Hey David,
I too can't find anyone matching that description in this thread. There only seem to be three types of people still posting:
A) Those with actual experience of hearing and having worked with the WE units and associated horns;
B) Those who are genuinely curious about the heuristic method apropos historical loudspeaker development and what it might teach us about pros and cons of contemporary methodologies; and
C) Those who haven't heard them whose intellectual sensibilities are offended by the very notion that group A or B might actually prefer a technology predating their existence on the planet.
As a member of group A I think we've made clear - time and time again - that there have been many, many improvements made since 1928, and yet group C still seem offended that group A can admit there have been improvements and still prefer the stuff made without all those "improvements".
Can we move on now? I'm still really keen to hear from those in group A and B.
Last edited:
Good point. But maybe compression driver have changed the least? Horns have gone thru quite a few changes.
suggest checking the output/sensitivity etc. in the simulation at an octave above that 80Hz. point. It will likely give a better picture. The 80Hz point might be at the hairy edge unless the simulated horn and setup is just right... just a thought.
I may have understated the top cut-off..it really depends on what instrument (pipe organ) they were attempting to mimic with this system...but 200 cycles would have been very close to the max cut off (+ 100 or 200 HZ or so) but I am only guessing of course, based on what organists are telling me.
I have since learned from a long time Chicago theater organ enthusiast that the Conn Horn and Jensens were designed solely and exclusively for the organ in question as the Chicago Convention Hall had it's own PA system...apparently Conn was attempting to commercially replicate the thunderous sound of the Morton Theatre Organ at the Chicago Stadium, which, up until the 1950s, was usually the site of such large conventions. He also believes the Bass Horns would have had a cut-off somewhere in the 200-300 hz region max and been devoted...just as the largest pipes in an organ...to bass pedal notes.
As to the "who's heard the WEs" "who hasn't heard them," keeping an open mind would seem to be the minimum we should expect...I appreciate the modern engineering perspective of Speaker Dave who, despite some assertions to the contrary has, I believe, contributed significantly and in detail to this discussion.
But I look at the build quality of many vintage drivers, which by today's ability to measure and engineer would be considered less than sonically optimal, and can marvel at the attention and detail put into things like finish, basket construction, motors, tags, cone material, surrounds etc....and yes, even performance.
For fellows operating in the dark and sans the measurement abilities of today they managed, somehow, to produce loudspeakers that performed and continue to perform quite remarkably even by today's standards (Quad ESL57 circa 1956 B&W DM70 circa 1970 and my SFB/3s circa 1956 are just a few that come to mind) Do they have limitations, yes! But from the subjectivists and objectivists among my group of acquaintances, the consensus is they are remarkable in more ways then they are deficient..and in ways that seem to defy the individual measured breakdown of each driver/x/over, cabinet etc....there is just this remarkable...synergy.
Their subjective performance IMO isn't miles behind what is being produced today, some six decades later, but maybe just a few inches or a foot behind. And when I think it more than possible that some psycho-acoustic nostalgic bias on my part is colouring my opinion I invite as many diverse ears as I can to the dance for their subjective observations.
These products were not out of the financial reach of the average audio hobbyist of the day but in comparison to the finish of most of today's products (with the exception of the boutique and uber-audiophile) simply don't show the same overall craftsmanship...again IMO.
I've followed this thread from the outset and learned things from the proponents and naysayers alike. I tend to be a vintage audio fan simply because of economics...its all I can afford. I've owned Altec compression horns, dabbled with JBLs and others etc...(likely considered middle-brow by some here, no doubt) and never found them to my liking (probably because they weren't set up properly) however, I'm open to the suggestion that properly implemented they can do some magical stuff.
And I've heard some new product that also is stupendous..so it's not as if old farts like myself can't appreciate how modern engineering and design can certainly deliver the goods..in spades.
What fascinates me is that these old WE horns are regarded so highly by a pretty wide spectrum of people who say they've heard them in the flesh and have heard so many other great systems (vintage and contemporary)...I am equally fascinated by the objectivists, who say unless you can measure it, it isn't real.
If there are measurements of these drivers..let's post 'em...but they'll only tell one part of the story...measurements and sims without sonic proof are insufficient on their own as many a DIY speaker builder can attest to.
A parallel argument about new vs.old technology specific to mechanical pipe and theatre organs vs. digital organs has been raging some time...and even further back it raged between proponents of Baroque Period Organs vs. Romantic Period organs...what it leaves us, of course, is a staggering array of design, engineering, manufacturing and voicing approaches.
As the original query in this thread was how far have we come in the last 100 years relative to the WE horn in question and related designs? the answer is probably not as far as many of the measurements crowd would suppose and farther than the subjective crowd realize.
Last edited:
Which is why I mentioned three categories, those who have heard them, those who haven't but are curious and possess an open mind and lastly, those who haven't heard them but insist the first group must be wrong. When Speaker Dave writes:
"I feel that there should be a couple of voices in this thread that balance out the unabashed and somewhat unrealistic love being showered on what where probably quite well engineered products of their time"
he outs himself as someone who thinks he needs to tell others how their "love" of WE gear is ultimately misguided and begs for us to point out where we should shower our affection instead. All without having ever heard them for himself, of course. (I put him in group C, and hypothesize he really wants us to shower our love all over his Snell XA Reference Tower instead...)
And yet as soon as those observations are applied to the WE gear a bunch of dudes show up waving their 30+ years experience demanding we prove it. Having never heard them, of course.
Because the question for me is still:
Have we, in the course of the last 90 years of mechanical and material scientific progress lost something in as much as we've gained something?
Why are Quads still relevant? Why is vinyl? Why are tubes? Horns? Compression drivers?
Anyone with even a modicum of basic engineering knowledge knows the above were abandoned in favor of newer technologies that were demonstrably objectively better. Because the latter were objectively better. The issue many have difficulty comprehending is why what measures best is not always that which is preferred - hence vinyl, tubes, horns and compression drivers. It's easy to measure something measurable and declare it better than something that measures less well.
IQ tests are interesting because they reveal the fallacy of "what measures better must be better". Does testing high (say, 130) in an IQ test mean you're more likely to be rich? Famous? Employable? Diligent? Honest? Moral? Ethical? For example, let's say George W. Bush comes out of retirement and wants to work at your Silicon Valley hedge fund - do you hire him? I mean, the guy has an (estimated) IQ of 125, which puts him squarely in the same league of (actual) MD's and PhD's, so, measurably speaking, he must be better than all those other guys who only have an IQ of 124, right? Right?
It's possible to measure almost anything a speaker does. So it's worth asking why measurably defective ones continue to be musically relevant and conversely, why measurably superior ones are often musically defective. This is not a subjectivist/objectivist argument - it's a discussion related to why there is a gap between objective measurements and subjective feelings and whether compression drivers and horns ala Western Electric theatre speakers (and vinyl and tubes) might tell us something about that gap and what can be done to bridge it. It asks whether a paradigm shift is necessary in the way we evaluate common notions of loudspeaker performance. It is absolutely not bashing technological progress or objectively measurable parameters - it's asking whether technological progress and objectively measurable parameters have any utility value apropos listening to music in the home - and if not, then what is the paradigm shift needed either individually or collectively for the construction of a dialectical we can all embrace. That would be progress.
Last edited:
Are you thinking that the horns were used as a sort of "electronic pipe"? That they were used not as a normal horn speaker, but as a resonant device like a musical horn? That's the only way I can see them reaching down that far. Of course they could have powered down below cutoff, as Tom Danley does in his horns. Hard to say.
On a similar note, I was thinking about Bill F.'s sim of diaphragm displacement at 80Hz in the 15A. Would the 15A actually load the 'fram down that low? I was thinking (from memory) that its cutoff was around 110Hz, which would make sense given the size of the mouth. Anyone know?
A couple of posts use the term "heuristic" and imply that Western Electric designed product was not designed under modern principals or should not be judged with any modern objectivist criteria.
Everyone should take a look at this:
http://charm.cchcdn.net/redist/pdf/maxfield_harrison.pdf
It gives a good glimpse into the thinking of Western Electric at the time. Specifically it details how they designed the electrical disc recording method and the Orthophonic Victrola playback units. (1925). Note that these were directly used in the first cinema talkie systems as the Vitaphone process. Large 33 rpm discs were played in sync to the film. Sound on film also existed but the quality of discs was initially superior. Later WE cinema systems follow the same design principles and were developed by the same team.
They start with stating the goal of sound reproduction: "perfect reproduction requires that the components of the reproduced sound reaching the ears of the listener should have the same relative intensity and phase relationship as the soundreaching the ears of an imaginary listener to the original performance".
From here they move into studio acoustics (quoting Sabine reverb times), then move into electrical equivalent circuits for solving the modeling of electrical record cutters. For reproduction of the discs they cover the reproducer (essentially a compression driver diaphragm minus the voice coil). They describe the Orthophonic horn which is A 115 cycle cutoff exponential horn folded much like a WE 15.
They finish with an acoustical response curve of the full phonograph vs. A previous generation victola. It is fairly flat from 115 to 5000 Hz vs the narrower band, less flat, Victrola.
All good objectivist engineering in the modern mold. So why do we insist that Western Electric product shouldn't be judged on objective terms, when clearly the designers of the products did just that?
David S.
Everyone should take a look at this:
http://charm.cchcdn.net/redist/pdf/maxfield_harrison.pdf
It gives a good glimpse into the thinking of Western Electric at the time. Specifically it details how they designed the electrical disc recording method and the Orthophonic Victrola playback units. (1925). Note that these were directly used in the first cinema talkie systems as the Vitaphone process. Large 33 rpm discs were played in sync to the film. Sound on film also existed but the quality of discs was initially superior. Later WE cinema systems follow the same design principles and were developed by the same team.
They start with stating the goal of sound reproduction: "perfect reproduction requires that the components of the reproduced sound reaching the ears of the listener should have the same relative intensity and phase relationship as the soundreaching the ears of an imaginary listener to the original performance".
From here they move into studio acoustics (quoting Sabine reverb times), then move into electrical equivalent circuits for solving the modeling of electrical record cutters. For reproduction of the discs they cover the reproducer (essentially a compression driver diaphragm minus the voice coil). They describe the Orthophonic horn which is A 115 cycle cutoff exponential horn folded much like a WE 15.
They finish with an acoustical response curve of the full phonograph vs. A previous generation victola. It is fairly flat from 115 to 5000 Hz vs the narrower band, less flat, Victrola.
All good objectivist engineering in the modern mold. So why do we insist that Western Electric product shouldn't be judged on objective terms, when clearly the designers of the products did just that?
David S.
Last edited:
BTW, I forgot to mention before that the 15' long expo horn with 16 sq. ft. mouth in half space calculates to be a 37Hz flare!
But like I said, I was just guessing on most of the horn and driver parameters. I'd love to get more accurate inputs to populate the model.
Nice and exciting topic and long too 😱
Everyone just wait and see what will happen if a modern large scale manufacturer undertakes to make a super system 😉
Imagine Westinghouse or General Electric or Yamaha or Mitsubishi (Panasonic/Technics) or another giant undertaking something big and perfect...
In the 30's there were Hispano Suiza - cars better than anything. Now we have a Rolls Royce Phantom that is far closer to ordinary cars than the super cars of the past.
Obviously in 1928 WE did a state of the art complete system from the source to the speakers. Imagine today's state of the art let's say modeled with super computers and executed with space equipment and materials 😉
Technology is just so advanced, that there is no need for extraordinary effort in order to achieve satisfactory performance.
And one more thing, state of the art should be compared to state of the art. I am quite certain that for about 20 years we do not receive state of the art because the technological giants backed from the audio market. Example is Mitsubishi, they did the SL1200 some 40 years ago and never put an effort to do better, not that they can't. For the present capabilities of human senses, there is a clear boundary of "good enough" and it won't change irrespective of technology.
Since good enough was achievable nearly 100 years ago it is obvious that the performance is much more Engineer dependant rather than materials or technology dependant.
Best Regards!
Everyone just wait and see what will happen if a modern large scale manufacturer undertakes to make a super system 😉
Imagine Westinghouse or General Electric or Yamaha or Mitsubishi (Panasonic/Technics) or another giant undertaking something big and perfect...
In the 30's there were Hispano Suiza - cars better than anything. Now we have a Rolls Royce Phantom that is far closer to ordinary cars than the super cars of the past.
Obviously in 1928 WE did a state of the art complete system from the source to the speakers. Imagine today's state of the art let's say modeled with super computers and executed with space equipment and materials 😉
Technology is just so advanced, that there is no need for extraordinary effort in order to achieve satisfactory performance.
And one more thing, state of the art should be compared to state of the art. I am quite certain that for about 20 years we do not receive state of the art because the technological giants backed from the audio market. Example is Mitsubishi, they did the SL1200 some 40 years ago and never put an effort to do better, not that they can't. For the present capabilities of human senses, there is a clear boundary of "good enough" and it won't change irrespective of technology.
Since good enough was achievable nearly 100 years ago it is obvious that the performance is much more Engineer dependant rather than materials or technology dependant.
Best Regards!
Last edited:
Any designer would be lying if he said he didn't want others to admire his products. But if others started "showering love" beyond what was warranted, ascribing perfection or overlooking faults of the XA References (or 11 ohm LS3/5a, or JBL 4430/4435, or Mac XRT24, or JBL KM1) then I would be the first to correct them.
Each product from every company is a bundle of compromises and could have been better if we had the development time, bill of materials money or especially the knowledge that later designers had.
In the end we all design products for a fickle market. Some catch on and some don't. Some are kindly regarded by audio history and others aren't. Curiously I am more proud of the annonymous PSB in-wall with a complex bracket that doesn't rattle or the crappy Essex 3-way that sounded okay in spite of a 2 element crossover.
David S.
I think the answer to this dilema is pretty simple if you stop and think about it with an open scientific mind instead of a mystical or dogmatic one - if the device that sounds better measures worse and visa versa, you're measuring (or assigning too much importance) to the wrong measured parameters.
In this regard modern measurement techniques such as computer based FFT may have actually made things "worse" by placing too much power at our finger tips and allowing us to quickly and accurately measure a whole plethora of parameters that simply weren't possible to measure 30+ years ago, at any price point.
The temptation is then to assume that because we can measure all this stuff so easily, "better" measured performance in one particular parameter of the design, for example harmonic distortion, "must" sound better, simply because we know gross distortion is "bad" therefore less must be "better".
But what if the threshold for detection is quite high, the speaker already falls below this threshold, and making distortion much lower makes some other parameter of a hypothetical design such as frequency response flatness worse ? Which sounds better ? Which measures better ?
Sometimes too much information can just make it hard to see what matters.
Find out through controlled listening tests which parameters really matter (and their relative importance) and which matter little, then focus on measuring and optimising the things that really matter, instead of the things that just seem like they should matter.
That is where the real progress is to be made, when modern engineers start to accept that for measurements to be meaningful in evaluating speaker performance they have to be related back to how we perceive sound, not some idealistic notions of what should and shouldn't matter from an engineering perspective. A speaker designer must also have a solid knowledge of room acoustics and human hearing perceptual science before he can make intelligent design decisions and tradeoffs. I would wager that many don't have the necessary breadth of understanding of those two.
Every decade since the 70's has had its "fad" where one particular parameter was prized above others, then dropped when the next fad came along. I'm not old enough to remember them all well or the exact order, but some include;
Very wide dispersion at high frequencies ? (beginning with the prototypical dome tweeters in the 70's) Time aligned drivers with 1st order crossovers ? (Dunlavy etc) Drivers operating entirely in their piston range ? (Seashell shaped B&W Nautilus) Exotic cone materials ? (Any of the plethora of modern materials like kevlar, metals etc) DSP based linear phase ? ("Perfect phase correction", at one point in space at least...Ultimate Equaliser, etc) How many of these things really matter at the end of the day, how many of them have been proven important in carefully designed blind listening tests, and how many just seem like a good idea ?
I think its really hard for some designers to accept that designing a speaker that sounds good is not just an engineering exercise. Although we have some idea of what parameters contribute to good sound, I think the book is not yet closed on the topic of which measurements matter and by what weighting factor.
As long as designers dogmatically cling to design philosophies that aren't necessarily based in the science of how our hearing works, you'll get speakers that measure "good" (at least in certain parameters which the designer thought was important) but maybe don't sound great, because the parameters that really matter have been neglected or compromised.
I have no doubt however that measurements - ones we already know how to measure today, can adequately describe what sounds good and what doesn't, we just need to know which measurements to trust and how to interpret them and learn to let go of the ones that are less important.
The irony is that one of the easiest to measure and one of the only things that was possible to measure in the very early days - amplitude response - still seems to be one of the most important if not the single most important parameter, far outweighing ones like distortion, power response etc. Not that those others don't matter at all, just that they don't matter as much as many people think.
Regarding old speakers, I have a lot of respect for some of the high end full range drivers of the early 70's, like the Coral's, that can sound really good even by todays standards with fairly minimal tweaking and EQ, despite such incredibly limited measurement technologies available around 1970 when they were designed.
They are pre-Thiele/Small yet they have T/S parameters which whilst requiring a moderately large box due to high sensitivity actually give a pretty decent non-boomy bass reflex alignment with a nice smooth rolloff. Anyone who has played with T/S parameters knows that this doesn't happen by accident - many cheap drivers with poor T/S parameters simply won't give an acceptable response in any box size or tuning. You could swear they understood T/S parameters when they were designed, yet its likely that they didn't, unless the Japanese had independently derived it and not published.
And frequency response in general - at the time all they probably had was a pen plotter that could measure amplitude response, no easy way to measure phase vs frequency, distortion measurement would have been a painful manual process, and things like CSD etc were not yet invented.
Yet they managed to make a large dual cone driver that is flatter and more neutral than most modern full range drivers with so little to go on other that a simple amplitude response measurement and probably a LOT of listening tests. My hats off to some of the old time speaker designers, it makes you wonder what they could have done with the same attitude to design but with modern materials and measurement techniques...of course a lot of old speakers were awful, we tend to remember only the very best examples of each era, just like we forget all the awful movies of each decade and just remember the classics 😀
Last edited:
And yet you happily accepted a "Class A" recommendation from Stereophile for the Snells - guess they warranted all the adulation with no need for you to point out any faults they had, right? (You might want to re-read your manufacturers comment, but I couldn't find anywhere you said "Well, thanks, but you should know this...")
See, this is the fallacy of logic applied to the scientific method all the time. I understand compromises - it's part of my day job. But your statement assumes that knowledge is an ever forward-moving progressive endeavour, and I'm arguing that forward progress leaves behind as much as it picks up. By moving away from groove modulations in polyvinyl chloride read by a cantilever and coil to binary samples encoded in polycarbonate read by a laser - we lost something. Sure, we gained impressive bandwidth and much better signal/noise, but I'd like to suggest we abandoned something inherent in musical replay of the analogue format by moving to digital.
I'm not anti-progress, Dave. I'm not saying that improvements of many kinds have not happened since 1928. I'm not saying measurements are evil or that designing something of value in an over-saturated market is not fiendishly difficult.
I'm saying there might be a reason to look at abandoned technologies, and see if we cannot learn something to improve the state of the art in music replay for the home. A good start might be to actually listen to a pair of WE drivers first, and go from there, no?
Because it's not 1928 anymore David. We've now had a chance to listen to many many more references since then, building that into a very wide, broad based understanding of music replay encompassing many innovations across all formats and mediums.
We can't pretend we've never heard a pair of Dynaudios or a Tivoli or reel-to-reel or SACD. And once we have, it becomes part of a narrative that we cannot un-experience. Therefore, we now have the luxury of comparing across time things that no engineer could have foreseen last week, let alone in 1928.
So now we have 15a's and 18" OB field coils in people's homes with DIY crossovers and paper-in-oil caps and tungar power supplies... They're re-creating a narrative that goes far beyond what the WE engineers could have ever intended or thought possible. That's a heuristic (experiential) approach and is why this thread is here on a DIY forum. WE engineers were scientists. They weren't necessarily using a heuristic methodology, but we are.
Last edited:
What does it calculate out to in full pi space? I'm guessing about 138Hz. Here's why:
Almost all horns I've measured have a cut-off frequency related to the mouth width. For a round horn that's the diameter, for a square or rectangular horn, the hypotenuse. If you find the frequency of that wavelength, then multiply by 0.75, you will be very close to cut-off. That being the frequency under which the response takes a rapid drive and distortion goes way up. I don't know how well that matches theory, but I've seen it across a large number of horns of different shapes and sizes.
The 15A has a hypotenuse of about 73", which equals 185Hz. 185x0.75=138.75Hz
When we ran it in a 4-way system, it was high passed at 250Hz.
But knowledge is an ever forward-moving progressive endeavor. You can lament that “they don’t make them like the used too” (and investigate the commercial reasons why that is probably true) but it has nothing to do with lost knowledge.
Before Thiele people thought that vented box design was about equalizing the impedance peaks. They were wrong. He created a simplified approach and everyone could benefit. It was a piece of the puzzle. Once analyzed and accepted by the community it became of benefit to all. We could move on and work on the next piece of the puzzle.
Fincham and Berman at KEF said, “I think we can take the magnitude and phase of loudspeaker units, add the crossover network, load it with the true impedance curves, account for the cabinet dimensions and get an accurate computer model of the final system.” Another piece of the puzzle (and now we have free spreadsheets to let us do that). Who first had an inkling that cabinet dimensions caused diffraction and that the frequency response was impacted by it? Now we all understand it (and have free software to model it).
In the Maxfield and Harrison paper I linked to earlier, they repeat that logarithmic (exponential) horn growth is the mathematically correct approach. They got it from A.G. Webster, a University professor. He studied horn shaped musical instruments, products of empirical design. He found that the common factor between the instruments and the reason for their characteristic performance was their exponential area growth. He then proved that this growth would give the most extended loading for a phonograph or loudspeaker. With this piece of the puzzle in place Maxfield and Harrison could work on other challenges, such as the new design reproducer head.
Read about concert hall acoustics where it wasn’t until the 80’s that researchers started to understand the importance of early lateral reflections. The failure of concert halls built to previous design rules suddenly made sense. Or read the fascinating account of Sabine, in the early 1900's dragging seat cushions into a piano recital hall and discovering that 8 cushions made the sound too dead and 6 cushions weren’t enough (the subjective discovery predated the objective understanding).
Toole gathered several dozen loudspeakers and applied carefully controlled listening tests to rank order them (and many repetitions of the test to prove that many listeners would consistently give the same rank ordering). After the subjective tests he set about to measure the units to find which technical factors seemed to matter and which didn’t. Olive took it a step farther and was actually able to mathematically weight the measurable factors and reliably predict subjective ranking purely on measurements alone.
Keele on constant directivity horns. Linkwitz and Riley on crossover networks. Allison on room boundaries. Vilchur and the acoustic suspension system. Shorter and Harwood on BBC monitor design. Each discovery or research paper adds to the body of knowledge and lets the next researcher, engineer, or scientist work on the next problem. None of the previous knowledge is lost as long as the next generation is willing to read and study what came before.
Should we study designs from long ago to see what was unique and admirable about them? Absolutely, just don’t try and convince me that there hasn’t been progress, a lot of it, in the intervening decades.
David
Progress implies a forward motion where there are continuous (albiet incremental at times) improvements to previous efforts.
I think that in the case of speakers and acoustics there have been improvements in terms of parsing understanding, and drilling into details, reducing to engineering or scientific method specific aspects. But, does this ensure or define "progress"?
Perhaps in some sense it certainly does define a type of progress.
But is there presently a speaker that is an improvement upon the Quad 57? I would suggest that there are speakers that when looked at across all possible aspects would appear to be "better", appear to have "better" measurements - at least they would clearly test better in several ways. I picked the Quads since they do have some easy to find limitations, and they are not Western Electric.
Now, here is the real question, and I am serious. The Quad 57 happens to sound terrific in the right situations, the right (not too large) room, and with the proper amplification and signal chain. It also measures very well in some regards. So, what speaker is truly going to be more appealing to the listener's ear, overall? Which speaker is going to show us an incremental or order of magnitude improvement over the Quad 57? Which speaker?
It is not easy to better the old Quad, except in terms of raw SPL and HF extension above 15kHz or so and LF extension below 60-70Hz...
All of the more recent studies and work, JAES articles, and the like still do not and have not given us another definitive speaker like that old Quad 57. One would think, IF there was that much "progress" that the Quad 57 would be relegated to the dust bin, and would be laughed at entirely. But it is not. It is still better to listen to than (imho) maybe 99% of all speakers built using all that "progress".
Think that makes the point well enough. Someone else can now have the soapbox.
_-_-bear
EDIT: one thing that is clearly an improvement is FEM analysis for magnetics and for surfaces (like cones). I think dynamic drivers are clearly better on average than before the advent of these tools. So the average is up. The question is how much better are dynamic drivers, or is the best really much better than in previous decades? Perhaps for dynamic drivers the best is better now. But, compression drivers and ESLs?
I think that in the case of speakers and acoustics there have been improvements in terms of parsing understanding, and drilling into details, reducing to engineering or scientific method specific aspects. But, does this ensure or define "progress"?
Perhaps in some sense it certainly does define a type of progress.
But is there presently a speaker that is an improvement upon the Quad 57? I would suggest that there are speakers that when looked at across all possible aspects would appear to be "better", appear to have "better" measurements - at least they would clearly test better in several ways. I picked the Quads since they do have some easy to find limitations, and they are not Western Electric.
Now, here is the real question, and I am serious. The Quad 57 happens to sound terrific in the right situations, the right (not too large) room, and with the proper amplification and signal chain. It also measures very well in some regards. So, what speaker is truly going to be more appealing to the listener's ear, overall? Which speaker is going to show us an incremental or order of magnitude improvement over the Quad 57? Which speaker?
It is not easy to better the old Quad, except in terms of raw SPL and HF extension above 15kHz or so and LF extension below 60-70Hz...
All of the more recent studies and work, JAES articles, and the like still do not and have not given us another definitive speaker like that old Quad 57. One would think, IF there was that much "progress" that the Quad 57 would be relegated to the dust bin, and would be laughed at entirely. But it is not. It is still better to listen to than (imho) maybe 99% of all speakers built using all that "progress".
Think that makes the point well enough. Someone else can now have the soapbox.
_-_-bear
EDIT: one thing that is clearly an improvement is FEM analysis for magnetics and for surfaces (like cones). I think dynamic drivers are clearly better on average than before the advent of these tools. So the average is up. The question is how much better are dynamic drivers, or is the best really much better than in previous decades? Perhaps for dynamic drivers the best is better now. But, compression drivers and ESLs?
Last edited:
David, I'm not trying to convince you of anything. You're trying to convince those who have heard the WE speakers that our "love" for them is misguided and disproportionate. But you don't get to call the shots as to what I choose to like or love, thanks.
I did say:
"I'm not anti-progress, Dave. I'm not saying that improvements of many kinds have not happened since 1928. I'm not saying measurements are evil or that designing something of value in an over-saturated market is not fiendishly difficult..."
right?
David, I agree with you. Lots and lots and lots of progress. Much of it has been hugely important for the development of loudspeaker design. I'm sure many of the DIY crowd here have benefitted from it immensely, and I'm sure it's informed your practice as a designer/engineer too. That's your job, right?
Actually, my life experience thus far tells me this: Each discovery or research paper adds to the noise and often makes the next researcher, engineer or scientist work on a problem unrelated to the primary purpose of the initial endeavor. It's not the amount of stuff you know, it's knowing what is not useful and the only way one can know that is to have articulated the gap between the initial primary purpose and where we are now. Then one can concentrate one's effort on what might contain the greatest utility value related directly to the primary concern (the experiential act of listening to music in the home) rather than creating a new ceramic driver with lower distortion because it has lower distortion.
Because if, as I'm trying to argue, the primary concern gets lost amongst the thrill of better living through chemistry (newer creamic driver materials or cleaner clearer waterfall plots) then we only end up here - debating progress via an objectivist/subjectivist dialetical rather than redefining one that gets to the core of the gap.
When you say none of that knowledge is lost as long as the next generation is willing to read and study what came before, I say it's not lost - just buried beneath too much new information that is often progress for progress' sake rather than addressing the primary core concern. Hence why we have 800+ posts on a technology from 1928.
So where are we now? You're probably still clinging to the amassed knowledge of Thiele and Finchman and Berman and Toole and Linkwitz and Reily. That's understandable since your background requires fundamental knowledge of those guys work. It's also understandable that it provides an acknowledged framework from within which to work. Without them we wouldn't have the modern loudspeaker.
But that's why paradigm shifts don't happen very often. It causes such a radical split for the practitioner that it forces one to revise a tremendous amount of heretofore accepted wisdom that was the backbone for their previous worldview. It's entirely possible that if you ever get a chance to actually hear music over a set of WE speakers, your world may not change. And that's fine. But mine did. So where are we now?
Last edited:
My take is that the effect we're seeing from the improved materials, knowledge and number crunching is that cheap stuff is better than ever and really good stuff is affordable. The $40 "Altec" speakers next to this monitor would amaze the engineers at the real Altec. They could do it, but not cheap.
I really can't say that the best haven't been bettered, but the old ones were great and can't be compared all that directly since their environment has changed and they're designed to different criteria.