First, a disclaimer: many audiophiles, probably most, can't hear the differences between magnet materials. They sound like the difference between transformer core materials - similar to the difference between 80% nickel, M6 steel, and amorphous cores. If you've not heard that kind of difference for yourself, the fairly subtle differences between driver magnets may not be audible. It is less audible than the much grosser effects of diaphragm resonance, horn modes, horn diffraction, internal standing waves in the cabinet, cabinet edge diffraction, and crossover design errors. The more severe effects of these dominate the sound of most high-end loudspeakers on the market; all of these design errors are audible in the first seconds of listening if you have some experience in designing loudspeakers.
To my ear, field-coils, Alnico, and neodymium magnets are all noticeably better than ceramic magnets - for lack of a better word, they have a "fresher", more vivid, more colorful sound. The field coils sound a bit quieter, more understated, while the Alnicos have a more vivid, more colorful sound. Neo sounds partway to the duller sound of ceramic, while having a fair amount of Alnico's vividness. Some say neo can sound "brash" on some program material, but a lot depends on details of gap saturation and magnetic path design.
This is a purely subjective assessment on my part, and not shared by other loudspeaker designers. Many designers - who I respect as good engineers who know their stuff - are convinced there is no audible difference between magnet materials at all, but then they listen to Class AB transistor amplifiers - we live in different subjective worlds. I find the sound of nearly all Class AB transistor amplifiers to be dull and flat, with much less resolution, tonal color, and spatial dimensionality than vacuum-tube amplifiers, particularly zero-feedback all-triode amplifiers.
A lot of loudspeaker designers treat Alnico and field-coils as exotic and super-expensive. A historical note: all loudspeakers, even for the cheapest radios and PA systems, used field coils until World War II. That's all there was. There was no other way to energize a Rice & Kellogg cone loudspeaker, compression driver, or cutterhead for a recording lathe. In the late Thirties, an aluminum/nickel/cobalt alloy was discovered and named AlNiCo, and in the space of a few years, replaced all field coils. By the end of the war, field coils had disappeared, replaced by Alnico. It was used in all loudspeakers, from the cheapest table radio to the highest-performance theater speakers (which represented the state of the art back then) right through the early Seventies. Alnico magnets, vacuum-tube amplification, and analog sources was the sound I grew up with - everything sounded that way, and was universally used in recording systems.
At the same time as the introduction of reliable transistor amplifiers (early Seventies) and mass-market phase-out of vacuum-tubes, a civil war broke out in Africa that interrupted the worldwide supply of cobalt. Since this war was covertly financed by the Cold War powers, it burned on for many years, and the supply of cobalt was reserved for military applications in the USA and Russia. Loudspeaker manufacturers had a problem, and one that had to be solved fast: what to replaced AlNiCo with? Cast ferrite, long used for RF chokes and filters, was pressed into service as a replacement, and Alnico disappeared from markets worldwide. Manufacturing techniques were developed that made cast ferrite magnets less labor-intensive than machined Alnico, driving down costs.
Notice NONE of this has anything to do with sound quality; but the demands of geopolitics. WWII was looming, and compact, powerful magnets were needed in many different sectors of the economy, so Alnico replaced heavy, inefficient, and hot field coils. A Cold War civil war interrupted the supply of a strategic material, driving it off the commercial market.
The cold war is long over, although the wackier fringe politicians in this country think it will somehow return out of nowhere (Communism! Socialism! Fluoridation! Booga-Booga!). Hello! The psywar today is about Terrorism, didn't you get the memo?
Cobalt is available again on world markets, but few loudspeaker manufacturers want to hassle with machining it. The soft iron used in field coils is easier to machine, and winding a few hundred turns of magnet wire around a soft-iron core is exactly what was done in the Twenties and Thirties for every loudspeaker made. All loudspeakers from the Twenties to the early Seventies were either field-coils or Alnico. So it's not exotic; it's just different than gluing on a precast ceramic magnet, which is the easiest and cheapest of all.
To my ear, field-coils, Alnico, and neodymium magnets are all noticeably better than ceramic magnets - for lack of a better word, they have a "fresher", more vivid, more colorful sound. The field coils sound a bit quieter, more understated, while the Alnicos have a more vivid, more colorful sound. Neo sounds partway to the duller sound of ceramic, while having a fair amount of Alnico's vividness. Some say neo can sound "brash" on some program material, but a lot depends on details of gap saturation and magnetic path design.
This is a purely subjective assessment on my part, and not shared by other loudspeaker designers. Many designers - who I respect as good engineers who know their stuff - are convinced there is no audible difference between magnet materials at all, but then they listen to Class AB transistor amplifiers - we live in different subjective worlds. I find the sound of nearly all Class AB transistor amplifiers to be dull and flat, with much less resolution, tonal color, and spatial dimensionality than vacuum-tube amplifiers, particularly zero-feedback all-triode amplifiers.
A lot of loudspeaker designers treat Alnico and field-coils as exotic and super-expensive. A historical note: all loudspeakers, even for the cheapest radios and PA systems, used field coils until World War II. That's all there was. There was no other way to energize a Rice & Kellogg cone loudspeaker, compression driver, or cutterhead for a recording lathe. In the late Thirties, an aluminum/nickel/cobalt alloy was discovered and named AlNiCo, and in the space of a few years, replaced all field coils. By the end of the war, field coils had disappeared, replaced by Alnico. It was used in all loudspeakers, from the cheapest table radio to the highest-performance theater speakers (which represented the state of the art back then) right through the early Seventies. Alnico magnets, vacuum-tube amplification, and analog sources was the sound I grew up with - everything sounded that way, and was universally used in recording systems.
At the same time as the introduction of reliable transistor amplifiers (early Seventies) and mass-market phase-out of vacuum-tubes, a civil war broke out in Africa that interrupted the worldwide supply of cobalt. Since this war was covertly financed by the Cold War powers, it burned on for many years, and the supply of cobalt was reserved for military applications in the USA and Russia. Loudspeaker manufacturers had a problem, and one that had to be solved fast: what to replaced AlNiCo with? Cast ferrite, long used for RF chokes and filters, was pressed into service as a replacement, and Alnico disappeared from markets worldwide. Manufacturing techniques were developed that made cast ferrite magnets less labor-intensive than machined Alnico, driving down costs.
Notice NONE of this has anything to do with sound quality; but the demands of geopolitics. WWII was looming, and compact, powerful magnets were needed in many different sectors of the economy, so Alnico replaced heavy, inefficient, and hot field coils. A Cold War civil war interrupted the supply of a strategic material, driving it off the commercial market.
The cold war is long over, although the wackier fringe politicians in this country think it will somehow return out of nowhere (Communism! Socialism! Fluoridation! Booga-Booga!). Hello! The psywar today is about Terrorism, didn't you get the memo?
Cobalt is available again on world markets, but few loudspeaker manufacturers want to hassle with machining it. The soft iron used in field coils is easier to machine, and winding a few hundred turns of magnet wire around a soft-iron core is exactly what was done in the Twenties and Thirties for every loudspeaker made. All loudspeakers from the Twenties to the early Seventies were either field-coils or Alnico. So it's not exotic; it's just different than gluing on a precast ceramic magnet, which is the easiest and cheapest of all.
Last edited:
Much appreciated and much to think about.
There's a triumphant narrative in the audio biz - mostly pushed by the idiotic reviewers - that we've been on a magnificent technological march forward from the Twenties to today.
That's true of computers - Moore's Law and all - but the "progress" of audio has largely been driven by events external to the industry, and have more to do with steadily driving down costs and marketing campaigns than anything else.
When you inflation-adjust prices from the Twenties through the Seventies, audio was actually quite expensive, right down to the cheapest table radio from Sears Roebuck. The catalog price in the 1950's might have been $29.95, but that's really $250~300 now - for the cheapest monophonic AM radio on the market. A dangerous radio with no power transformer, a live chassis directly connected to one side of the AC line, five tubes, a single 5" speaker, and a plastic cabinet. If the plastic insulation failed and the 2-prong non-polarized power plug was reversed, you could get a 120V shock. (Electrical devices like this are now illegal in nearly every country.)
A good-quality RCA or Zenith 21" color TV set in 1958 was $699 to $999 (depending on cabinet - the chassis was the same), with no discounts, and the worst monochrome "portable" (which means on a rollaround cart) 17" TV was $159. In real dollars, a price range from $1200 to $8000. The picture on the cheap 17' monochrome set - the kind of TV most people had back then - was nearly unviewable in modern terms, with no DC restoration, atrocious geometry, and a power supply with almost no regulation. The color sets had a pretty good picture, aside from the hassle of adjusting the "tint" control when the network went from New York to local, but the extremely high costs meant that almost nobody had them.
Electronics were expensive in the vacuum-tube era, while automobiles and houses (in relative terms) were cheaper. In real dollars, the minimum wage was about $11~13/hour, and the top marginal tax rate was 91%. The Eisenhower era was a different world.
That's true of computers - Moore's Law and all - but the "progress" of audio has largely been driven by events external to the industry, and have more to do with steadily driving down costs and marketing campaigns than anything else.
When you inflation-adjust prices from the Twenties through the Seventies, audio was actually quite expensive, right down to the cheapest table radio from Sears Roebuck. The catalog price in the 1950's might have been $29.95, but that's really $250~300 now - for the cheapest monophonic AM radio on the market. A dangerous radio with no power transformer, a live chassis directly connected to one side of the AC line, five tubes, a single 5" speaker, and a plastic cabinet. If the plastic insulation failed and the 2-prong non-polarized power plug was reversed, you could get a 120V shock. (Electrical devices like this are now illegal in nearly every country.)
A good-quality RCA or Zenith 21" color TV set in 1958 was $699 to $999 (depending on cabinet - the chassis was the same), with no discounts, and the worst monochrome "portable" (which means on a rollaround cart) 17" TV was $159. In real dollars, a price range from $1200 to $8000. The picture on the cheap 17' monochrome set - the kind of TV most people had back then - was nearly unviewable in modern terms, with no DC restoration, atrocious geometry, and a power supply with almost no regulation. The color sets had a pretty good picture, aside from the hassle of adjusting the "tint" control when the network went from New York to local, but the extremely high costs meant that almost nobody had them.
Electronics were expensive in the vacuum-tube era, while automobiles and houses (in relative terms) were cheaper. In real dollars, the minimum wage was about $11~13/hour, and the top marginal tax rate was 91%. The Eisenhower era was a different world.
Last edited:
Well Lynn, glad to see you've gone with a classic design. As an owner of VOTT A5 speakers (+super-tweeter) I'm very happy with it. Not too far from your approach.
Although I agree with you about the content of modern movies and the beauty of 70mm, don't be too fast to discount digital cinema. I do digital projection as a living. Right now I'm working with a system that has at least 7x the resolution of full HD video and more than a 70mm film print. It's rather stunning, to say the least. Good things are coming, very good things.
Although I agree with you about the content of modern movies and the beauty of 70mm, don't be too fast to discount digital cinema. I do digital projection as a living. Right now I'm working with a system that has at least 7x the resolution of full HD video and more than a 70mm film print. It's rather stunning, to say the least. Good things are coming, very good things.
I'm looking forward to 4K projection - haven't seen it around here (Denver area) yet. What I've seen of so-called Imax Digital is that is dim and nowhere close to Imax 70mm quality.
By the way, since you're in the digital-projection biz, what's happening at the theater when they switch projectors from the local ads to the real movie? Is the projector for the ads not 1920x1080, or is the 2K main projector much sharper somehow? The projector for the ads is dim and soft, and the pixel structure is very obvious, while the pixels on the main projector are much subtler, and the dynamic range is a lot better looking.
I also have to give digital credit for finally getting rid of the annoying film weave. Film weave isn't so bad in a really well adjusted projector, but we all know what's happened to projector maintenance in most theaters.
By the way, since you're in the digital-projection biz, what's happening at the theater when they switch projectors from the local ads to the real movie? Is the projector for the ads not 1920x1080, or is the 2K main projector much sharper somehow? The projector for the ads is dim and soft, and the pixel structure is very obvious, while the pixels on the main projector are much subtler, and the dynamic range is a lot better looking.
I also have to give digital credit for finally getting rid of the annoying film weave. Film weave isn't so bad in a really well adjusted projector, but we all know what's happened to projector maintenance in most theaters.
Last edited:
Although I sound like a curmudgeon about a lot of new techology, I don't worship the old stuff. You have to step outside of history, and look at the technology on its merits, not the marketing myth of perpetual progress. Some of the old stuff was superbly engineered, and has not been significantly improved in the present - I put a lot of Altec engineering and the family of direct-heated triodes in this class. The Partridge output transformer of the late Forties is still considered one of the best transformers ever made.
But the best modern - like the ESS Sabre 9018 converter, or the somewhat older Burr-Brown 1704 DAC - is superb engineering by any standard. I own a Panasonic P58VT25, and am genuinely impressed with the color quality and depth of the picture. The modern RAAL ribbons are in the same league as the Hill Plasmatronics of the late Seventies.
But the best modern - like the ESS Sabre 9018 converter, or the somewhat older Burr-Brown 1704 DAC - is superb engineering by any standard. I own a Panasonic P58VT25, and am genuinely impressed with the color quality and depth of the picture. The modern RAAL ribbons are in the same league as the Hill Plasmatronics of the late Seventies.
Last edited:
"... But the best modern - like the ESS Sabre 9018 converter, or the somewhat older Burr-Brown 1704 DAC - is superb engineering by any standard ..."
The Datasat AP20 Digital Audio Processor runs highly tweaked Burr Brown PCM4104 DACs and has gained a well earned reputation in Commercial Cinemas as a good sounding box. AP20 | Datasat Digital Entertainment
After a rough start Digital Cinema is getting better!
The Datasat AP20 Digital Audio Processor runs highly tweaked Burr Brown PCM4104 DACs and has gained a well earned reputation in Commercial Cinemas as a good sounding box. AP20 | Datasat Digital Entertainment
After a rough start Digital Cinema is getting better!
Paper probably sounds the best, for the simple reason it has mechanical characteristics similar to the speaker cone, which is typically made of paper. Materials with dissimilar speed-of-sound create reflections at the glued mechanical joint, which is very undesirable in the time domain.
But paper is a terrible material in other ways. Over time, it tends to burn and carbonize from the heat of the voice coil, which eventually leads to catastrophic failure. Since the VC, cone, dustcap, surround, and spider are integrated, the entire assembly must be replaced when the VC former crumbles and fails. Reconing a speaker essentially makes it a new loudspeaker, not something you want to do once the original paper formulation becomes unobtanium.
Kevlar, Kapton, and Nomex all sound different from what I understand, with very different mechanical properties, which in turn affects the sound of the loudspeaker.
Going deeper, the glues used to attach the VC former to the cone have a quite substantial effect on the sound as well. You can't say "Kapton sounds like this when combined with a paper cone" without discussing the sound of the different glues as well, since all glues have a degree of elasticity. The elasticity of the glue has a strong effect on the reflection between VC former and the cone - but if it's too elastic, it may fail under strong acceleration, or not have the right thermal characteristics.
The elasticity is also dynamic, since the edge of the VC former is exposed to the intense heat of the VC as it heats and cools under dynamic conditions. This is a subtle argument for high-efficiency loudspeakers, since less of the amplifier power is wasted in heating the VC (typical audiophile speakers are less than 0.5% efficient, leaving 99.5% of the amplifier power in the small VC assembly). If a speaker is ten times as efficient, for a given SPL at the listening position, the VC sees ten times less heating.
This is the reason that small, not-visible production changes can have a large effect on the sound and measurements of a given loudspeaker - and most driver vendors make production changes of this type all the time, to simplify production, improve QC results, and lower costs. They may look exactly the same, but sound quite different. Look at the glue that attaches the VC former to the cone, the glue that adheres the VC itself to the former, the location of the dustcap relative to the cone/VC connection, and the symmetry of the lead-wires going from the frame to the VC. All of these things affect the measurements and the sound of a driver.
But paper is a terrible material in other ways. Over time, it tends to burn and carbonize from the heat of the voice coil, which eventually leads to catastrophic failure. Since the VC, cone, dustcap, surround, and spider are integrated, the entire assembly must be replaced when the VC former crumbles and fails. Reconing a speaker essentially makes it a new loudspeaker, not something you want to do once the original paper formulation becomes unobtanium.
Kevlar, Kapton, and Nomex all sound different from what I understand, with very different mechanical properties, which in turn affects the sound of the loudspeaker.
Going deeper, the glues used to attach the VC former to the cone have a quite substantial effect on the sound as well. You can't say "Kapton sounds like this when combined with a paper cone" without discussing the sound of the different glues as well, since all glues have a degree of elasticity. The elasticity of the glue has a strong effect on the reflection between VC former and the cone - but if it's too elastic, it may fail under strong acceleration, or not have the right thermal characteristics.
The elasticity is also dynamic, since the edge of the VC former is exposed to the intense heat of the VC as it heats and cools under dynamic conditions. This is a subtle argument for high-efficiency loudspeakers, since less of the amplifier power is wasted in heating the VC (typical audiophile speakers are less than 0.5% efficient, leaving 99.5% of the amplifier power in the small VC assembly). If a speaker is ten times as efficient, for a given SPL at the listening position, the VC sees ten times less heating.
This is the reason that small, not-visible production changes can have a large effect on the sound and measurements of a given loudspeaker - and most driver vendors make production changes of this type all the time, to simplify production, improve QC results, and lower costs. They may look exactly the same, but sound quite different. Look at the glue that attaches the VC former to the cone, the glue that adheres the VC itself to the former, the location of the dustcap relative to the cone/VC connection, and the symmetry of the lead-wires going from the frame to the VC. All of these things affect the measurements and the sound of a driver.
Last edited:
It varies from place to place, but a lot of houses run a smaller projector for the ads - like they used to run a slide projector for the ads. Don't know why they do that. I'm long out of the cinema projection biz, I do live events now.
Digital isn't bright enough, yet, but the lack of film weave, scratches, dirt, bad splices, faded prints, misalignment, edge fall-off, etc. is a blessing.
Last edited by a moderator:
Hi Lynn
I have read your opinion on the tonal quality of ceramic, alnico and field-coil drivers, but what do you think of neodymium magnets? Are they much the same as ceramic? Thanks.
The British author John Watkinson has written on this subject in a series of articles called "Speakers Corner" appearing in Electronics World several years ago. I would have the article, but not to hand. His theory is that ferrites, being insulators, allow the magnetic domains to move about when subjected to forces of attraction/repulsion as work is being done moving the cone. He further suggests that this movement is not benign but rather granular/quantum. The end result is that the sounds produced are accompanied by noise sidebands, leading to the conclusion that if you are listening to a CD on a ferrite magnet speaker you are effectively not hearing what 16 bit resolution is capable of! He did concede that ferrite magnets do have uses including holding messages on the 'fridge door and picking up the metal turnings from the lathe when making experimental pole pieces.
Of course the other thing that characterises ferrites is the unpredictability of their magnetic properties. I remember being attracted to ferrite cored tape heads in the 1970's mainly in relation to their hardness and long life. A conversation with someone more knowledgable on the subject soon convinced me that the hardness was the only thing they had going for them. It was difficult to control the gap width and the inductances of a multitrck head could vary from 150 mH to 600 mH.
Keith
In any magnetic material, whenever magnetic domains change orientation, there is always hystersis, and they reorient in unsmooth jumps and cascades. The degree to which the magnetic materiel is electrically conductive will allow the formation of partially shielding/cancelling eddy currents. However, eddy-current formation is a phenomenon that increases with frequency, so they offer little help in the mid/bass. Here's a simulation of what I am talking about, using the example of copper shorting rings.
IMO, it's useful to think of a magnetic motor as a transformer. Like a transformer, it should be designed to operate over a certain bandwidth and within a certain power range, so that its operating point on the BH curve is in the most linear range. And the more power it has to handle, the less ideally it will behave. This is one reason why high-efficiency drivers can sound so good--there's less voice-coil flux pushing the magnetic material around to create a given SPL.
Excellent link, Rob.
JBL engineers compensated for poorer flux-modulation performance of ferrite vs. AlNiCo by adding back in a path for eddy currents to flow--a shorting ring--that AlNiCo magnets provided natively with their higher conductivity. The net result works very well (except at lower frequencies, as I mentioned, where AlNiCo wouldn't fare much better in terms of flux modulation).
A "theory" ("conjecture" is a more appropriate term) is nice and all.
But without measurement data and listening tests, who cares? Without data, it's just someone spouting off.
Don't forget "the power of statistics". When something is found good statistically, let the theorists find the reason why. We just enjoy the "reality".
I've enjoyed the ME2 version for quite awhile now, (bi-amped). Several years ago I started noticing some "leanness?" in the mid-bass. I ran a sweep, sure enough a drop-off/hole in the FR at 95-100 hz. I ran each P13 and thought they were mislabled MH's!! FS has risen to 102 hz to 112 hz on the four. I would guess this is caused by the surrounds aging, (brush a finger across the surrounds and they feel like an old pencil eraser). Seems like any "treatment" would cause it to completely degrade. Any ideas?
For now I moved the x-over point to 150 hz.
