With the "dipole or not" discussion, these seems apropos:
http://www.6moons.com/audioreviews/yamamoto9/speaker.html
http://www.6moons.com/audioreviews/yamamoto9/speaker.html
I continue to not understand how anyone can talk about phase response of acoustic signals through crossover by discussing the electrical filter characteristics. This assumes no acoustic phase modification from the drivers, which, for the types of systems being talked about here, is anything but true.
With all due respect, I have seen the Booke's approach and the problem I have with it is that, in a two way, the tweeter is conected with inverted polarity. Such a system can not be coherent because inversion introduces a nonlinear delay of the form Td = 1/(2f). Such an effect is clearly shown in the reproduction of square waves, particularly as the frequency of the square wave approaches the crossover frequency.
Please note, if not clear, that when I refer to phase I specifically mean the net phase of the crossover filter pluse the driver. When I say the phase much match the target through the crossover, the target transfer function is the acoustic output set by T(acoustic) = T(driver) T(filter). The development of the filter transfer function is what ever is necessary to achieve T(acoustic).
And let us not forget that alignment for correct impulse, regardless of what the definition is, is only valid at a single point, usually taken at some distance on a specified design axis.
John
Then I would have to say that you are the exception since most people look only at the filter characteristics of the DSP box or whatever and completely ignore the acoustic effects. When the upper driver is a waveguide then the acoustic effects are a major factor and the passband isn't even electrically flat for CD devices.
I figured that you would understand this, but I find it troubling how many people don't.
I had not gone out of my way to design phase nuetral crossovers as I am not a big believer in phase audibility, but then as I worked through the crossover to yield the best polar response, i.e. both on and off axis, I found that this did in fact yield the lowest group delay in the impulse response and hence the most linear phase system response. If this is coincidence or not I don't know, but it only makes sense that a minimizing the phase difference between two drivers at the crossover - from whatever causes - will yield the best polar response.
I understand. I see publications on the web and in journals about crossovers and then see the authors going to length showing how the transfer functions can be constructed either passively, or actively, analog or digital and it's all pretty moot unless the drivers are dead flat to start with.
Anyway, with todays DSP it seems to me that the best approach is to design a system to have good polar response and the desired frequency amplitude response. If symmetric lobing is what is desired then the LR crossovers are the way to go. Then the system impulse can be corrected using DSP. I'm actually working with someone on such a projects right now. Here are some preliminary results using my NaO Mini, panels as a test mule. The panels response is 100 Hz LR4 high pass, LR4 2k Hz mid/tweeter x-o. The system amplitude remains unaltered after the DSP preprocessing.
System Phase, before and after DSP linearization:
System Impulse, before and after:
System 1k Hz square wave, before and after:
With a little adjustment to the phase linearizing DSP I can make the square waves even better. I just don't have a picture of the fine tuned result.
Do I take this to mean you are referring to using LR filters over BW filters? This is a recent development?
I'd like to back this approach, I think it is the best design strategy.
Not having to regard the total system phase response offers a degree of freedom for the designer (no need to compromise other things in favor of total system phase), as it can be corrected afterwards, with zero side effects, to minimum phase (of the overall system bandpass) or even quasi-linear phase near the lower passband corner and below. Of couse the correction only applies a smooth rollback of the overall allpass-phase, minute ripples in the response remain and need to remain to get a correction filter convolution kernel with zero side-effects (no ringing). Just like in your graph, which looks pretty much like what I got.
I did this (the quasi-linphase thing, that is) with my two sets of speakers which are fed from DSP-preprocessed signals (PC is the sound source) and once you get used to the effects, there is absolutely no way back, uncorrected they just sound lame, bass lags behind and has less punch as well as the depth of the soundstage decreases noticably. Also I found absolute polarity to be audible for the first time after this change.
All of this are partly due to the speaker's construction which both have lots of group delay in the mids/lows. With better speakers in this regard the effects may be less prominent.
- Klaus
Last edited:
The problem I have with the omni-top isn't the horn per-se, but rather that the horn mouth and/or flair isn't large enough *not* to be audibly noticed.
Moreover not only do you have all the additional problems with cabinet vibration, but you also *hear* those idiosyncrasies far more with the horn design, including the quality of the driver (..which may not have anything detectable with modern measurements). While you can hear what they do right, after a while you also can't hear anything but what they do wrong. (..rather like lowthers.)
The driver suspension in particular starts to become critical, and NOS drivers start having a significant advantage. I *also* have a problem with any design that puts a bass reflex that near the average that high up in freq.. Delay/Phase goes to "pot" with these designs and damping factor actually is noticeable. With horns like these there is no free-lunch here, in fact it's where I started, but finally gave-up on in frustration.(..just trying to save you some agony..
)
I think the A7 is similar to the more modern Exemplar bass-bin (that was based on the A7), i.e. around 27"W x 33"H for a single 15" driver. That will get a slightly lower "cut-off" (especially with the floor coupling like the Exemplars), but these designs were at best pushing 102 db near the average with a lower tunning freq. for the vent (when compared to the omni-top), and with the bass region significantly below the average (even considering room-gain).
I remember first seeing the Exemplars and thinking: "yeah, that's probably about the best you can do for a front-loaded mid-bass horn in a domestic setting that doesn't use the rooms walls". (..the La Scala's and Klipshorns sound wrong in the upper mid.s to me.) The one design that goes beyond domestically acceptable, (but actually does extend below the lowest voice range), is the Living Voice Air Partner.. but not to many want a pair of SubZero-sized horns in their living room.
I'm not a fan of the eXemplars, unfortunately - to my ear, they were/are very bass-thin, very forward midrange, and I don't care for the HF horn at all. Well-restored Altec A5 & A7 theater speakers are a mixed bag - hints of greatness shine through, but there are a lot of heavy vintage colorations as well. I see them as technically interesting starting points, but nothing I'd want to replicate as-is. They do need the theater "wings" to sound correctly balanced, and they are unlistenable closer than 3 meters distance.
Of this genre of speaker (shorthorn with reflex loading & HF horn), the most successful example I've heard were Christian Rintelen's Blue Thunders (in Zurich). They were very good indeed - and Altec 515 bass and the best TAD large-format compression drivers didn't hurt. The Blue Thunders are superb speakers with no real commercial equivalent.
I am particularly interested in what you hear with this type of cabinet (as I joked earlier, I can't hear what other people hear, so I have to ask them - I might hear the same things, but I might not). I expect the horn to "magnify" the character of the driver, faults and all - they have a way of doing that, Lowthers, AERs, and Feastrex included.
If I read your post correctly, you hear artifacts created by too-small mouth size and/or incorrect flare rate, along with group-delay artifacts from the relatively high fourth-order cutoff of the vented box. What do these artifacts sound like (to you)? I'm being serious here, not joking around.
I'm not very sensitive to delay artifacts - the few times they've been drawn to my attention, they sound like rather subtle timbral shifts, while other more tuned-in listeners are driven from the room. I do notice driver and cabinet artifacts pretty quickly, and don't care for them at all; with real-world systems, it comes down to accepting colorations I can tolerate, instead of aiming for no coloration at all (which I have never heard from any hifi system, at any price point).
The catch is that everyone I meet seems to have different perceptions of colorations, so a recommendation for XYZ driver or cabinet only makes sense if I'm met the person, heard their own hifi system, and have some idea of their tastes. Then I can get some vague impression of what they're talking about if they say ABC or XYZ sounds like this or that. If they have different tastes than mine, of if their hifi sounds really weird to me (by audition), then they can be great friends and all, but I can't reliably assess their subjective comments in a useful way. Of the people I know, only a few (John Atwood, Gary Pimm, and Gary Dahl) seem to share similar tastes - everyone else seems to be listening for other things - not good, not bad, just different.
This is a serious question, ScottG: What artifacts and/or wrongness do you hear from given examples of bass shorthorns with reflex loading? Hangover? Cabinet colorations - woodiness, thickness in the bass range, that kind of thing? Colorations with male voice in choral works? Heavy or slow-sounding percussion from drums? A general impression of artificiality, or a flattening-out of tone colors? I'd like to know.
Last edited:
I remember what I read in Sound Practices about the Blue Thunder is that the woofer is a Focal unit. And it's mentioned in the article that the Focal was chosen because it needed less volume to reach the LF goal. (or did that change after the pubilshing?)
Such a short MB horn+ported cabinet design in my eyes is a too obvious compromise in size. It controls and reinforces where not very much needed while forgoes the major drawback. Maybe it'd getting right when reaching the scale of A4 or A2. Or maybe they are good in theater (PA) but a no no in homeuse. I've never heard any A4 or A2, so I'm not sure. A5 impressed me very much in the very beginning but soon I couldn't stand the tonal balance (or unbalance)....
Such a short MB horn+ported cabinet design in my eyes is a too obvious compromise in size. It controls and reinforces where not very much needed while forgoes the major drawback. Maybe it'd getting right when reaching the scale of A4 or A2. Or maybe they are good in theater (PA) but a no no in homeuse. I've never heard any A4 or A2, so I'm not sure. A5 impressed me very much in the very beginning but soon I couldn't stand the tonal balance (or unbalance)....
No, no, no. I hope that is not the case. That's why I was asking for a practical guide to time and phase alignment. It has to include the drivers, of course! The final result is what matters. Use whatever electrical slope you need to achieve the desire acoustic result.
Learning what techniques and tools other use to achieve this goal would be a big help to the DIY community.
Hello,
For the 4 quasioptimal crossovers mentionned (which are, apart the 6dB optimal crossover, the only crossovers that can be build conventionally, with passive elements by example), the HF and the LF loudspeakers (2 ways considered here) operate quasi in phase at all frequencies below 4kHz.
The small distortion you mentionned on square waves is only related to the part of the spectrum of the signal above 4kHz and, as I said previously, is inaudible (psychoacoustics). Specially, all who use those crossovers can testify that the small distortion you can see on square waves, is not assiocated to a lesser impact on transient of instruments. (And, no musical instruments I know delivers a transient equivalent to the transient of a square wave, even if that one is bandwith limited to 20Hz-20kHz).
For people refusing DSP in their audio systems those quasioptimal crossovers are the best solution at the moment.
I have to add that because thier "in coincidence curve" is very flat and nearly overimposed to the response curve (this curve is equivalent to the peaking curve rarely mentionned in RANE technical notes), the quasioptimal crossovers allow a perfect axial lobing in all the frequency range.
Best regards from Paris, France
Jean-Michel Le Cléac'h
For the 4 quasioptimal crossovers mentionned (which are, apart the 6dB optimal crossover, the only crossovers that can be build conventionally, with passive elements by example), the HF and the LF loudspeakers (2 ways considered here) operate quasi in phase at all frequencies below 4kHz.
The small distortion you mentionned on square waves is only related to the part of the spectrum of the signal above 4kHz and, as I said previously, is inaudible (psychoacoustics). Specially, all who use those crossovers can testify that the small distortion you can see on square waves, is not assiocated to a lesser impact on transient of instruments. (And, no musical instruments I know delivers a transient equivalent to the transient of a square wave, even if that one is bandwith limited to 20Hz-20kHz).
For people refusing DSP in their audio systems those quasioptimal crossovers are the best solution at the moment.
I have to add that because thier "in coincidence curve" is very flat and nearly overimposed to the response curve (this curve is equivalent to the peaking curve rarely mentionned in RANE technical notes), the quasioptimal crossovers allow a perfect axial lobing in all the frequency range.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
I would like to add that when two drivers are used in inverted polarity, the effects are very clearly audible with recordings using simple mic techniques and mixing, and drivers that decay quickly.
Hello Soongsc
You are right if a classical crossover is used (Butterworth 3rd order, Linkwitz-Riley...), a polarity reversal of one of the loudspeaker is perfecly audible.
But you have probably never experienced a quasioptimal crossover which allows:
1) both an excellent "in phase" operation of the 2 loudspeakers below 4kHz (2 ways example considered here for conveniency) like the Linkwitz-Riley crossover does (even if the LR2 requires an inverted polarity of one of the driver)
2) a very low phase distortion (unlike the LR which has a very bad phase distortion)
You'll find more information on one of those quasioptimal crossover here:
http://www.diyaudio.com/forums/showthread.php?threadid=121175
See also:
http://freerider.dyndns.org/anlage/LeCleach.htm
Best regards from Paris France,
Jean-Michel Le Cleac'h
You are right if a classical crossover is used (Butterworth 3rd order, Linkwitz-Riley...), a polarity reversal of one of the loudspeaker is perfecly audible.
But you have probably never experienced a quasioptimal crossover which allows:
1) both an excellent "in phase" operation of the 2 loudspeakers below 4kHz (2 ways example considered here for conveniency) like the Linkwitz-Riley crossover does (even if the LR2 requires an inverted polarity of one of the driver)
2) a very low phase distortion (unlike the LR which has a very bad phase distortion)
You'll find more information on one of those quasioptimal crossover here:
http://www.diyaudio.com/forums/showthread.php?threadid=121175
See also:
http://freerider.dyndns.org/anlage/LeCleach.htm
Best regards from Paris France,
Jean-Michel Le Cleac'h
Not true. There are artifacts. For example, consider an LR4 crossover, neglecting any small ripples in the response. The HP and LP sections are in phase at all frequencies and the all pass correction with makes the system linear phase also makes the individual HP and LP section linear phase as well, with the subsequent ringing associated with linear phase filters.
Therein lies the real challenge, not the DSP to flatten the response along some axis.
Hello Jean-Michel,
It seems these are just talking about stand alone filters. How does this relate to actual speaker drivers being used in different polarities?
Let me summarize what I see. You have a crossover with for which first arrival times are approximately aligned, and which has symmetric polar response. Now that is all good. But that is all it is. It is not phase coherent or transient actuate. For the case you are discussing the crossover frequency (in the power point presentation) is 1 K Hz. The HP and LP in the filters you mention are not in phase at all frequency below 4k Hz. They are not in phase anywhere but in the immediate region of the crossover frequency where the offset delay aligns the phase of the HP section with the LP section.
An externally hosted image should be here but it was not working when we last tested it.
The part of the spectrum you claim to be in error, above 4k is also misleading because the presentation looks at a 125 Hz square wave for which the majority of harmonic content is below the crossover point. Move the crossover point lower, as would be required for a 3-way, or the square wave frequency higher, and the distortion become clearer.
The crossover can not be transient accurate because one driver must be connected with inverted polarity. Again, this, aside from the fact that the attack of any transient is in the wrong direction, introduces a NONLINEAR, frequency dependent delay of the form 1/f. Note, 1/f behavior is not linear, though it is stated as such in the power point presentation.
I don't know particularly which 4 crossovers make up your grouping, but back in the 70's SPICA used a related approach of physical offset to time align their speakers. Their approach does (did) not require inverting polarity of any drivers and, while again, not perfectly linear phase, was closer than what you have. Granted, the Spica crossover approach does not provide symmetric polar response, but it can be constructed passively.
I accept that your crossover approach may be useful for improving the characteristics of 2-ways systems but it is not appropriate for 3-ways.
I guess I am a bit of a perfectionist, but I figure if I am driving down the highway, even if I am going the same speed as other cars and in the same direction, if I'm driving back wards I don't have it right.
Additionally, a 3rd order, quasi Linkwitz type crossover is easily constructed by cascading a B2 with a B1 filter and delaying the high pass section by 0.25 wave lengths at the crossover frequency. The response error will be +0, -1.2dB max. This is easily constructed using passive elements.
For the HP,
C1 = 1/(4 Pi R F), C2 = 1.2 /(Pi R F), L = 1/(2.4 Pi R F)
For the LP
L1 = R/(Pi F), L2 = R/(4.8 Pi F), C = 0.6 /(Pi R F)
Offset tweeter by 0.25 wave lengths at the x-o frequency and connect with inverted phase. I don't know if this is in your "4" or not.
In any event, there is another problem with using offset to time align. Since the acoustic centers are purposely misaligned and do not lie in, or close to in the same plane, moving off axis in the horizontal direction causes loss of the alignment.
Last edited:
Hello Soongsc
Better to not mix 3 different problems.
1) The trivial problem of the wrong polarity due to a negligent people.
2) The trivial problem problem of the absolute polarity
3) The question: is it possible to reach a quasi linear phase over the widest audible frequencies interval without using FIR and DSP.
I don't think this is a place to adress the problem 1) and 2). Better to concentrate on 3)
John has correctly written that using conventional (causal) crossovers only the 6dB/octave can reach the optimal solution. ( note that people using 6dB/octave crossovers don't pay enough attention on a possible polarity reversal of one of the polarity of the loudspeakers. We have to remind that for the 6dB/octave crossover if the 2 drivers are in phase opposition, the frequency response curve is still flat, with no hole. But if one loudspeaker has a polarity inversion then the phase is no more linear as it is with the 2 loudspeakers having the same polarity)
I don't want to speak about digital linear phase crossovers here. They are theorically perfect but few people report hearing the preondulation of the filters used (don't forget that our hearing is causal)
Several people (John, Francis Brooke, Samuel Harschn myself,...) are trying to overcome the weakness of "classical" crossovers ( weakness of the phase linearity of the LR crossover is an example. Bad dephased operation of the 2 loudspeakers when using the classical Butterworth 3rd order is another...). They brought solutions that have each of them advantages and their weakness but which are a true ameliorations of existing crossovers.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Best regards from
Better to not mix 3 different problems.
1) The trivial problem of the wrong polarity due to a negligent people.
2) The trivial problem problem of the absolute polarity
3) The question: is it possible to reach a quasi linear phase over the widest audible frequencies interval without using FIR and DSP.
I don't think this is a place to adress the problem 1) and 2). Better to concentrate on 3)
John has correctly written that using conventional (causal) crossovers only the 6dB/octave can reach the optimal solution. ( note that people using 6dB/octave crossovers don't pay enough attention on a possible polarity reversal of one of the polarity of the loudspeakers. We have to remind that for the 6dB/octave crossover if the 2 drivers are in phase opposition, the frequency response curve is still flat, with no hole. But if one loudspeaker has a polarity inversion then the phase is no more linear as it is with the 2 loudspeakers having the same polarity)
I don't want to speak about digital linear phase crossovers here. They are theorically perfect but few people report hearing the preondulation of the filters used (don't forget that our hearing is causal)
Several people (John, Francis Brooke, Samuel Harschn myself,...) are trying to overcome the weakness of "classical" crossovers ( weakness of the phase linearity of the LR crossover is an example. Bad dephased operation of the 2 loudspeakers when using the classical Butterworth 3rd order is another...). They brought solutions that have each of them advantages and their weakness but which are a true ameliorations of existing crossovers.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Best regards from