Some are saying (on that thread) that perhaps somewhere down the line Jim realized it was of no sonic consequence, but kept doing the time/phase coherent design because Thiel had already built a reputation marketing that characteristic. I think that is nonsense. John Atkinson at Stereophile once said that if everything else being equal, he did notice that speakers with time/phase coherent have and advantage in soundstage presentation. The difficult part is how to determine whether a pair of speakers is superior to another pair of speakers because of its time/phase coherent or something else. For example, the CS2.4 may have better soundstage vs. another pair of speakers but maybe because it is just a better design with better driver integration and not because of the time/phase coherent aspect. How can you 100% sure the CS2.4 is better because of its time/phase or something else? Maybe the CS2.4 superiority comes from its coax driver and the quality of the xover? So you end up comparing apples to oranges. The proponents of time/phase always point out to the "step response". But then if "step response" is so important, then you would think that non time/phase coherent speakers shouldn’t be able to reproduce music at all period, since in theory, if you can’t replicate the actual input electrical signal, then in theory, the output is all wrong and therefore what you hear should be all garbage. But obviously, non-coherent speakers can reproduce music just fine, therefore it is a contradiction, and therefore the "step response" is not a valid criterion, right? I’ve been thinking about this but nothing came to fruition. I have a couple of explanations but really it could be anyone’s guess. First, maybe our hearing is very tolerant. Even with non-coherent speakers, if it comes close to reproducing music, our hearing won’t really care much. But if the speakers happen to be coherent, then it would be icing on the cake. It’s like baking a cake. Anyone can bake a cake and most of the time, any cake would be fine, but if a really nice coherent cake comes a long, it would wake up our taste bud. Secondly, and this one may be related to the first, is that the step response in theory has infinite frequency bandwidth, but our hearing is only limited to 20KHz. I won’t go into the mathematical details about the infinite bandwidth stuffs but you could look up. So the step response is not a valid "test" for our hearing since our hearing won’t care much for any high frequency content. I would imagine that if we human being has supersonic hearing capability all the way to the MHz range, then I am sure we could clearly hear differences between coherent vs. non-coherent speakers and the step response would be valid. Of course if a pair of speakers are just plain garbage then well anyone can tell :-) Anyway, I’ll try to capture a step response in the next the post to illustrate the bandwidth limited theory. Looking at a simulation step response from one my design, it is consistent with what I said above with respect to our hearing bandwidth limited. Regardless of time/phase or not, I DO see an advantage in first order design vs. higher order based on various listening experience. First order filter is the only filter that does not have phase distortion. |
My study of audio and auditory neurology reveals that multiple parallel tracks decode the auditory stimulus, and the whole body is involved including the ears, mastoid process, sinus cavities, solar plexus and skin envelope - all working together to sense, decode and decide on the nature of incoming sound.
Hi Tom, Do we know if our ear drum can vibrate at much higher frequency than 20KHz? In order for out brain to process higher frequency, I guess at least mechanically, our ear drum is not the bottleneck which is something that can easily be determined. We evolve from primitive animals and I am pretty sure they all possess ability to hear at much higher frequencies because is critical for their survival, but as we evolve it is not as critical for us so I guess our ability to process high frequency is no longer there.
One circumstance in play is that the temporal domain is not limited to the 20kHz frequency domain limit. Onset transient form and integrity which we can reliably hear, translate to wave-forms in the 200kHz range
That is an interesting claim. Theoretically I suppose that's possible but how to prove it I can see it could be problematic. I am no longer as young as I used to be, but when I play a 15KHz tone, I swear I could not hear it :-) But music is more than just a single sinewave tone, so I guess it cannot be used as a proof. Raise your hand if you can hear a 20KHz tone. God blesses you :-) But let remove our hearing aside and look at thing objectively. Let's say if you were to design a speaker that acts purely as a transducer - that is it required to convert an electrical signal to acoustic sound pressure. Usually you would come up with a spec that say something like: My transducer can work from 0 - 200KHz or 2MHz or some frequencies with a certain harmonic distortion. So you would have to be able to show data to prove the spec. What you would do is playing various sinewave tone from 0 - 200KHz or to 2MHz and measure the sound pressure at various sinewave frequencies including distortion. My guess is the higher the frequency, the higher the transducer will show distortion and phase shift, and up to a certain frequency, the distortion will get so large that the transducer will no longer able to produce a clean sinewave. So with this method, you could objectively compare two different transducer. The problem with step response is it has such a wide range of frequency bandwidth that it is not easy to be used to compare or to characterize. Back to speaker design, I would suspect a true time coherent speaker will be able to produce higher frequency tone vs non-coherent speaker with less distortion. And of course, as we go higher and higher frequency above 20KHz, the distortion on average will get higher and higher for any speakers. Back to Tom's claim that we can actually process signal as high as 200KHz, and as I have said in my previous post that the higher frequency that human can process, the more likely we can hear the difference in coherent speaker. |
Hi Tom,
By tonality I suppose you meant frequency domain. But time and frequency domain are the same. You can convert from time domain to frequency domain and vice versa.
For example, if you have a waveform in time domain, you can perform Fourier analysis into frequency domain, but then later on, if you want, you can convert the frequency domain back to the original time signal with no loss of information.
You probably had in mind steady state frequency response. But when you convert from time domain to frequency domain, the phase information is still there, so no information is lost and the frequency domain is just as valid as in time domain. One is no superior than the other.
|
Hi Tom,
Thanks for your detail feedback. Something just struck me that could give me a clue. I keep thinking that music is limited to 20KHz, but that's not true, it's our hearing that is supposedly limited to 20KHz. BUT musical instrument has no frequency limitation. For example, when a drum is struck, the energy could be well above 20KHz. So the speakers whether we are aware of it or not, are required to reproduced music at quite a bit higher than 20KHz.
Anyway, more to come. |
Horses can "hear" a train more than 5 miles down the track. They sense that 4Hz wave through their hooves.
In that sense, I was wondering if dogs are just opposite? They can better sense high frequency sound. That's why my dog tends to act up when I play Diana Krall :-) |
but anybody try Acoustic BBQ speaker cables with their Thiels? I need to get some speaker cables with spade connectors (long story) and leaning towards trying these, unless someone can make a strong case for something different for less than $400. Same old requirements - less harshness in the treble, no loss of resolution or harmonics. Currently have AQ CV-8s which are the only audiophilesque speaker cables I've tried.
For high end speaker such as the Thiels, in order to bring out the best, I think you need some really decent cables. I am using Acoustic Zen Hologram II and the sound is a lot more sure footed than some low end cables. These can be found for about $500 here at Audiogon. |
We are barely scraping the surface here. The point is that the inner functioning of the ear mechanism itself is vastly complex, even before the neuronic energy is sent to the brain for processing. Hi Tom, I was under the impression that the only part of our body that can "sense" sound is the ear drum, but then what you’re saying is that since our ear drum is only part of a complex network that forms a mechanism to "sense" sound, therefore it’s possible our brain can get sound information from other than our ear drum. So there are three options to this argument: 1. If what you said is true, then even if our ear drum can only vibrate at 20KHz max, other parts of the ear has the possibility of vibrate at higher then 20KHz. I mean our body works like a machine in that sense. You need something to vibrate at higher than 20KHz for our brain to process signal. Otherwise where would the brain get the information from? 2. But if the ONLY part of our body can vibrate is our ear drum, then I don’t see how we can sense sound at higher than 20KHz. 3. The third argument is this. Considering the equation Y = 1/x. Y is approaching zero but never gets to absolute zero. Y will get smaller and smaller but never zero. In general that’s the nature of our analog world. Nothing is absolute zero. So our ear drum is like that. Is possible that the ear drum does not get cut off right at 20KHz, but its response after 20KHz gets smaller and smaller but just enough to be able for the brain to obtain information. And when we say "we don’t hear it", it’s our brain that "told" us that we don’t hear it, but that does not mean our brain did not process it. Think about it. If we are constantly being interrupted by supersonic sound, we would be driven to crazy. Trust me. I know. I used to have a girl friend lols. A lot of people say that truely low bass can only be sensed, and not heard. Our whole body can be thought of as a giant ear drum but for only very low frequency. Our body can actually vibrate but only at very low frequency. So when we "hear" low bass, I guess it’s more like the sensors in our skin telling our brain about the low bass (how else would we know if somebody punched us?). In that respect, not only our ear drum can hear low bass, our entire body can hear it too. |
1) connections on XO boards where multiple coils, resistors, and caps come together (especially joining the Litz wire to the bundle), and 2) connection to driver posts, all the more so with the Litz (but even the OEM solid core was pretty much a solder connection there).
Hi beetlemania, Does it mean you also did swap out the stock internal xover wires? |
Ours would act up at Kathleen Battle.
I would too :-) |
I’ve been thinking about the time/phase coherent and to be honest I’ve been back and forth - it matters, then it matters not like a girl choosing who to marry. One is for money and one is for love lols. A lot of it is more like personal experience than actual reasoning - like marrying.
But I think now know who but I need to know if she’s for sure lols. You guys probably thought I was gay right :-) ha ha ha ha ha |
my present work on these products strongly suggests that "neutrality vs musicality" is not necessarily a dichotomy. The assumption that we must sacrifice Articulation / Neutrality / Resolution in order to get "Musicality" is not necessarily so.
That is true albeit at a rather high cost. If you were to build low cost speakers, then it's not likely you can get all three- neutrality, resolution and musicality. Depending on the cost you may only get one out of three. And the more you're willing to spend on R&D and manufacturing, the more likely you'll get all three in one package. The Thiel CS2.4 is one of those rare speakers that are close to getting all three. In stock form, it has neutrality and musicality. Its weakness is in the high frequency extension resolution. Based on what others have done, it probably came down to the xover and internal cables, and it seems like if you upgraded the xover and cables, you can get quite good high frequency extension resolution as some have reported. Of course by the same extension, the same would apply to preamps, amplifiers, digital front ends, cables and so on. |
I was wondering if the CS2.7 uses the same bass driver as the CS2.4? At least they visually look the same and about the same size. |
A few previous posts ago, I said that the most apparent difference between first order vs. higher order is in the treble.
I also notice that first order filter speakers such as the Thiels make early 1980's recordings sounding more "musical" than they really are. I have quite a few early recordings back in the early days of digital recordings that generally sound sort of flat, analytical, and very lean with very little presence. But through first order, they actually sound musical and the sound has a fulsome quality that not there in typical conventional systems. |
Still, the little tiny Spendor S3/5 speakers I have can sound spooky accurate to real human voices. And at the recent Toronto audio show I attended, voices played via the Harbeth speakers (thin wall, wider baffle design) sounded more human than any other speaker system I heard there, regardless of price.
I think it might be just a coincidence. A lot of the "human" sounding probably has to do with driver selection and speaker voicing. I don't think a "thin" wall speaker has any inherent advantage in term of able to reproduce a "human sounding" vocal although I know what you meant. I've listened to "hard and stiff" wall speakers and a lot of them could reproduce voices very well so much so that listening to Diana Krall makes me think twice about getting married :-)
This is put in relief in comparison to the Joseph Audio speakers I have now (no I’m not getting rid of my Thiels!). The high frequencies of the Joseph speakers are shockingly pure and grain-free, without brightness.Cymbals pop out of the mix like a scrim of hash has been wiped away, and ring with more of the beauty of the real thing.
As Tom pointed out, the Joseph uses of the expensive soft dome tweeter from Seas probably has a lot to do with that. Soft domes have come a long way. In the past, soft domes although haave the sweet sound but they lack the transparency of hard domes, but nowaday the newer ones such as the Seas have both. I think Thiels tweeters use metal based material (aluminum alloy I think) and although aluminum has a lot of details and extension, it may not sound as sophisticated or sweet as the best soft domes from Seas.
I think, no doubt, that the designer will say this is a result of the benefits that can be found in going with a higher order crossover, and (in the case of the Joseph speakers, at least purportedly having a steep crossover shelving), to allow drivers to operate optimally within their range with lower distortion and less crossover interaction.
I think it may be a "red herring" in this case, since I don't think "high order" is what responsible for what you've heard. The drivers used in the Joseph play a large part to the characteristics of the sound. |
Given some of the recent posts about cable upgrades, I wonder if you notice that the "larger" the system, the more apparent the difference makes. For example, if you have small monitors in a small room vs. a large system such as a real four way speakers in a large area room, any upgrade such as cables will make more of a difference.
Let's take this analogy. If you have two digital pictures with different resolution, says one is 0.3 megapixel and one is 7megapixel, and zoom them on to two large computer screens, the difference in quality will be very apparent and the 7megapixel will have superior picture quality vs. the .3megapixel. But now if you display these two same pictures on very small computer monitors, the difference is not quite apparent and it is related to the picture resolution.
Now back to audio system, the larger the system will expose the quality of the equipment better because the sound will have to be projected in a larger air volume vs. a small systems. The difference will be amplified in a large system in similar way with pictures. I think that's why different people will perceive the same set component differently and some say the difference is significant while others may say the difference is minimal. From personal experience, I used to have a small system in a small room and it was hard for me to tell the difference, but now I have a much larger living room and much more high end system and any differences from different components are very easy to tell. Not only that, any difference when the system not fully warmed up is also very easily to tell.
Of course we have to consider the affects of "subjectivity", but what I said above plays a significant roll in how the sound is perceived. |
For those who are interest in a more detail discussions of first order/time phase coherent design, I have just created a thread specific to this subject. I actually have simulations, plots, graphs and not just subjective philosophical rubbish :-).
Personally I find it's a fascinating subject and hopefully I'll post some of my findings. Just be warned, some of my findings may go against your belief (but not too much). Tom may not like what I will say but ultimately we need to search for the truth. Vincent van Gogh tried to cut his ears off but I don't think he found any. Hopefully I won't cut anything off :-) |
"Since if a speaker can produce a step response correctly, therefore it is time-phase coherent, and therefore it must be "good" - is not something I hear him saying. He does say that "all else being equal, phase coherent speakers tend to produce exceptional imaging". (My semi-quote)
Yes, I think that's what he said. I may have mistaken "exceptional imaging" for what he felt as "good". I think I've read the same article that you read and he did say "exceptional imaging". It's been awhile since I've read the article (and I can't seem to find it anymore), but I also believed he said something to the extent "if everything else has been taken care off", that is if a speaker was well designed in other regards, then "time-coherence" may have an advantage of providing "exceptional image". Anyway, what I will write probably won't be based on my personal opinions but more based what I've seen from simulations with actual real world data. Others will have to draw their own conclusions. I only used John Atkinson's statement as an "introduction", but I will try to stay clear of using my own subjectivity.
I think that phase coherence (as an objective or success) increases the difficulty of making a good speaker by a large multiple, and that many attempts fail in many ways, including Thiel's attempts. I agree that time-coherence is not easy and that's why besides Vandersteens, I don't know of anyone else. I have seen a few measurements from Stereophile done by John Atkinson and he did question the validity of some of the claims of being time-coherent (not Thiel though). My document is not meant to be a vehicle for judging anyone's work, but it is meant to discuss "time coherence" in a objective way with actual simulation using real measured data (when possible), plots and graphs, and as I said, I will not include any of my subjective opinions. And for whatever reasons, this topic is weirdly fascinating to me, and I probably won't sleep well until I put the document together :-)
|
Atkinson and others have hijacked "coherence" to include the smooth phase transitions in non-coincident systems with large phase shift - as long as the phase transitions are smooth, some call it phase coherent
That is certainly what I understand as one class of "phase coherence".
Thiel, Vandersteen and many physicists reserve the term "coherence" to mean that the phase response remains minimum, ie. it does not depart from flat.
I look at some of Vandersteen designs and I am not sure how he could achieve that.
A good measure is "excess phase" approaching zero.
Yes, in the strictest of the definition, that is true. But in reality, I am not sure if any speakers can meet that definition but I could be wrong since I have not seen all the designs out there. The latter Thiel speakers use coax drivers so maybe "time coherence" can be achieved without excess phase on the tweeter as it approaches 20KHz. In the case of Vandersteens, he uses conventional drivers with the tweeters offset from the midrange so I am not sure how he could do that without some amount of excess phase. So it's possible he could be sacrificing "smooth phase transition" at the expense of the tweeter excess phase? I've seen John Atkinson measurement of the CS2.4 step response, and the peak on the step response does suggest that the tweeter response as it approaches 20KHz may not meet that "strictest of definition" as you pointed above. If a speaker meets the "strictest of definition" of "phase coincidence" , then the step response should be "smooth" on top of the step response where there lies the tweeter phase vs. the rest of the frequency (for example below the tweeter cross over point). For example, if you take the phase at the cross-over point to be the reference, the tweeter phase will always deviate from that reference as it approaches 20KHz and therefore the speaker does not meet your claim of "phase coincidence" and therefore by definition, the tweeter response will not be "phase coincident" with the rest of the frequencies. And of course, some speakers are the worse offenders than others with this respect. |
While I commend Stereophile for actually measuring some of the gear they review. The discussion is not really about Stereophile measurements but more about the underlying meaning of "time coherent" or "phase coincidence" (Tom terminology). My argument is a perfect "phase coincidence" is not physically realizable in real world environment irrespective of measurements. I think you can have a speaker that is "phase coincident" but only within a certain frequency range, NOT from 0 - 20KHz. Even with a speaker that only has one driver, it will have different phase at different frequency, and yes it will have a proper step response, but based on Tom strict definition then it is not "phase coincident" at least not at all frequencies. I am not criticizing Thiel products in anyway, it just seems that Tom was using an argument that is a bit unreasonable. Tom, If you look at Stereophile measurement of the Thiel CS3.7 here:
https://www.stereophile.com/content/thiel-cs37-loudspeaker-measurements
You can see there is a rather pronounce peak on top of the step response which means the tweeter phase is not well aligned with the rest of the frequency range, otherwise there would not be such pronounced peak. Again I don't mean to criticize here but more like trying to understand your definition of the term "phase coincidence". You definitely threw a curve ball into the conversation :-) It looks to me the tweeter phase is deviating from the frequency cross over point between the mid and the tweeter. John Atkinson would call this "phase coherent" and I would too. But then it appears that from your above post, you would disagree. If that is the case then it would imply the CS3.7 does not meet your criteria as "time coherent". |
Hi Tom,
Actually I appreciate your timely feedback and I have said many times I have a lot of admiration for Thiel’s products.
I’ve looked at a lot of Stereophile measurements it seems like there are speaker manufacturers that pursued "time phase coherence" but at too much expenses at other parameters that I think are just important. And some are using the term "time phase coherence" too loosely as a marketing gimmick than from actual engineering.
Thanks again for your feedback. I was going to ask you if you would care for another theoretical discussion but my guess is you probably have had enough for one day :-)
PS: As for measuring step response at 8ft, my guess is although the step response will show valid result, but the frequency response may suffer because the tweeter at 8ft distant may exhibit dips or bumps at some frequencies because the phase will not be at same say at 4ft for example. Anyway, just my 2cents. |
I have to eventually ask Tom if all Thiel products are optimized for far field? That is if I measure the frequency response at far field then repeat at near field, should it deviate a lot? At far field, room interactions may hide things that near field will reveal.
I have seen some step responses measured by Stereophile for some of the Vandersteen speakers and John Atkinson has to move the microphone up and down to get it right otherwise the treble response will peak too much (probably due to deviation in phases). But I think it's like cheating. The listener probably will probably not know which way to move to get the best step response. Also it seems like Vandersteens may sacrifice the freq. response too much in order to achieve time phase coherent. For example, the tweeter crosses over the mid at 1KHz which think is pretty low for a tweeter and I wouldn't personally do that just to get time phase coherent. At least I don't think Thiel speakers do that.
I think when you measure at far field such as 8ft, the drivers will in general integrate better because the phase a better aligned at far field vs at near field. For example, if you place a microphone at 1meter at the midrange height, the acoustic distance from the mid to the tweeter will be, let say 4in. Now if you move the microphone at 8ft distant, the relative acoustic difference will be less than 4in. It's like looking at things from far distant and your movement will probably not change the field of vision very much vs. if you're really close. |
I am not sure why but high-end audio is a pretty tough business. I think partly because the younger generation does not care much of "sound quality". Actually it’s hard for me to strike a conversation regarding high-end audio with anyone since I find most people are not really into it either.
For most people as long as there are some music playing from a pair of speakers, they’d be fine. I guess I belong to a dying breed. I tend to be rather particular about my system. |
Oh, and should you still be worried about the build quality of these products then the 10-year (!) warranty should instill confidence.
I got burned by Chinese products in the past. Not sure I can trust them. Hopefully they will be still around ten years from now. |
Careful...spend too much time together and you'll soon be spending lots of money on cables ;-)
and worth it. |
As for some of the comments with regard to tubes driving Thiel speakers, well I have a pair of Cary Six Pacs Monoblocks rated at 50W in triode mode. I tried them to drive the CS2.4, but they just do no have any dynamic and the music lacked any life and vitality. My conclusion is that the Six Pacs do not have enough power to match well with the CS2.4.
On a larger issue, I am not sure why Thiel in general are a bit difficult to drive. In my experience, time-phase coherent speaker designs tend to be a bit less efficient than other more conventional designs but it shouldn't be difficult to drive. Is it because the Thiel driver themselves have very low impedance?
I myself designed a few time-phase coherent speakers (which I hope to share soon), but the impedance curve or the overall efficiency are not that different if I were the do the same but with using more conventional approach. I supposed it's not because of time-phase coherent but something more intrinsic to Thiel specific driver specs. |
With respect to "ease of drive", I looked at the CS1.5 which I believe do not use "Thiel designed" drivers but more or less "off the shelf". Based on Stereophile measurements, the impedance minimum is about ~3.7 Ohm at around 200Hz. Based on that, I assume the drivers used in the CS1.5 are 4 Ohm design. The overall impedance of the CS1.5 is not that different from most 4Ohm speakers using conventional design approach.
For comparison, I ran a quick simulation on my time-phase coherent design, all using 4Ohm drivers:
1. For a 2-way design, minimum impedance is 3.6 Ohm at also 200Hz.
2. For a 3-way design, minimum impedance is 3.3 Ohm at around 160Hz.
But using a more conventional NON time phase coherent methods, the impedance is about the same, therefore my guess is time-phase coherent is not the real reason for being "difficult to drive".
|
Note that the 1.5 drivers were entirely developed by Thiel in-house OK, I see. I just thought they did look similar to some of the Vifa drivers.
I have an eccentric theory about why phase-coherent x time coincident speakers are "harder" to drive
I also notice that if I disconnect the tweeter, the sound becomes subjective less "dynamic". I also notice that higher order filter speakers do have more perceived "slam" vs. first order filter speakers as if the drums become more "dynamic". It may have to do with how high frequency affects our listening. But then beyond this, I am just guessing ...
|
There is a real limitation to dynamics in first order filters. On the one hand, there is no time smear or phase rotation in the crossover range for very realistic immediacy, but on the other hand, the drivers, especially the tweeter, do more work, generate more heat, and experience anomalies coupling to the air beyond their 'natural range'. Thiels are actually small-signal machines which do pretty well in normal rooms at normal volumes. But they are not and were not intended to be screamers.
In interesting with respect to two different design philosophies. After a short exchange with the designer of the "Infinite Slope" of Joseph Audio in another thread, which optimized for frequency domain whereas time-phase optimized for time domain, and "Prof" has listened to both Thiel and Joseph Audio, and he said he like them both. I personally think there is a unique sound with first order that cannot be found in the higher order design. Maybe "Prof" can further describe if there is any intrinsic difference between these two type of speakers. |
The Infinite Slope speakers are very good. His approach sequesters the phase misbehaviors to very narrow bands, so they are noticed very little, and the time misbehaviors are managed for least damage
Richard Modafferi, Infinite Slope
designer, did claim that the group delay within the overlap region is
very minimum therefore, at least within this region, the transition
should not be a problem in the time domain. But as you said,
there is still a problem of phase rotation in which the signal at
20KHz will arrive at 360 degree ahead of the signal at say 300Hz.
But the other problem with steep slope is the rapid change in phase
which will be addressed below.
From what I can gather so far,
the high order suffers the following main two things:
1. As
pointed above, phase rotation where high frequencies will arrive
earlier than low frequencies and how much it depends on how the
filter is designed. It could be 270 or 360 or 180 degree or
somewhere in between . At the same time, I am not sure how
sensitive our hearing is to this type of phase shift. Our hearing is not sensitive to the sound delay if this delay is within a few mlli second. For example, at 17KHz, 360 degree would equate to 0.05 milli second. Could our hearing tell of this small delay? I mean with diffraction from the baffle, I would assume sound delay from baffle reflection could be more than 0.05 from various high frequency components. On the other hand, at 500Hz, 360 degree would be 2 mill second which is somewhat close to our hearing threshold. My conclusion is as at higher and higher frequencies, this type of phase shift may not something our hearing is sensitive to. Thankfully, most xover even using higher order filter, do not have this type of phase shift issue at low frequencies. And considering most xover crosses at 3KHz below which more or less covers most of the musical contents. So regardless of xover, most of this type of phase shift will only occur at above 3KHz where our hearing may not be so sensitive. I have
designed speakers that is first order time-coherent (no phase shift)
and first order NON time-phase coherent. The common denominator here is first order - but one is time coherent but the other in NOT time coherent. The non coherent version has 180 degree phase shift at 20KHz. BUT I had a hard time telling
the difference in the sound quality or the intrinsic type of sound between the two xover types. Although I do prefer the
time-coherent version, but the frequency response and oxver point
between the mid and tweeter are all different between these two
version so I don't if most of the differences come from the
time-coherence or because of different types of voicing. So where do I think the differences come from? I
think most of the differences that we have talked about will be
because of #2 below.
2. All drivers have non-linear distortion
which means the drivers will produce more extra frequencies than the
frequencies putting in at the input. But why would this affect high order
more than first order? I think because of of the rapid phase
shift of steep filter slope, which in turn produces more over-shoot
or more high frequencies energy vs. lower order filter with shallow
phase shift. This extra high frequencies energy when applied to
the non-linear distortion of the drivers, will produce more extra
high frequencies contents that was not part of the original source. This extra high freq. will affect
perception and it is consistent with what I and other (Prof) have
observed that the treble of high order speaker seem to be riding on
to of the music as if it is not part of the music but our mind will
include it as part of the re-construction process. What I said
above also implies that first order filter will suffer from this as
well since it will inevitably have some overshoot (from parasitic such as
drivers coils ...) but not to the same degree. This will also
implies that if quality drivers are used which have low non-linear
distortion, this problem will be minimized even with using high order
filter.
#2 also explains why solid state components cannot produce
treble as well as tubes (I won't try to go into much details), due to
the nature of solid state physics, the electrons movement within the
medium will experience high order filtering affect from parasitic,
impurities and such. With tubes on the other hands, electrons
only have to move in the vacuum from the anode to the cathode without
having to go through any other medium therefore won't be affected by any filtering. And of course solid state amplifier does not have any xover that can explain the differences.
|
When it comes to first-order coherent and non-coherent, there has been some misundertanding with regard to the step response. For non coherent speakers, invariably, the polarity of the different drivers will be different, that is the tweeter may be inverted polarity and the woofer will be positive polarity. Therefore in the step response, the tweeter will "APPEAR" (notice capital letter for emphasis) to go negative with respect to the woofer. But that misses one subtle but very important point that will lead to a crucial insight to the phase of coherent vs. non coherent design.
Most people came familiar with step response measurements by way of Stereophile John Atkinson measurements, but his measurements only show the overall step response without the measurements of each single driver. Once each single response is shown, there is further truth to be gained. I will use one of my project to illustrate the misunderstanding. See pic of speakers here: https://www.flickr.com/photos/185616271@N05/49106259616/in/album-72157711891600612/
Below is the xover design. Notice the tweeter is inverted polarity and the woofer is positive. Also notice the step response on the lower right corner. Also notice the phase of the speaker response on the upper right corner of the plot. That is the phase starts at about 0 degree, and gradually go to -180 degree. Also notice each individual component freq. response and the cross over point. https://www.flickr.com/photos/185616271@N05/49106444352/in/album-72157711891600612/
At first glance, the tweeter seems to go completely opposite of the woofer (the initial dip of the step response). The time-coherent proponent will say – wait a minute, that’s all wrong since how can the treble be completely opposite of the woofer? But that is not quite right. The tweeter ONLY opposite of the woofer at very high frequencies, but at low frequencies, below the xover point, the tweeter and the woofer actually go up together.
I here have a zoom in pic of the step response. You can see the tweeter and woofer go up together after the initial high frequencies. It’s subtle but it’s there and it’s very important. Green – system Red – tweeter Yellow - woofer https://www.flickr.com/photos/185616271@N05/49106444417/in/album-72157711891600612/
What does this mean? If you look at the frequency response below, you see that the phase of the system is only at -180 at 20KHz, not the entire plot. But below the xover point, the tweeter and woofer are more or less “in-phase”. For example, at 2KHz the phase is only at -60 degree, at 3KHz the phase at -80 degree. As I said before, since our perception is not very sensitive at high frequencies, most of us may not “hear the difference”. Most of the musical contents occurr below the xover point which is about 3KHz in this case and that’s where most of the phase shift has not occurred. Also the phase change is not at a constant -180 degree everywhere but only at exactly at 20KHz. For example, the phase change at 20KHz is only a few degree vs. 19KHz and likewise 19KHz is only a few degree vs. 18KHz … and so on. It’s a gradual shift in phase so our hearing may be able to adjust to it. But on the other hand, if this -180 degree occurs at 500Hz, trust me, you will definitely HEAR it! I think high order can get away with it because the phase shift occurs at high freq. where our hearing may not be so sensitive. https://www.flickr.com/photos/185616271@N05/49106492271/in/album-72157711891600612/ So how does an ideal time-phase coherent design look like. Below is a design using ideal perfect driver. Notice how the phase of both the tweeter and woofer tracking each other at 90 degree apart. But the overall phase of the system (the green curve) is at exactly zero. https://www.flickr.com/photos/185616271@N05/49106060213/in/album-72157711891600612/
But real world system does not look that clean and the phase is not going to be exactly at 0 degree. I’ve modified the xover so the tweeter and woofer are all positive polarity, and the phase of the speaker stays more or less within 0 degree (the green curve). Notice the step response where the tweeter and woofer go up together. https://www.flickr.com/photos/185616271@N05/49106502197/in/album-72157711891600612/ |
Phase Coherence is used in various ways by various people for various purposes.
I understand. I am personally got lost with all the different terminologies.
Very simple and very difficult!
Being working on DIY, it usually takes me much more time than a typical high order speaker. But still, I hope more speaker manufacturers will do it. It may be very difficult but not impossible. I think it's just an "excuse" that most people won't try it. I was thinking about how our hearing may not be so sensitive to high frequencies and the high freq. phase shift. In our normal environment, high freq. is very reflective and when sound reflects, its phase will get change as well. Also diffraction will cause delay in high freq. and phase shift as well. If human is sensitive to these, it would drive us crazy. That's why our hearing is only sensitive to long delay, not short delay, and the long delay we hear is what we call "echo". Imagine if we listen to high order speaker and we hear all "echo", I guess that would not be very good :-) But having said all that, I still think time-phase coherent could make a difference, and if I could, my speaker would be all time coherent. |
|
|
Have you heard Ryan Speakers based in Riverside CA? Try to catch an audition and post your impressions here. Similar to a Thiel without being a pure first order filter design.
Thanks for the heads up. But if they are too high-end, I won't be able to afford it :-) |
|
I was playing around with a three-way mock-up. Basically I use a setup with what I already have. Making the cabinets would just take a lot of time. The tweeter is ScanSpeak AirCir, Seas Nextel 5.5in. as mid, and Scanspeak Illuminator 5.5 in as woofer. I don’t current have a dedicate cabinet for 8in. woofer but the freq/phase behavior should not be that different. Measurement was done at 1Meter distant between the tweeter and the mid which is about 33in. high. There are quite a bit of ripples on the freq. response probably due to room reflection which also shown in the step response being taking a bit longer time to settle at the low frequencies. The xover seems a bit complicated, but that’s typical of first order, time-phase coherent, time-coincident speakers. The 50uf cap on the tweeter should only affect the lower freq. so it probably will be OK. Actually I don’t need it there for freq. response, but I put it there to protect the tweeter. Pic: https://www.flickr.com/photos/185616271@N05/49129808921/in/album-72157711891600612/ Xover:
https://www.flickr.com/photos/185616271@N05/49129320988/in/album-72157711891600612/
|
Someone pointed me to Bang & Olufsen Uni-Phase xover design. At least based on theory, the filter will be able to achieve 0 degree phase shift across the frequency band, hence the name Uni-Phase. I have not seen the step response but the square waves all look very good - not too unlike that from a time-phase coherent/time-coincident design. Unlike the Thiel or Vandersteen, though, the Uni-Phase uses 2nd order filters, not first order filters. The advantage is 2nd order does not stress out the drivers as much as first order filter. For reference :
https://www.tonmeister.ca/wordpress/2015/10/29/bo-tech-uni-phase-loudspeakers/
|
|
Playing with speaker position and room acoustic will yield differences that swamp what you’d get switching around digital sources.(And btw, I’m not in the camp that says every digital source sounds the same).
That's missing the point though. Every adjustment will have an affect on different aspects of the sound. Speaker position, room, ... will have some specific on the sound, just as cables, DAC will have their own affects that cannot be found in position ... |
Maybe if Tom could help with this.
I was wondering how the baffle of the Thiel speakers are manufactured? The curved baffle geometry probably presents some difficulties. If I were to do it by hands, I probably would use a combination of a planar and chisel and probably would take me a few weekends, but I assume that would probably be too time consuming and would cost a lot of time and money as well in actual production.
I suppose if you were to mass produce something like that, you could come up with a machine that either can do the entire baffle, or most of it that will reduce most of the manual labor. |
Thank you Tom. The cabinet seems to be the most time consuming part of making a speaker at least in the DIY world. I suppose with CNC equipment then maybe the process will be much faster. Personally for me the cabinet takes about 70% of the time and the rest for the electronics. But even with CNC, I suspect the cabinets will always be the most expensive part of making a speaker, especially with the high end speakers. One of the most understated part of Thiel is in their cabinetry in term of being "furniture friendly" as most people usually talk of Thiel as just "first order coherent". I’ve assumed (or maybe read somewhere) that Thiel designed their speakers to be used with the grills on (as in the sonic effects of the grill were accounted for in the design). I used to listen with the grill off and I notice the upper frequencies were a bit more "open". Also just like most first order design, the sound balance of the Thiel is a bit sensitive to "toe in". I listen with the speakers pointing straight forward. With "toe in", the upper frequencies were slightly more pronounce than I would like. I wonder if the grill only affects the sound pressure to the affect of about "1dB" as Tom said. It’s possible that the grill also affect the "resolution" as well and ultimately the "openess" of the sound. My thinking is that the grill may affect more than just "1dB" less. That is even if you fine tune the treble energy to account for the "1dB", it still may not be the same as with "grill off". There have been some claims with respect to "audible transparent" cloth but I doubt it though, especially where a system has a lot of resolution that will high-light any characteristic of the equipment chain. |
For some reasons, Tom reminds me of this. Oh well maybe I’ll use the stereo illusions ... But I’ll take my gloves off at the challenge. https://m.media-amazon.com/images/M/MV5BNjhiMjk1YWYtMjgyYy00YTFhLTk0NTMtN2Q5MDZjMWEyYWI1XkEyXkFqcGdeQXVyMTQxNzMzNDI@._V1_.jpg It may have to do with the filtering affect of the grille. In a linear system, every frequency passes through the system will only be affected by a constant value, either a gain or an attenuation, which will be applied to all frequencies all equally. This would be ideal or at least preferable. My System Theory 101 is a bit rusty, and I am not sure if a filter can be called a "Linear Time Invariant" (I am pretty sure about the Time Invariant but not sure about the Linear part), but I am pretty sure any filter will affect the phase of a signal and not just the amplitude. In this sense, a grille will act as a filter, a high pass filter albeit a mechanical one, because it will attenuate the high frequencies more than the low frequencies, and just like any filter, therefore introducing a phase shift at the high frequencies. Therefore the attenuation affect of the grille is not a "constant", and probably therefore one cannot be compensated by a "fix" amount on the treble to counteract the variable attenuation of the grille. And I suppose the phase shift introduced by the grille at high frequencies is what affects the "openess" of the sound that I notice. I suppose one could match the treble response to counteract perfectly to the frequencies response of the grille, but that’s probably very difficult and probably not worth the trouble. It’s probably a lot easier to listen just with the grille off :-) Gloves put back. |
Andy2, I’m assuming you mistyped, because a high-pass filter will attenuate the *low* frequencies more than the highs!
OK, I was high on weeds :-)
But the more ’open’ sound without grilles IMO is as much the lack of an acoustic obstruction between you and the tweeter as it is a slight (relative) exaggeration of the highs.
I suppose that is a valid point. I'll let others "bring it on" as Tom will probably have some say to that :-) |
I agree that the CS2.3 baffle is cleaner vs. the CS2.4 in term of "diffraction". I never noticed the difference before. Maybe the motivation for the CS2.4 baffle was about the "aesthetic" appeal. |
Hi Tom,
The other part of the equation with respect to the CS2.3 vs. CS2.4 is evaluating with our listening test. Personally, I don't know if the difference as reported is due to the baffle "edge". The difference could may as well be the xovers, bass driver and so on ...
|
JA - if and as a small company might be created to keep Classic Thiel speakers updated, all models are in the mix.
Hi Tom, Please remember the American moto: "Close but no cigar". We at the end will want "cigar" :-) |
I agree with the ported bass - it's definitely a compromise. Well you can't have it all I guess. Wanting clean bass is not cheap :-) |
Not meaning to turn this thread to "tubes vs. solid state", but I currently have a Conrad Johnson 17LS and Pass Labs XP10, but I prefer the 17LS for music listening, but the XP10 is a really good preamp and in some way actually more refined than the 17LS. |
I have the CJ Premier 16LS2 and love it! Same of course with my other CJ gear.
Been saving for the $25K GAT but will see if my saving or my life will get there first :-) |
|
