• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Diminishing returns in the higher end

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,051
Likes
12,149
Location
London
Ideally you should compare loudspealers in your own room, switching instantly ( whilst not perfect) is the most valid form of comparison.
Keith
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
I'll throw my 2 cents into the wishing well.

I design drivers and finished systems and have since 1989.

Was a pretty accomplished musician years ago and still have perfect pitch. I notice things. And also can get fooled. Go figure on that!

So diminishing returns.

Not flogging any brand is number one.

Number two is most people listen with their eyes. That's a sad consequence really.

Importance of what you listen to.

If you like small ensemble jazz or classical, not to loud. Well there are a lot of speakers that do this exceedingly well. Even under $500. I have been humbled quite a few times working on a uber driver that is over $200 ex factory chasing changes that are noticeable. And then listened to very inexpensive drivers, I mean under $20 a piece that really sound pretty good!

No replacement for displacement.

If you want to go low and want it loud. There are precious few 6 or 8 inch sized drivers that have the X-max to do it justice. And that sums up the preponderance of available expensive high end loudspeakers. A few small drivers that are ok to about 45 hertz. If you are looking for lower you need something more.

Why do Ultra highend companies not publish measurements?

Many times they are that different from a great many other speakers. They are selling sizzle, not the steak.

Some other things to take note of.

Loudspeaker driver motor type does make a difference.

For one client we have gone from overhung small gap long coil, to XVL^2 to underhung short coil long gap. What does this mean? The area that the magnetic energy is concentrated in is either very small as in an overhung motor or very large as in an underhung motor.

What are the differences?

Distortion. There are distortions that are inherent in the overhung motor type. That are not present in a properly designed underhung.

Previously mentioned client called me up at the first listen to the underhung and told me something I will never forget.

He used the same piece of music to critique his speakers for over 30 years. Not only one, but this one was his litmus test. What blew him away was somebody moved. He turned to see who it was. Nobody there. It turned out to be a music stand was bumped and picked up by the recording mics.
Incredulous was I. So I asked. Same recording on the previous model? And did you hear it? He had an A/B system setup. And a blind switch operator. He could hear it on the new drivers, but not the old XBL^2 drivers.

Masked. The finer details were masked by nonlinearities that are a part of the relationship between the voice coil acting as an electromagnet and the permanents magnetic fields in the dual gap XBL motor. These distortions are there in all drivers. But are less there in underhung driver motors. There are quite a few papers that demonstrate this very effectively. So why don't we see this type of motor used more? Cost. Plain and simple. It costs a lot to make a good underhung motor versus a overhung motor that will essentially do the same thing. And loudspeakers are all about profit. To truly understand the cost of what is inside even the most expensive loudspeaker systems will shock you. Magico is a great example. Morel and Accuton make their drivers. And they are Ok drivers. But take them apart and compare them to many others and there is not that much difference. There is no truly secret sauce. There is good engineering. Sometimes the expense and the engineering meet and shake hands. And it works. Sometimes you are being sold the sizzle. Close your eyes and listen. You might be a little surprised.

Speakers are quite simple. Play with magnets and like poles repel, different pole N,S attract. Nothing magical about them. Your amplifier delivers voltage that is sent to a coil of wire that creates a magnetic filed and depending on the polarity of the voltage the cone moves forward or reward.

The biggest differences are what happens to the static magnet field in the motor. The one that is energized by the Ferrite, or Neodymium magnet.
Most amplifiers are quite capable of sending an acceptably low distortion signal. And most loudspeakers are actually pretty good at reproducing that signal with acceptable fidelity.

The diminishing return on my end of the loudspeaker is how well I can chase and exorcise those evil little distortions.

That being done. How I voice the crossover and apply the driver to their best effect.

2 cents over and out.
 

mohragk

Member
Joined
May 5, 2021
Messages
54
Likes
50
Location
The Netherlands
I'll throw my 2 cents into the wishing well.

I design drivers and finished systems and have since 1989.

Was a pretty accomplished musician years ago and still have perfect pitch. I notice things. And also can get fooled. Go figure on that!

So diminishing returns.

Not flogging any brand is number one.

Number two is most people listen with their eyes. That's a sad consequence really.

Importance of what you listen to.

If you like small ensemble jazz or classical, not to loud. Well there are a lot of speakers that do this exceedingly well. Even under $500. I have been humbled quite a few times working on a uber driver that is over $200 ex factory chasing changes that are noticeable. And then listened to very inexpensive drivers, I mean under $20 a piece that really sound pretty good!

No replacement for displacement.

If you want to go low and want it loud. There are precious few 6 or 8 inch sized drivers that have the X-max to do it justice. And that sums up the preponderance of available expensive high end loudspeakers. A few small drivers that are ok to about 45 hertz. If you are looking for lower you need something more.

Why do Ultra highend companies not publish measurements?

Many times they are that different from a great many other speakers. They are selling sizzle, not the steak.

Some other things to take note of.

Loudspeaker driver motor type does make a difference.

For one client we have gone from overhung small gap long coil, to XVL^2 to underhung short coil long gap. What does this mean? The area that the magnetic energy is concentrated in is either very small as in an overhung motor or very large as in an underhung motor.

What are the differences?

Distortion. There are distortions that are inherent in the overhung motor type. That are not present in a properly designed underhung.

Previously mentioned client called me up at the first listen to the underhung and told me something I will never forget.

He used the same piece of music to critique his speakers for over 30 years. Not only one, but this one was his litmus test. What blew him away was somebody moved. He turned to see who it was. Nobody there. It turned out to be a music stand was bumped and picked up by the recording mics.
Incredulous was I. So I asked. Same recording on the previous model? And did you hear it? He had an A/B system setup. And a blind switch operator. He could hear it on the new drivers, but not the old XBL^2 drivers.

Masked. The finer details were masked by nonlinearities that are a part of the relationship between the voice coil acting as an electromagnet and the permanents magnetic fields in the dual gap XBL motor. These distortions are there in all drivers. But are less there in underhung driver motors. There are quite a few papers that demonstrate this very effectively. So why don't we see this type of motor used more? Cost. Plain and simple. It costs a lot to make a good underhung motor versus a overhung motor that will essentially do the same thing. And loudspeakers are all about profit. To truly understand the cost of what is inside even the most expensive loudspeaker systems will shock you. Magico is a great example. Morel and Accuton make their drivers. And they are Ok drivers. But take them apart and compare them to many others and there is not that much difference. There is no truly secret sauce. There is good engineering. Sometimes the expense and the engineering meet and shake hands. And it works. Sometimes you are being sold the sizzle. Close your eyes and listen. You might be a little surprised.

Speakers are quite simple. Play with magnets and like poles repel, different pole N,S attract. Nothing magical about them. Your amplifier delivers voltage that is sent to a coil of wire that creates a magnetic filed and depending on the polarity of the voltage the cone moves forward or reward.

The biggest differences are what happens to the static magnet field in the motor. The one that is energized by the Ferrite, or Neodymium magnet.
Most amplifiers are quite capable of sending an acceptably low distortion signal. And most loudspeakers are actually pretty good at reproducing that signal with acceptable fidelity.

The diminishing return on my end of the loudspeaker is how well I can chase and exorcise those evil little distortions.

That being done. How I voice the crossover and apply the driver to their best effect.

2 cents over and out.

So what's your take on DSP to mitigate some of these distortions? I read a KEF paper where they explained applying a sort "inverse of the distortion" to there woofer drivers in real time, basically nulling that distortion.
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
So what's your take on DSP to mitigate some of these distortions? I read a KEF paper where they explained applying a sort "inverse of the distortion" to there woofer drivers in real time, basically nulling that distortion.

I'll say it like this. If you need to correct something why is this the case? I have worked with servo subwoofers. Usually relatively cheap drivers that are "corrected" to a desired frequency response. I have taken apart and thoroughly inspected Velodyne subwoofer drivers. They seem to go through periods of great drivers and then cheaper and cheaper until people catch on that they are not so great again. The servo circuit cannot make a bad driver perfect. It's a band aid. If you spend time on designing a very linear motor and on a very linear suspension a servo will have little valuable effect.

Lets jump to DSP. For crossovers, sure I love it, for Phase compensation, as in time compensation of the signals coming from multiple drivers awesome. In room correction it can work to a limited degree over a fairly useable area. Distortion compensation is a big a thing to do. As the effects that are easy to correct are the tip of the iceberg. If DSP is promoted as the cure for the ills of a loudspeaker why is it needed in the first place? Are the drivers being pressed into areas of operation that they were not designed for? If the loudspeaker designer woefully ignorant of what the drivers truly can and cannot accomplish? ( I could tell you hair raising stories on that front. ) I know I am throwing a lot of questions at you. But they are to this point. A well engineered driver is actually a pretty good reproducer of the frequency range that it is designed to reproduce. If it isn't then corners were cut and silicon and clever programmers are there to make you believe that something has been accomplished that most probably could have been done other ways. If the correction is modest then it will be useful. But all to many times the correction is ham fisted and sounds terrible when compared to a cleanly reproduced system.

So to the point. Jusicous amounts of correction is a good thing. Like spices when cooking it's easy to add to much.
 

Tokyo_John

Active Member
Forum Donor
Joined
Mar 6, 2021
Messages
214
Likes
289
I recently visited a shop in Tokyo-Akihabara, Dynamic Audio, which had ~7 floors of high-end gear. The nice fellow on the first floor explained that the prices go up as one ascends to the next floor. On the 2nd floor they were already into ~$10k/component prices, which was well beyond my price range, and I only made it to the 4th floor at which prices were getting into $50k-$100k/component levels. While I am curious to know what is on the top floor, I didn’t feel comfortable ascending further, as I was already way outside of my element. The systems started off having some familiar names, focal, magico, TAD, PS Audio, but as I ascended there was a point at which there was nothing I recognized...the domain of tiny specialty companies that only make a few items every year, at enormous cost for each. They had systems playing on each floor, sometimes wired up by cable bundles as thick as my arm, and of course everything sounded good subjectively. But is it even possible to get enough out of an audio system to justify spending $100k or more? That is for each individual to decide...in modern times of course there are some who have more money than they know what to do with, and they’ll surely spend it without caring about the details so much. This is a rarified market, if you know how to sell platinum-coated snake oil, there is a lot of money to haul in.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
I recently visited a shop in Tokyo-Akihabara, Dynamic Audio, which had ~7 floors of high-end gear. The nice fellow on the first floor explained that the prices go up as one ascends to the next floor. On the 2nd floor they were already into ~$10k/component prices, which was well beyond my price range, and I only made it to the 4th floor at which prices were getting into $50k-$100k/component levels. While I am curious to know what is on the top floor, I didn’t feel comfortable ascending further, as I was already way outside of my element. The systems started off having some familiar names, focal, magico, TAD, PS Audio, but as I ascended there was a point at which there was nothing I recognized...the domain of tiny specialty companies that only make a few items every year, at enormous cost for each. They had systems playing on each floor, sometimes wired up by cable bundles as thick as my arm, and of course everything sounded good subjectively. But is it even possible to get enough out of an audio system to justify spending $100k or more? That is for each individual to decide...in modern times of course there are some who have more money than they know what to do with, and they’ll surely spend it without caring about the details so much. This is a rarified market, if you know how to sell platinum-coated snake oil, there is a lot of money to haul in.
Man. I wish I had a store like that near me.
 

PierreV

Major Contributor
Forum Donor
Joined
Nov 6, 2018
Messages
1,437
Likes
4,686
Ideally you should compare loudspealers in your own room, switching instantly ( whilst not perfect) is the most valid form of comparison.
Keith

Ideally, yes, I agree. However, in practice, that isn't doable because

1 - ideally, both speaker pairs should be at their optimal location which may or may not conflict with the other pair.
2 - ideally, they should be perfectly level matched, possibly - to satisfy the requirements of some of our contributors - with a voltmeter, which isn't going to happen with active vs passive.
3 - ideally, the test should be blind in order to remove my own bias on the device sizes.
4 - ideally, all rooms and use cases should be tested.

Assuming that the ideal is somehow miraculously achieved, the potential customer could then hire a statistician and obtain his own Harman-like preference scores. If the potential customer knows a bit about stats, he will ideally not use the general Harman preference score because it is a loose, multiple regression model(*)

Realistically, this is not going to happen. You know it and that is why you offer auditions in your room for potential customers (as do all dealers). I know it which is why I keep going to listen to speakers (although not often recently).

Are you refusing to sell speakers to people who listen to them at your place and immediately want to buy them? I am guessing this isn't the case :)


(*)

1620282189770.png
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,051
Likes
12,149
Location
London
I believe but stand to be corrected that level matching, is probably best conducted with a steady state tone and an SPl meter.
I do conduct the odd unsighted comparison of the client requires, customers are always welcome here if they want to whittle down between models, but eventually hearing them in your own room is the ideal.
Generally a decision between speaker models is formed pretty quickly, and again generally the benefits of contemporary active designs are strikingly obvious in the clients home, although that does depend of course upon the clients current system.
Keith
 

FeddyLost

Addicted to Fun and Learning
Joined
May 24, 2020
Messages
752
Likes
542
I really don't think any Speaker out there can perform better than the Dutch & Dutch 8C, and they're 12k$ a pair.
However, with that said, the Dutch & Dutch 8C distortion is not pretty, and I wonder how big a Speaker must be to have a similar radiation pattern with lower distortion. Nevertheless, reviewers rarely bring this up, which makes me question how much distortion can humans hear when everything else is 'just right'.
Different listeners, different program material ... and then you have to understand that distortion ( i mean all types of distortion) profile is one of the components for "sound of the brand".
And then we'll come to point that if we make some "good enough" speakers, most of customers will be satisfied. But for enthusiast, that can pay extra over "standard solution" we'll have to make something special, not universal and significantly better than standard. It's the PITA in my opinion.

So what's your take on DSP to mitigate some of these distortions? I read a KEF paper where they explained applying a sort "inverse of the distortion" to there woofer drivers in real time, basically nulling that distortion.
DSP is good when you need to get rid from 50% of HD from small woofer and your subwoofer will not grunt until voice coil kicks into magnet.
But it will never allow one to get really low distortions with bad motor and significantly non-linear components in midwoofer and higher. One will have to design something correct first...
 

mohragk

Member
Joined
May 5, 2021
Messages
54
Likes
50
Location
The Netherlands
If DSP is promoted as the cure for the ills of a loudspeaker why is it needed in the first place? Are the drivers being pressed into areas of operation that they were not designed for? If the loudspeaker designer woefully ignorant of what the drivers truly can and cannot accomplish? ( I could tell you hair raising stories on that front. )

DSP is good when you need to get rid from 50% of HD from small woofer and your subwoofer will not grunt until voice coil kicks into magnet.
But it will never allow one to get really low distortions with bad motor and significantly non-linear components in midwoofer and higher. One will have to design something correct first...

Of course, it's not a cure all. You have to start with good drivers. But I feel that DSP can help elevate a great driver to an excellent driver.
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,780
Location
Oxfordshire
I'll say it like this. If you need to correct something why is this the case? I have worked with servo subwoofers. Usually relatively cheap drivers that are "corrected" to a desired frequency response. I have taken apart and thoroughly inspected Velodyne subwoofer drivers. They seem to go through periods of great drivers and then cheaper and cheaper until people catch on that they are not so great again. The servo circuit cannot make a bad driver perfect. It's a band aid. If you spend time on designing a very linear motor and on a very linear suspension a servo will have little valuable effect.

Lets jump to DSP. For crossovers, sure I love it, for Phase compensation, as in time compensation of the signals coming from multiple drivers awesome. In room correction it can work to a limited degree over a fairly useable area. Distortion compensation is a big a thing to do. As the effects that are easy to correct are the tip of the iceberg. If DSP is promoted as the cure for the ills of a loudspeaker why is it needed in the first place? Are the drivers being pressed into areas of operation that they were not designed for? If the loudspeaker designer woefully ignorant of what the drivers truly can and cannot accomplish? ( I could tell you hair raising stories on that front. ) I know I am throwing a lot of questions at you. But they are to this point. A well engineered driver is actually a pretty good reproducer of the frequency range that it is designed to reproduce. If it isn't then corners were cut and silicon and clever programmers are there to make you believe that something has been accomplished that most probably could have been done other ways. If the correction is modest then it will be useful. But all to many times the correction is ham fisted and sounds terrible when compared to a cleanly reproduced system.

So to the point. Jusicous amounts of correction is a good thing. Like spices when cooking it's easy to add to much.
For me the main benefit of DSP and cheap high powered amps in subwookers is compensation for LF roll off with either a "too small" cabinet or open baffle.
Yes the driver still has to be excellent to take the power and displacement but a smaller sub with the performance of a big one is a benefit for many.
 

Jim Matthews

Major Contributor
Forum Donor
Joined
Mar 25, 2021
Messages
1,051
Likes
1,286
Location
Taxachusetts
A more fair comparison would be a tally of component costs prior to comparison. The products that perform well for little money, using readily available parts would define value.

Those that command high prices for exotic materials with bespoke assembly methods reach the other extreme.

Plot ELAC on the value end YG Acoustics on the other
 
Last edited:

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
A more fair comparison would be a tally of component costs prior to comparison. The products that perform well for little money, using readily available parts would define value.

Those that command high prices for exotic materials with bespoke assembly methods reach the other extreme.

Plot ELAC on the value end YG Acoustics on the other


Also the amount of realistic R&D involved is worth noting. As is limited manufacturing quantities. The small runs are usually made by very skilled people that are not cheap to employ. And the people with enough experience to actually make the cutting edge drivers, the woofers and the tweeters etc are also not an insignificant cost.
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
Of course, it's not a cure all. You have to start with good drivers. But I feel that DSP can help elevate a great driver to an excellent driver.


I'd argue against that statement. If you are trying to linearize a driver at what point do you stop? And can a DSP algorithm truly keep up with all the potential situations of correction required? It is an engineering exercise to be sure. But it is flying a large head wind against the very precepts of highend audio. To use the best engineering to get the best reproduction. It's a stopgap measure really.
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
Different listeners, different program material ... and then you have to understand that distortion ( i mean all types of distortion) profile is one of the components for "sound of the brand".
And then we'll come to point that if we make some "good enough" speakers, most of customers will be satisfied. But for enthusiast, that can pay extra over "standard solution" we'll have to make something special, not universal and significantly better than standard. It's the PITA in my opinion.


DSP is good when you need to get rid from 50% of HD from small woofer and your subwoofer will not grunt until voice coil kicks into magnet.
But it will never allow one to get really low distortions with bad motor and significantly non-linear components in midwoofer and higher. One will have to design something correct first...


I agree 100% With the DSP statement. If a driver is exhibiting a great amount of harmonic distortion you have three potential sources. The soft parts, the surround and the damper or spider, or the cone. Problems like this are well understood. So are the solutions. It takes an experienced engineer to know how to identify these problems and how to mitigate them. If inferior tranducers like these get into a product you have a number of reasons. Number one is price point. Poor engineers get the short end of the stick all to often ahead of what the marketing guys feel they can successfully sell. And conversely sometimes the engineers run amok and there is a product that is very complicated and really solves nothing special in particular.
 

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
It depends on the type of music one likes, the size of room, prefer loudness level, etc.

My setup in my small computer room is pretty satisfying. But it lacks big sound stage illusion due to room limitations. If I move my computer room setup to my much bigger living room, it can provide bigger stage illusion. But, as soon as I play electronica with lots of sub bass loud, the setup simply loss it's magic and struggle with its ability to cleanly "move" me. This is where bigger setup with lower distortion is needed.....Bigger and lower distortions typically means more expensive.....
 
Last edited:

audiofooled

Addicted to Fun and Learning
Joined
Apr 1, 2021
Messages
530
Likes
588
IMHO, competently designed loudspeaker must be able to deliver any type of music, any time, with neutral presentation. Maybe enough of headroom and bass extension, where size matters, may be expensive... Room (size, modes, distance to listening position, etc.) being the very important part of the equation, sometimes a pair of very expensive towers may not be enough. A bit less expensive ones but reinforced with a couple of good subs may perform better in some situations.
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
IMHO, competently designed loudspeaker must be able to deliver any type of music, any time, with neutral presentation. Maybe enough of headroom and bass extension, where size matters, may be expensive... Room (size, modes, distance to listening position, etc.) being the very important part of the equation, sometimes a pair of very expensive towers may not be enough. A bit less expensive ones but reinforced with a couple of good subs may perform better in some situations.

BINGO!

That is what I try to accomplish. Many well documented psycho acoustic tests done in North America, Europe, Japan and in Asia have verified that clean and accurate sound is readily recognizable to people of all walks of life. They simply have to know and understand what a reference true to life sound is in the first place. That has been a perennial problem for me. Some of the simplest things to do for myself working with new clients id to understand what they consider as being accurate sound. It is not that easy to do if all you have ever listened to is mono recorded pop music that is really really processed and produced.

Not saying that enjoying your tunes is wrong. Setup what ever turns your crank. But I am talking about what is true to life. And that requires true to life sounds. Midbass, a car door slamming. It's a huge transient that has frequency content that is very recognizable. Especially if it's your car door. A loved ones voice. Can't fool anybody there right. It sounds real or it doesn't. Low bass, in the right location down town or in an overpass bus stop a bus has astonishingly loud ultralow bass. Pay attention to it. Not that hard to grab a recording of it. We all have cell phones and a calibrated mic is not that difficult to plug into your cell phone. Heck Dayton sells one that is what $16 or $18 USD?

Attend a live unamplified ( key here ) concert. Understand what real room acoustics sound like. What a guitar or a trumpet sounds like. A drum set. They don't sound much like what I hear on recordings.

Anyway. Off of my soap box.
 

mwmkravchenko

Active Member
Joined
Nov 20, 2018
Messages
180
Likes
187
Location
Perth Ontario
I believe but stand to be corrected that level matching, is probably best conducted with a steady state tone and an SPl meter.
I do conduct the odd unsighted comparison of the client requires, customers are always welcome here if they want to whittle down between models, but eventually hearing them in your own room is the ideal.
Generally a decision between speaker models is formed pretty quickly, and again generally the benefits of contemporary active designs are strikingly obvious in the clients home, although that does depend of course upon the clients current system.
Keith


Level matching is best if you want to compare a loudspeakers. We all have phones that are more than capable of taking relative measurements. Absolute values are not important. Matching to a half db is a good idea. Or even better than a half db if you are patient. Having the same room position is also very important. Rooms influence the sound much more than a few points of a db. And even placing side by side can change response. Having loudspeakers in the room that are dead but ported in the right places can even influence the rooms sound. Ever hear of resonant bass traps? That's what an un-powered loudspeaker can become in the right proximity and location in a room.
 
Top Bottom