difference and you get a way better sound image as well.
Thank Chris good job, Jim Thiel made great speaker and a great company.
Interesting point of view Mr.Brunhaver. I listened to the Eikon Image 1 DSP speakers at the 2019 Capital Audio Fest with Gayle Sanders sitting next to me answering all of my questions. This was of course before the horrific COVID. I was truly impressed with what I heard and if I had $25K laying about for a second system I would be tempted to purchase.
Glad to hear from and see Chris in this video series. Kind hard to say no to the boss. :-) but Paul generally knows what he's doing... 99% of the time any way.
There are compromises with every facet of speaker design, and you pick and choose the compromises that matter to you. For me, time and phase alignment make absolute sense. Having drivers purposely firing at different times, make no sense to me.
Interesting subject.
I too have recently started using DSP as I have added a sub to my set up to complement the Magnepans. So I'm using the DSP for cross over and time alignment between the speakers and the sub.
I just thought it was worth mentioning something that you touched on regarding the DSP approach adding additional stages of A to D and D to A conversion. This isn't true if you are using an all digital, pre output signal path like I am. So I run the DSP software on the same computer that I stream from.
I just wanted to get that out there incase your comment regarding conversion misled some people in regards to using DSP.
Obviously I totally appreciate that you would be referring to DSP in the context of speaker design so this would be handled inside the speaker after the analogue input and would therefore be subject to the additional conversion which you speak of.
Otherwise, great video and I look forward to hearing more on the subject of speaker design.
Hey Chris I thoroughly enjoyed listening to you explaining this info about speakers, I found it very interesting thanks.
Great presentation Chris. As with most products where one design characteristic is touted as the big game changer, the most accurate response is “Actually, it’s a bit more complicated than that”.
My read: don't place too much weight on just one parameter or measurement.
Or: there's no such thing as a free lunch
My read: every speaker design is a compromise. Some focus more on one thing, others on another thing. Essentially, that's what you said.
I have 3 BBE balanced pro-audio processing in between my Outlaw pre-pro and Yamaha amplifiers in a 5.2 surround system. It sounds more open and 3 dimensional compared to them switched out of the signal path. I love it, never getting rid of em.
This is a great topic to address. The concern involves only those frequencies that are shared by two drivers (the crossover region), and most importantly, the mid driver and the tweeter (not so much the bass and mid driver). Important to understand that the length of waves at @ 2-3 kHz (the common crossover region) is between 4.5 and 7 inches in length. So if the tweeter was out of phase by @90 degrees or @270 degrees (a 180 degree out of phase condition is easily managed by reversing the polarity of the tweeter), then the driver position would have to be shifted by 1-2 inches, creating a step and can create as much difficulty as the out of alignment condition.
This is kind of old but still. Using Audiolense for time alignment is audible. My experience, and others, is that it improves imaging especially the centering of the vocals. Correcting (EQ) frequency response is much more audible. Compare the same FR correction with or w/o time alignment is definitely audible. This is done using active design i.e. separate amplification for each driver and digital xover using FIR filters.
Fantastic to hear mr Brunhaver thoughts
Very entertaining
Thanks Chris,
Very informative insights!
Dirac does work well. It was smart of MiniDSP to put the Dirac button on the middle of the remote so user's can A/B test it instantly. It clearly works.
I imagine in PS Audio will ultimately produce a system that has a mic calibrated streamer/crossover/fir filter/preamp that feeds multiple amps that are perfectly matched to each driver. At least I hope they do.
The difficulty is that it would require a lot of wires and stacks. So the preamp would need to be wireless and the amps mounted in the speakers.
...and DACs in front of each amp of course
Hi Chris, my name is Hans, and I was told that you have answers to some questions that I have. I am thrilled with the opportunity to communicate with you!
I purchased a FreeDSP-Aurora, and will be tri or quad amping. I'm in the process of re-capping, so don't have everything in place. The Aurora allows fairly long FIR filters to be programmed. So much to do, so little time!
Fantastic information Chris. I really enjoyed listening to you explain time alignment. Looking forward to your videos in the future. Nice to finally see you in person. Paul has been telling us great things about you. I cant wait to buy a set of speakers from you folks!
this stuff is fascinating, and when its done right, its a whole new world.
I've always been amazed at the difference it makes in clarity and depth when you reverse the phase on a tweeter. It doesnt always work with every build, but when it does, that tiny mod is a life saver when you're trying to get a speaker to sound just right. It's useful when you are running the tweeter on a 12db slope and your mid or woofer on a 6db slope.
You always reverse the polarity of a second order crossover.
@@Smart-Skippy It's not always beneficial. Nothing about speaker building is guaranteed in my experience. But hey, whatever works for ya
Kudos for the engineer speak. Don't think you could have pulled this one off Paul.
Why do you think I let Chris do it? He's much more informed than me.
I think it was an answer to my question regarding Cabasse Galion V speakers. Thanks therefore. For me, as a non native english person, it was heavy to understand, but I Look forward to the upcomming videos. Thx
Great video Chris hope there's many many more to come
From what I remember speakers like the Dahlquist DQ10 were time aligned phase arrays which tended to image very well.
Hey, thank you for answering my question! It's Adam.
Very well done.
Hey Chris, that was great. I've just built a new tall 3-way with my old 20+ years old drivers (by intention!) to see what I can achieve with them with a 8ch MiniDSP processor and a UMIK-1. So far the results are mind blowing, I think I'll keep these drivers in these boxes as is although they represent absolutely 0 value by now. I'm just playing around to see the difference between a heavily corrected system and a slightly corrected one and I think here one has to find her/his own taste, which kind of overall sound and behavior to choose. Now I seem to love a moderate bass correction (always the most problematic part due to room modes) and just minimal midrange/tweeter adjustment, leaving some naturally occuring smaller anomalies in the upper range.. not much, just a bit. That combined with the somewhat more corrected bass provides a great blend of the "old world" and the DSP'd world and the result was also for my wife and stepdaughter like "wooow". Great to see DSP not being the devil of audiophile societies but an option for further improvement - if applied well, with care (!).
You need to read anything written by Dr. Richard Heyser, Dr. Floyd Toole, Dr. Christien Heil or Thompson Ambrose and then learn how to use a TEF or Easera. Correct phase at crossover point is paramount.
Funny you should bring up Dr Toole, I was wondering what Harman's research found on this topic. Time to Crack the book
You can measure the phase with holmimpulse. At crossovers invert polarities on speakers to align the phase curves. You can verify this on CSD with ARTA : Red colours all the way. At the end you move back the tweeters and mediums to have the best 0° phase. Holmimpulse and ARTA are free.
It's clear you know what you are talking about! Great vid, i just hate you look shy despite how much you know. Thank you for the info!
I feel so lucky to have purchased my Medowlark Kestrels a true time coherent speaker. Yes it makes a difference. Just watched another video about the other big debate - do cables especially speaker cables make a difference. What is was looking at is the higher end cables seem to "time coherent" than the cheaper cables. It isn't about the frequency responses or anything else. In the end after many many auditioning times with many many speakers the Meadowlarks just fit perfectly with me. That is the key in all this. You find what fits you and moves you. Been over 20 years with my speakers and have never wanted for anything else. I feel sorry for those audiophiles that are in constant change searching for that "perfect"" sound. Time to stop listening for the problems and enjoying the beauty of the music.
Very informative video! Learned something new here now. Quite nerdy too. Which I like
Good info Chris..
Here are some things many need to know,yet evidently,do not.First of all,we have a more diverse set of listeners than ever before.WE have people like me,who listen to music on UA-cam,on my aligned system,on my tablet,from records,from turntables,and on anything from MP3/MiniDisc to a Richard Long system.I have heard expensive gear,and boy oh boy,some of it was amazing,in 'proper' listening rooms.The problem is,I never really have time and interest to sit in a specific spot,in an acoustically treated room,and spend my whole listening experience there.It's just not practical,and I personally do not know anyone who does.
Now we have way more Class D amps,more digital listening/recording devices,and a host of converters,both ways,to suit all prices and preferences.I recall going to someone's place in Harlem,and in his studio,he has Tannoys,and yes,I was impressed.I have been equally impressed by systems with EAWs,Mcintosh,PSBs,while others have come to hear me spin records on Alesis Monitor 2s and Kenwoods,with a Parasound power amp,etc,etc.Like others here have said,there is no magic bullet,no magic speaker,and what is revered by few,can be hated by many.This is yet another subject like which cables,what tonearm,what plinth,what type of wall treatment,and the FACT is,rarely,if all ,can you ever tell the differences blindly.
I do have an old time delay box from way back,and it works for me,and I found it in a thrift shop for $20.I poked around,and tested over and over,until I heard real improvements,yet then again,I love to mix and match products anyway.Point being,today,people want it loud,quick,easy,and uncomplicated,and everything does not have to be scientific,which works for us all,because diversity brought us here,to the Internet,with gaming PCs,Mac Minis,Iphones,Notes,Wacom tablets,VR headsets,and advanced AI.Those who get stuck on anything,get left behind..
You ask 10 different ‘experts’ the same question, you’ll likely get 10 opposing opinions. They can’t all be right. I think Chris was making every effort to be true to his beliefs!
The recorded event is captured by the mics. Loudspeaker drivers reproduce the recording and launch it into our listening room.
The goal is to recreate the recorded event for enjoyment.
Putting it all back together as it's launched and propagates into our room is no trivial task... even before the room enters the equation.
What we the listener experiences while enjoying our systems is the direct energy combined with a great deal of reflected room energy.
That said, the image recreation benefits from the reflected energy being phase accurate uniformly with the corresponding energy from the other channel (ie., Left to Right, or Right to Left).
So, overcoming the challenges of phase shift from coils and caps, physical discordant drive units, etc, present significant obstacles to imaging and soundstaging.
However physical and acoustic ASYMMETRY within the listening room can shred the image and soundstage apart.
Yes the Vandersteen speakers are quite awesome and they do have a special time alignment thing going on with the sound.
Personally I don't care what has and what has not, I just want the damn thing to sound good, and I agree with and without it can sound pretty good.
My speakers has the front of the cabinet sloped where the tweeter is 2 inches behind the woofer. My tweeter will sound great & 3 dimensional, but n0t the woofer. When I move the woofer toward me 2 inches its 3 dimensional sounding , not the tweeter. I can tilt the speaker forward, where the tweeter is even with the woofer, but this makes the woofer point down
Try with inside boxes just like I recommend in my videos, it isolates all the unit sounds of your speaker, bringing the sound quality up to an entirely different level. Also having you speakers relatively close to the back and sidewalls will maybe also help.
I have a pair of Theil tweeters left over from when I worked at Audio Gallery and they are sweet tweeters as long as you use them in their range. They sounded awful with any kind of lows entering them. We had the 3.6's and the 5's.
Time and phase domain is one of the many important parts of speaker design.
Fat Rat Pretty much everything he says I agree with. Most simple designs don’t require a whole lot of time alignment built into the crossover but even with some simple 2 ways it can be an important part to the final tweak of the speaker. Always exceptions.
Fat Rat Indeed. Many of the designer type love to chat at shows and the like. I’ve chatted with Andrew Jones before, such a nice guy. Sandy Gross I got to talk to at Cedia last year. He sold Golden Ear tho just lately.
Fat Rat You would of of really enjoyed it. They are all passionate guys like you and me. Real cool
I got my Dahlquist DQ-10's . . . staggered baffles for tweeter to mid-range . . . I just re-capped the complex crossovers to handle five drivers per channel . . . time and phase aligned. Jon Dahlquist and Saul Marantz. Alignment to the woofer is less critical than hearing range 150Hz and up.
The DQ-10's great speakers specially if you consider their age. The DQ 10 and the Time Windows used to be among the best.
Wow, really interesting topic, as a newbie I appreciate it so much.
Greetings from Brazil!
Excellent video. Just so you know, Finite impulse response technology definitely does work. KS Audio, a German company, perfected the technology probably 20 years ago. Other companies are doing it now. I looked for your links below but did not see them.
Highly informative! You should mention active speakers. These designs can better control both crossover and phase. Nevertheless your point is correct - we don’t need absolute phase accuracy - “smooth” is really what is important and sharp changes in phase do cause audible issues. MQA and minimum phase filters mess up phase and I find this to be audible with careful listening if the phase delays occur at audible frequencies (high frequencies are delayed). My preference is linear phase. Most speakers have phase issues in the bass but this is much less audible then phase delays at mid range or higher frequencies. The manifestation of phase delays from minimum phase filters is usually less accurate imaging. Some minimum phase filters like MQA also cause mild compression (apodizing type filters). Active speakers aren’t perfect but if well designed they possess several advantages over passive designs.
Good one Chris! Excellent debut, really enjoyed the critical point of view and it'll be fun trolling around here in the out of phase waves you've caused.
Linear phase active crossovers are always an option for higher order crossovers that have phase coherence. It's the method I use, and it's amazing!
I have switched between minimum phase and linear phase crossovers digitally, and there certainly is a massive difference that I heard anyway. Linear sounded so much natural, but I do think you need the speaker itself to be perfectly time aligned first!
Thank you for the video. I've never thought about something this difficult as I'm a very casual audiophile (if such thing even can exist), I just enjoy listening to my music and trying to get the best performance out of the hardware at hand. But now when I hear "time aligned" I will know what to ask about it.
I like this format, at least answer starts at very beginning of video 🙃
Great info. Thanks for the presentation. It was very informative.
Great to see others from PS Audio share their knowledge, hope we see more of the technical / engineering team have guess slots on Pauls UA-cam show. Hope we see more of your insights soon
Great video with a lot of information. Just to double check, first order crossover does not produce phase irregularities??
Where can we see the impulse and step responses of the FR-30s?
Thank you Chris for sharing your knowledge of speakers. I suggested paul put you into some videos since he has mentioned you in previous videos. Hope to see more! Feel free in future videos maybe talking about your background and how all you designed for and perhaps speakers you had a hand in designing. Thank for the video! Perhaps scott can get in more videos as well. Haven't seen many from him. Enjoyed the few he did. Thanks!
OMG can’t wait for the new book from Paul’s series: “0% truth: how I moved from getting skinnier to actually becoming younger!” ROTFL
Just joking, excellent explanation. Well done Chris!
OOOoohh!! Killing the Wilson Audio sacred cow!!
Nice explication. Tray to look in the camera . And hope for more videos
It’s probably difficult when talking nonsense. Done correctly with proper crossovers, the results are amazing!
Well then there is surround systems and car audio in which as far as I can see there is no way to get all sound waves to your ear correctly, or am I a bit dumb?
I also discovered (as you mentioned) some other variables that effect phase irrespective of passive crossover components and driver position, and I postulated that these variables are related to the great difference between a mid driver and a tweeter in the mass of the two driver elements, and the physical resistance to motion exerted by the size of the cone anchoring material (just a surround for the tweeter, and both a surround and spider for the mid driver).
Things that never change: Loud speaker engineers agreeing on speaker engineering and design theory
Good job, Chris; hope to hear more from you in video circuit.
I was always suspicious of "time alignment" when the difference of the sound reaching your ears in was probably a 0.00000000076543 difference.
Ya and I would think in a live setting
the high frequency sound would be
traveling faster then the Lows.
All so very confusing, my Harbeth’s
sing so I’m happy. 😎
Some people be going out buying time aligned speakers with no acoustic treatment in their room. Unless it's in a live setting with multiple speakers in the same range in an array I'd say time alignment between tweeter and woofer matter a hell of a lot less than phase issues from first reflection points
The offset alignment of drivers often allows the crossover to sum properly at the listening position, so time alignment in a way can be misused (or assumed) and is not the same as phase coherent. I can never understand why audio enthusiasts when given information instinctively run the wrong way with it. :) Music is based on time --- phase, time, and Q always important. :)
cein o'dowd , Hear, hear!! Well said brother.
In terms of geting the phase right, comb filtering probably has more the most noticeably effect, since it will result in multiple peaks and dips in the ampitude response of the loudspeaker
And this is why you should just listen for yourself and buy what sounds best to you. There are a lot of theories on speaker design but a few are commonly offered as the holy grail and time alignment is one of them. It's completely misunderstood by most audiophiles and likely, as Chris points out, by most designers. Acoustic centers are not determined by where the voice coils of a speaker lie so very often the stepped baffle or raked speakers are not actually time aligned regardless of the crossover design. Crossover design, however, does play a critical role and there are plenty of ways to design a crossover to approximate time and phase alignment. Even so this will not guarantee good sound. There is no singular part of a design that if you get it right then, shazaam, audio nirvana materializes. And that's why you can design speakers many different ways and arrive at a great end result. Or, conversely, you can execute one design theory to perfection and still end up with something that sounds like a first time DIY'er built it.
Excellent description of time alignment 🙌🏻!
That's not entirely correct: absolute phase is what's difficult to hear. Relative phase (that all drivers should be moving in-phase) isn't. Also, what's being time-aligned, if done correctly, are the so-called acoustic centers of the drivers, which isn't necessarily their membrane surface as claimed in this video, but where energy conversion is taking place in each one of the drivers (usually closer to the centre axis of the voice coil, but really, no educated guess will replace an exact measurement). I've designed loudspeakers of the kind discussed here, and can vouch for their ability to produce significantly more realistic sound, but yes, they're a major PITA to design and build, and it's also true one needs wide-band drivers to begin with. I'm convinced time- and phase-coherent speakers are worth the effort. And no, DSP-based systems, while easy to implement, sound unnatural by comparison - those just a short-cut for the lazy if you ask me. A properly designed system will always be superior to one using makeshift solutions to problems one wouldn't have in the first place if said system were designed from scratch.
I would add to things. 1) If there is a vertical offset between the drivers, and one is closer to ear level than the other, then this will force a difference in their relative positions. 2) Nice name. Leon Fleisher died too young.
@@jonathanknight8251 Referring to the legendary pianist and teacher born in San Francisco in 1928, he's still alive (at least I'm hoping so?). And yes, what you're referring to vertical offset is the same I meant when I say all acoustic centers of multiple drivers must ideally be placed at an equal distance from the listening position. It can't really be done without measuring equipment, but it's worth the effort, and so are filters that get the drivers to move in phase.
I can't listen to music on conventional multi-way speakers. I hear only non-logical sounds.
Phase and time alignment are the most critical thing about a speaker. That's why I'm listening to an open baffle with linear phase crossover and the necessary delays to different drivers.
The only one better speaker than that is full-range driver. :)
I agree dsp and delays to the multiple drivers is the best.
What amps do you use?
Active crossover? What brand did you go with?
Nah...dbx or peavey or rane or ashly or
Pro audio rackmount stuff
@@zeusapollo8688 class d or class ab? Monoblocks for each driver? Stereo amps for each bank of drivers? Or a multichannel receiver?
@@thunderpooch oh stereo banks of power amps
How come midrange drivers not called Squawkers anymore?
For the same reason that horns are called waveguides. It's a marketing thing, squawkers doesn't sound Hi-Fi.
Very interesting opinion and useful information!
What I get from your description is that time alignment works and from my reading of professional speaker builders Wilson,B&W,PSB,Magico,YG,I have read some of there reasons for there builds learning of sound and how it transmitted and time alignment is very necessary,so all I can surmise is that I feel sorry for Paul hiring you
Time alignment is important with multi channel (3+) systems, 2 channel systems you can have a dipole with a crude out of phase delay channel bouncing off the front wall so who cares. "Folds" in phase hide details (information) just like peaks in linear response. Don't forget amplifiers often have significant phase rotation. When the system is slopped together phase doesn't matter (ergo some of the academic research on the audibility of phase never tested the source or the amplifiers for phase distortion). As the audio system's obvious flaws are continually diminished smaller flaws suddenly become "a problem". Passive loudspeakers are obsolete because of what you described, but the inherent flaws of 2 speaker playback are greater than the flaw of phase distortion. So you have to correct listening to only 2 speakers first. :)
Aka time alignment doesn't actually "align" the sound waves, but it sounds different, and therefore we think that the difference is the sound of aligned sound waves?
Very interesting Chris. Great topic.
Are those Buchardt S400's in the back? Time-aligned drivers :)
Time aligning drivers (physically aligning them) can help with phase summation in the crossover. But if I had to choose, I'd rather have the highs and mids hitting my ear a smidge before the bass. Your brain hearing the first consonant or sibilance sound of a word, before the bass arrives, gives the impression of clarity.
I often like to place the subwoofer in front of the speakers but then switch the phase 180 degrees. It sounds quite clear for vocals, plus the sub hits you harder and doesn't interact with the back wall as much.
To the point of the video, they are physically aligned (because of the tweeter set back in the waveguide) but are not using a time/phase coherent crossover (not that they claim to) That was sort of the point of my video. Again, I think that that theirs is the correct engineering decision to make but they are non minimum phase etc.
@@cbrunhaver Hi Chris, I understand your point and it's very informative. I was just happy to see my speakers in the PS Audio office, it's a good look (both aesthetically and for Mads at Buchardt :) Thanks for the video!
supergiddy3 yes, I believe that our companies are rather friendly with each other and I have found his engineering work and designs to be very well implemented.
Meadowlark Audio had some nice time aligned speakers and now switched to a DSP approach.
The only time alignment I've had contact with was room correction set up that the amp did you positioned a special mic at the listening position and let the amp do its thing and you could really hear the difference but isn't that really a form of DSP ? (Sorry it had nothing to do with the speakers themselves)
Howdy Chris.... you've met your Aussie match here btw ..!
I'm an electro acoustic engineer and a recording engineer at Sony ...
I go along with what you say here but the trick with speakers is to select the right ones that go well together in the first place .. after that it's a matter of phase and amplitude control using XO networks ... first orders are excellent for lower frequencies yes but for HF drivers 2nd or 3rd order networks are essential .
I have software that measures both group delay and plots decay responses shown in the familiar waterfall plot . I use these extensively to achieve the right result . Do you at PS Audio use component valuation software in your XO designs ? This software works like pspice or ORCAD and changes the values of XO components to fit a particular response envelope with minimum group delay ..
I guess you'll be handling all things speaker from now on as Paul does refer to you as the speaker guru ..!
I used to work at Sony for a few years and helped support Kaku and Yuki with the SS-AR1 and SS-AR2 product.
Yes, I have a couple of crossover optimizer/simulation programs that we use. We also design all of our own drivers and work heavily to optimize those for the application as well.
Fat Rat I designed prototypes for Duntech audio ... was an audio consultant back then
cbrunhaver ~ how interesting.. you and I could chat and argue for hours on speaker designs I'm sure lol 😂 No two engineers always agree ..!
Yes these component simulators only optimise as you said .. the final test is in open space using LMS software as I do to perfect the correct coherent phase and amplitude response . I keep an open mind about making your own brew of in house drivers .... my first reaction is why reinvent the wheel ? Big names in electrodynamic transducers ( Dynaudio comes to mind ..) have done nothing else BUT design high end and very predictable drivers with amazing consistency in characteristics. I know ... a very large number of high end speaker manufacturers seem to build their own drivers but in reality they contract out the designs . Whether this is good I'm still not really sure .. new designs require extensive testing and honing in order to narrow down any 'production spread '.
I'd like your comments on any readily available software for T/L enclosure designs ... as these things do not use any lumped constants and this is what makes them so hard to design ...
Cheers for now ...
@@janinapalmer8368 Well, I'm always glad to chat! My email is chris b (at) psaudio.com.
LMS doesn't actually measure acoustic phase (it can derive it through a hilbert transform but that assumes a minimum phase system). A shame about Chris' death, as I was looking forward to his new system and was a huge fan of his work.
Well, the likes of Focal and Dynaudio don't sell their drivers anymore (except to car audio), as it competes with their box speaker sales. JBL came to much the same approach in the pro world. For us, it's a value add - when buying an off the shelf driver, you're paying for their engineering staff and supply chain etc. Some guys have incredible robotic assembly lines and smart purchasing and are hard to beat, as compared to a lot of high-end driver manufacturers, we can get more for our money and something uniquely tailored to our application by doing it ourselves.
Very interesting - thank you!
Good job Chris. It is very difficult for engineers to talk and you are doing very well. Most engineers I know are happier in their lab with a computer and lots of test instruments, not so much talking in public. We would like to know you better as there is a lot of knowledge in that young head of yours.
Interesting video! Looking forward to more.
Is Paul on Vacation, I heard he has a time share..?
Lol!
I'd love some answers on how to get into speaker design! I'm a software engineer with a past passion in electronics and with all the free time I've supposedly got on my hands thanks to lockdown I'd love to get a more in depth experience working on audio and electronics engineering projects. Any resources you would suggest?
Interesting post!
My speakers has the front of the cabinet sloped where the tweeter is 2 inches behind the woofer. My tweeter will sound great & 3 dimensional, but n0t the woofer. When I move the woofer toward me 2 inches its 3 dimensional sounding , not the tweeter. I can tilt the speaker forward, where the tweeter is even with the woofer, but this makes the woofer point down.
It never ceases to amaze me that companies, engineers, and MANY others; go to the end of eternity to design "speakers" and all of its entities; to get the BEST sound possible; and then turn around and set it so that nothing comes out of the speakers, but "distortion", and "noise", etc, etc and Etceterrrrra, called "Music"!
What in the pluperfect hallelujah hath mankind wrought?
Oh well
thanx for this. yes, there is math/measurements. i find multi ch vids try to measure sound quality. in the end, i think it is about experience. 2, 4 or 20 channels, is perfect, if u can master sound n acoustics of room which is purely subjective.
Phase alignment is only important for recording. In the home the only thing that matters is how good it sounds. That said, the human ear cannot detect phase alignment above 5KHz. The ultimate phase alignment is between your front and rear hearing perception (talk about sound stage!) -- .006 seconds
Now I'm *really* looking forward to a PS Audio (Sprout?) bookshelf speaker -- Paul hit the jackpot when he hired this dude!
@Fat Rat -- The Buchardt is good, but certainly not unbeatable, especially given its price point. www.audiosciencereview.com/forum/index.php?threads/buchardt-s400-speaker-review.12844/
With the onset of affordable and qaulity dsp settups. The analog passive crossover delay in "old school" high-fidelity speaker systems is a non-issue.