aaand the proof of that is... where???My take-away from that table is that "the price to be paid" for the Purifi's remarkable Xmax and Fs is that as a system, it will not be as agile in starting and stopping as is the Accuton (or the Eton)--....
Apparently, you didn't read a single science paper (AES, IEEE...) in a whole your life. In all those published science papers names of the DUT are named as A, B, C, etc.BTW, the reason our paper was published in a high-end audio magazine, and not somewhere you might think it more likely to be published, is that other outlets apparently were leery of naming names of the various drivers in Table 1.
Lousy excuse! All you need is: free ARTA software, cheap microphone capsule costing $3 (which doesn't need to be calibrated) and a PC sound card.I don't have the time, money, or the equipment (poor poor pitiful me, Dear Lord, Don't buy me a Mercedes Benz, buy me a Klippel NFS--with due respect to Janice Joplin) to make measurements of all those drivers.
Ahw, come on!
The (low-frequency) operation of a dynamic driver is described by the conventional Thielle-Small paramers. As far as I can tell, the newly proposed "Load Factor" is defined as LF = Mms/Sd, so it's just a simple combination of the existing TSP (Mms and Sd). In other words, it does not extend on the set of the existing Thielle-Small parameters in any way. LF is not an addition to the existing set of Thielle-Small parameters.
The same can be said of the "Agility Factor" (AF).
(Note that I am not saying that it's not interesting or useful to look at LF or AF.)
The (low-frequency) operation of a dynamic driver is described by the conventional Thielle-Small paramers. As far as I can tell, the newly proposed "Load Factor" is defined as LF = Mms/Sd, so it's just a simple combination of the existing TSP (Mms and Sd). In other words, it does not extend on the set of the existing Thielle-Small parameters in any way. LF is not an addition to the existing set of Thielle-Small parameters.
The same can be said of the "Agility Factor" (AF).
(Note that I am not saying that it's not interesting or useful to look at LF or AF.)
Multiple Entry Horn, I'm sure that'll ring a bell.A TLA i am not familiar with.
dave
@mterbekke: That is then a misunderstanding, my bad. I was referring to the idea that woofers can start and stop. No physical system can start and stop in the sense that it is meant here. It starts moving under acceleration which depends on it's mass, gradually getting going or slowing down to a stop.
Furthermore, there are no start and stop signals in music. It's all band-limited.
Many think of transients as suddenly starting and stopping signals, but you can't start and stop suddenly in a band-limited system.
Jan
Furthermore, there are no start and stop signals in music. It's all band-limited.
Many think of transients as suddenly starting and stopping signals, but you can't start and stop suddenly in a band-limited system.
Jan
Agree, then again, iirc it was said it could be added to the pdf, as is done on sites with EBP. It's just a rough estimate and helps in filtering, nothing more, nothing less.Ahw, come on!
The (low-frequency) operation of a dynamic driver is described by the conventional Thielle-Small paramers. As far as I can tell, the newly proposed "Load Factor" is defined as LF = Mms/Sd, so it's just a simple combination of the existing TSP (Mms and Sd). In other words, it does not extend on the set of the existing Thielle-Small parameters in any way. LF is not an addition to the existing set of Thielle-Small parameters.
The same can be said of the "Agility Factor" (AF).
(Note that I am not saying that it's not interesting or useful to look at LF or AF.)
the load factor (Mms/Sd) has some meaning: it indicates how long a port needs to be relatively. As Thorpe and Small showed: there is a mathematical relationship between box volume, bass extension F3 and sensitivity. pick any two and the 3rd is given. A driver that plays deep in a small box as is used nowadays (unlike retro style speakers ) needs to have high Mms and thereby high LF. This requires a very long port and this is why a passive radiator works better for such small boxes."B x l," also known as "Force Factor," is one of the canonical Thiele-Small Parameters.
As explained in the linked-to paper, Jim Tuomy and I believed that a new Parameter, complementary to and analogous to Force Factor, would be helpful. We call that new Parameter "Load Factor."
Then, when one expresses the relationship between Force Factor and Load Factor, we call that result the Agility Factor.
Load Factor and Agility Factor are “Quick and Dirty” methods to quantify what an experienced designer might in any case be able to intuit from scanning the data sheets on two different drivers.
No more; but certainly nothing less.
A comparison to the well-accepted parameter Force Factor (B x l) illuminates. Force Factor has no need of complexity. Force Factor has no need of calculus, the symbol Pi, or constants. It would be very difficult to have fewer “moving parts” than does Force Factor.
Force Factor relates a physical characteristic (voice-coil length) to an electromagnetic phenomenon (flux density) by the simple arithmetic operation of multiplication.
Load Factor is analogous and complementary to Force Factor. Load Factor is merely an expression of density rather than mass.
In much the same way as Force Factor is derived, Agility Factor relates a physical characteristic (cone density) to an electromagnetic phenomenon (Force Factor). Except, in this case, the simple arithmetic operation is division.
Looking at the Table of drivers under consideration, it is almost impossible to escape the conclusion that the woofer-mid with an Agility Factor of 71.96 is more Agile (or, more responsive; or, just plain better at starting and stopping) than the woofer-mid with an Agility Factor of 26.78.
We believe that the burden is on the naysayers to prove that the above conclusion is illusory.
https://positive-feedback.com/audio...proposed-additions-to-the-thiele-small-canon/
john
The agility factor is pretty much a proxy for the sensitivity Bl/Mms•Sd and a driver that plays deep in a small box has low sensitivity and thereby low AG. The sensitivity says nothing about bandwidth/speed so the agility factor name is really misleading. Eg a diamond dome tweeter is heavy but has despite that very high bandwidth (fast).
Amazing that Thiele and Smalls work is forgotten.
cheers
Lars from Purifi
exactly. to further elaborate: https://purifi-audio.com/2019/12/10/a-fast-driver-needs-a-light-cone-or-does-it/@mterbekke: That is then a misunderstanding, my bad. I was referring to the idea that woofers can start and stop. No physical system can start and stop in the sense that it is meant here. It starts moving under acceleration which depends on it's mass, gradually getting going or slowing down to a stop.
Furthermore, there are no start and stop signals in music. It's all band-limited.
Many think of transients as suddenly starting and stopping signals, but you can't start and stop suddenly in a band-limited system.
Jan
@jan.didden I was a bit direct, to try and get this thread not going off the rails, sorry if that meant I came across pushing a bit.@mterbekke: That is then a misunderstanding, my bad. I was referring to the idea that woofers can start and stop. No physical system can start and stop in the sense that it is meant here. It starts moving under acceleration which depends on it's mass, gradually getting going or slowing down to a stop.
Furthermore, there are no start and stop signals in music. It's all band-limited.
Many think of transients as suddenly starting and stopping signals, but you can't start and stop suddenly in a band-limited system.
Jan
Yeah, that's true, but not entirely sure about the not starting signals part. Most signals die out gradually for sure, but a kick drum, timpany or equivalent is fast in it's attack. The reason I doubt, is because if you listen to these instruments live and compare them with the recording (I have done this with our local NNO Orchestra a few times), the playback of the recording comes across as slow and heavy. This is independent of the recording (analog/digital). I know this is subjective and purely speculative, but it's mind blowing how many systems don't do this right, and the ones that do, seem off in frequency range. Most get better by DSP, or have big and light cones (western electric etc), which my technical mind has trouble accepting (bottom end missing, expensive, rare etc). Either way, the experience with light and big speaks volume, and until I can understand it (which I don't), I seem to have to restrict myself in underestimating BL, weight of the cone, Fs and Sd versus Frequency Range. By that I don't mean I disagree, it's just that I don't know where the limits should be exactly. It could be just group delay, haven't checked. And for one other suspect cause, directivity, I have no idea how to test that scientifically. Meaning: maybe in-room FR and directivity play a game where it "deconstructs" the signal to that extent, that the speed of the kick etc seems lost. Like Gedlee said a few weeks back, there should be more research done in that area. I couldn't agree more.
Unity horn, Multiple Entry Horn, Synergy Horn, afaik they're all the same, could be just a naming scheme because of patents, but yes!\
Like a synergy?
dave
Can you explain post #21 in light of no difference between mass of cones and it's "speed", in this case bandwidth.exactly. to further elaborate: https://purifi-audio.com/2019/12/10/a-fast-driver-needs-a-light-cone-or-does-it/
Because I can't get around a restriction to the statement. E.g. BL changed otherwise bandwidth (and sensitivity) will in fact change. I'm not sure where to draw the line between a woofer and a compression driver, even though TS don't apply to compression drivers, yet no idea why the same rules wouldn't apply for mass, force etc.
Curious.
The misconception is the blanket statement all woofers have the same bandwidth. If you explain where those woofers are I'll send you the popcorn.
What many people don't seem to understand, and why "bandwidth" is important, that is is the driver AND THE CROSSOVER that determine the bandwidth. More specifically, it is the highest frequency in the passband that limits how fast the driver will response to an input. What is laughable is that people go to all these extraordinary lengths to say why a low Mms, high BL driver will be "fast" (and therefore better I guess) but then seem to have no clue that when they slap a low-pass crossover onto it they have just handicapped the driver's response "speed" significantly. Maybe you know this important point already, but I suspect not.
So, generally speaking, two drivers that have the same low-pass crossover point in a loudspeaker application will have the same "speed".
I always had the idea that it was a low in-cabinet Q that gave people the idea that the driver was "fast" due to the somewhat thin sound due to a rolled-off low end.
P.S. I recall some analysis a couple of decades ago published by Adire audio about "fast bass myths" that concluded that it was the inductance that was limited "speed", not other factors like Bl, etc. I will see if I can find that. Ah, yes, here you go:
https://www.docdroid.net/WRLDjtx/adire-audio-woofer-speed-by-dan-wiggins-pdf
Last edited:
That is ironic !! 😀
I had that very same discussion with a a fellow DIYer quite a few years ago, as the question was should I use a driver of the lowest Q..........
And I said pick a driver that has the lowest THD with the lowest Q you can of course but even if the Q is a bit higher than what you want in the first place go with the lowest THD. 😉
As it was for an OB design I told him by the time you add all of the extra EQ'ing now to get the thing to match up to an ESL panel and you music to make it flat or what ever, you just destroyed all that effort of selecting and paying the extra price for an ultra low Q driver in the long run!
Now, nevertheless there is No question the system sounded great however, but I never did find out what the final selection was or how it was set up in the end, it was all built with quality components to begin with. 🙂
Now considering if you are using active filtering it is not so bad as the amp still has ample dampening control over the driver but not as much so with a Passive Crossover.
FWIW
jer 🙂
I had that very same discussion with a a fellow DIYer quite a few years ago, as the question was should I use a driver of the lowest Q..........
And I said pick a driver that has the lowest THD with the lowest Q you can of course but even if the Q is a bit higher than what you want in the first place go with the lowest THD. 😉
As it was for an OB design I told him by the time you add all of the extra EQ'ing now to get the thing to match up to an ESL panel and you music to make it flat or what ever, you just destroyed all that effort of selecting and paying the extra price for an ultra low Q driver in the long run!
Now, nevertheless there is No question the system sounded great however, but I never did find out what the final selection was or how it was set up in the end, it was all built with quality components to begin with. 🙂
Now considering if you are using active filtering it is not so bad as the amp still has ample dampening control over the driver but not as much so with a Passive Crossover.
FWIW
jer 🙂
Dear Lars,the load factor (Mms/Sd) has some meaning: it indicates how long a port needs to be relatively. As Thorpe and Small showed: there is a mathematical relationship between box volume, bass extension F3 and sensitivity. pick any two and the 3rd is given. A driver that plays deep in a small box as is used nowadays (unlike retro style speakers ) needs to have high Mms and thereby high LF. This requires a very long port and this is why a passive radiator works better for such small boxes.
The agility factor is pretty much a proxy for the sensitivity Bl/Mms•Sd and a driver that plays deep in a small box has low sensitivity and thereby low AG. The sensitivity says nothing about bandwidth/speed so the agility factor name is really misleading. Eg a diamond dome tweeter is heavy but has despite that very high bandwidth (fast).
Amazing that Thiele and Smalls work is forgotten.
cheers
Lars from Purifi
Your statement "pick any two and the 3rd is given" strangely enough made me think of "Hofmann's Iron Law."
Had you ever encountered it--that's going back a very long way.
https://en.wikipedia.org/wiki/Josef_Anton_Hofmann
I was gobsmacked to learn that the audio engineer JA Hofmann, of loudspeaker "Iron Law" fame, was the son of the legendary Polish concert pianist Josef Hofmann, who was the only private piano student of Anton Rubinstein, the founder of the St. Petersburg Conservatory, and therefore founder of the Russian school of pianism. That was in 1862!!!!!
Just a few weeks ago I had lunch with Winslow Burhoe, who was Edgar Villchur's lab assistant at Acoustic Research in the late 1950s and early 1960s. Winslow pretty much designed the AR4a on his own--Villchur was off doing other things. Winslow later invented the inverted-dome tweeter. He also invented the signal processing for the Zvox sound bar. I am sure Winslow had crossed paths with JA Hofmann, and perhaps worked with him at KLH, where Winslow later worked before founding his own companies.
Lars, thank you for your thoughtful contribution.
I have never claimed to be an engineer. My day job is as a classical-music record producers. I prototype loudspeakers as a sideline.
Strangely enough, two weeks ago I was recording a man who was a student of Nadia Reisenberg, Josef Hofmann's teaching assistant at the Curtis Institute:
That's iPhone audio, but the venue is Mechanics Hall (Worcester Massachusetts) and that's a relatively new 9-foot Hamburg Steinway.
It's brief, but I hope you enjoy it.
john
...Winslow Burhoe...
I was getting into hifi seriously just when Winslow’s EPIs were hitting the local shop. Myself and my friend sowned many. And Scott & i have a couple tribute designs published.
What the point is, is that Winslow did not use T/S parameters, he went right to the real parameters they are based on.
dave
You're missing the point a bit Charlie, and that is that, all things being equal (if there could be such a thing) a high Mms, low BL driver will likely be a worse driver. Bar exceptions. Unless you want to build closed cabinets with low efficiency, that's fine. And of course I'm not saying a driver whith a Q of 0,0001 should be the goal, but bigger drivers get more energy in the room, standard physics, it depends on personal preference as well e.g. at what level is being played. That wish doesn't result in small woofers, so that's as far as "better" goes. Then again, with small woofers you're heavily restricted in SPL, which you don't have to like, but it leaves out having the option to feel a drum, so in reality, small low efficiency woofers are less by design, whichever qts or mms, bl they might have..What many people don't seem to understand, and why "bandwidth" is important, that is is the driver AND THE CROSSOVER that determine the bandwidth. More specifically, it is the highest frequency in the passband that limits how fast the driver will response to an input. What is laughable is that people go to all these extraordinary lengths to say why a low Mms, high BL driver will be "fast" (and therefore better I guess) but then seem to have no clue that when they slap a low-pass crossover onto it they have just handicapped the driver's response "speed" significantly. Maybe you know this important point already, but I suspect not.
So, generally speaking, two drivers that have the same low-pass crossover point in a loudspeaker application will have the same "speed".
I always had the idea that it was a low in-cabinet Q that gave people the idea that the driver was "fast" due to the somewhat thin sound due to a rolled-off low end.
P.S. I recall some analysis a couple of decades ago published by Adire audio about "fast bass myths" that concluded that it was the inductance that was limited "speed", not other factors like Bl, etc. I will see if I can find that. Ah, yes, here you go:
https://www.docdroid.net/WRLDjtx/adire-audio-woofer-speed-by-dan-wiggins-pdf
I'm not sure at all about that docdroid link. Am I supposed to deduce speed of that driver was scientifically proven to give better SQ with the missing of the resonances in the Puls response figure with the added 28 grams of weight? I look at speed as a driver that can follow the amplifiers' signal. How does that added mass do that better than the raw one? Did you just choose to not see that big hump a few dB above the original, coming back from the negative pulse? How is that not slower? It's total energy is off more than the others, this generates more questions, not answer them imho. If speed was important, the impulse response with added mass did show resonances that the induction and raw driver both didn't have, yet was lower in level halfway through. What to make of that, a slower driver, more mass to be less bothered by internal or external air disturbances, feedback of some sort? No idea.
I do see the induction having slightly earlier roll off, but I don't see how any of that is a discussion at all. First of all, pulse following of speakers are highly depending on impedance, mass and frequency tested. It maybe more depending on the Rdc of the induction used, or the amp and cables driving it, or the sum total. By which I mean that we can wave our hands about it, call me thick, but I don't see proof that the added mass gives comparable signal as one without. Or are we all going along with the idea pulses don't exist (at all, maybe sometimes, not enough to be significant?) in music, because then we'll have to agree on what impulses are part of it. We sure don't listen to sine waves.
Let's see if we can get a clear picture about the pulse response and why that extra mass is or isn't a problem. I do think it's a thing that is overlooked, because it doesn't show up significantly in FR with woofers, is that a thing we can't enhance because drivers' Re is generally too high, DG too low, BL too low, what exactly?
Winslow is an amazing guy--and still designing.I was getting into hifi seriously just when Winslow’s EPIs were hitting the local shop. Myself and my friend sowned many. And Scott & i have a couple tribute designs published.
What the point is, is that Winslow did not use T/S parameters, he went right to the real parameters they are based on.
dave
He's one of the few people I have known who can operate at the frontiers of higher math--such as Quaternions.
You might enjoy this, if you have not seen it already:
https://www.stereophile.com/content/direct-acoustics-silent-speaker-ii
John Atkinson is a hard guy to impress--but just look at how close the Silent Speaker II's in-room midrange response is to JA's personal reference pair of BBC LS3/5As! JA concluded that Winslow's budget speaker was "greater than the sum of its parts."
ciao,
john
help John, thank you very much for all this. I was of course referring to Hofmanns Iron Law but my knowledge of him did not go beyond the Wiki article that I have read earlier. The link to Rubinstein is amazing!Dear Lars,
Your statement "pick any two and the 3rd is given" strangely enough made me think of "Hofmann's Iron Law."
Had you ever encountered it--that's going back a very long way.
https://en.wikipedia.org/wiki/Josef_Anton_Hofmann
I was gobsmacked to learn that the audio engineer JA Hofmann, of loudspeaker "Iron Law" fame, was the son of the legendary Polish concert pianist Josef Hofmann, who was the only private piano student of Anton Rubinstein, the founder of the St. Petersburg Conservatory, and therefore founder of the Russian school of pianism. That was in 1862!!!!!
Just a few weeks ago I had lunch with Winslow Burhoe, who was Edgar Villchur's lab assistant at Acoustic Research in the late 1950s and early 1960s. Winslow pretty much designed the AR4a on his own--Villchur was off doing other things. Winslow later invented the inverted-dome tweeter. He also invented the signal processing for the Zvox sound bar. I am sure Winslow had crossed paths with JA Hofmann, and perhaps worked with him at KLH, where Winslow later worked before founding his own companies.
Lars, thank you for your thoughtful contribution.
I have never claimed to be an engineer. My day job is as a classical-music record producers. I prototype loudspeakers as a sideline.
Strangely enough, two weeks ago I was recording a man who was a student of Nadia Reisenberg, Josef Hofmann's teaching assistant at the Curtis Institute:
That's iPhone audio, but the venue is Mechanics Hall (Worcester Massachusetts) and that's a relatively new 9-foot Hamburg Steinway.
It's brief, but I hope you enjoy it.
john
I play the cello btw and appreciate good classical recordings - hard to come around.
cheers
Lars
- Home
- Member Areas
- The Lounge
- Going Where Angels Fear to Tread, We Propose Two Additions to the Thiele-Small "Canon" of Driver Parameters