How Do You Buy High-End Speakers?

[AcuDefTechGuy]

And the moral of this story is?

For one thing, you never know what the future holds. Your needs may change. Your preferences and ideas may change. You may buy a new house.

You can read all the reviews, audition all the speakers for 8 hours each, and you still won't know until you have lived with the speakers for a long time.

I think it may just go back to the OP's idea or definition of a "perfect" system - before others have influenced or changed his mind along the way.

Does the OP have an explicit and decisive definition or goal of his perfect dream system?

 

[walter duque]

I am sure, since it seems like the high dollar stuff has a lot of wiggle room rite at the showroom alone, I am good friends with a couple shop owners but the stuff that is in there doesn't really interest me, and most of the time a closeout or demo will be available online for less than they paid...
Now I have listened to some really expensive systems, I have heard the focal utopias, I have heard a few sets of speakers that cost twice what my duramax powered super cab cost me and the thing is when comparing what I heard to what I have in my parlor for a few thousand dollars, I couldn't justify spending $10K on them systems never mind $100K+ just on the speakers...
Heres another thing, I am happy with my budget systems, sure I can buy something insane, but rite now I can box what I have up and start all over, say replace my evos with ascends and hsu subs, my csb1s with phils and sb2000's, my tektons with phil 3's, ect... I could throw the speakers I own away and still only spend $7000 replacing them in 3 systems, lol... Not a big deal... I like being able to swap and play around with the systems, I don't take it too serious, other wise I would put curtains over the glass, I would put rugs down, treat the corners, treat the walls, ect, but thats not what I am looking for...

Now don't get me wrong I am not saying that I don't like the sound and looks of say the sonus Aida's over my evo-2's lol, but I don't like them 120 times more, I got to listen to some high end sonus towers in Tennessee not that long ago, powered by a pair of levinson mono blocks {of course}, with a tube preamp {I forget the number maybe a c2300} and matching cd player, they were awesome, the look, the sound, the power, I love subs and for some reason even the bass was great to me... But that system was around $150K, so I don't think I could imagine the sound and exotic veneer it would take for me to spend that kind of money since I could buy a simple $3000 system and a G series, lol... Now I know that is out of reach for a lot of people, and $10K is in reach for a lot more, but that is my point, at $10K it is even less impressive.. The salk scapes would be my pic for around 10K and I can't talk myself into that when I hear the difference... I am all about music, and I like the entire package, I wanta good looking speaker, I want it to sound good, I want it to be worth almost what I paid when I find the next set I want...

I did listen to the Utopias (the big ones) he was asking a little under 40k for them, demos of course with the wooden inserts for covers. I regret not picking up these Salons. They just came in on a trade in and I could have bought them for 3.5k. Great deal, but way to big and heavy for my place. Can't imagine dragging these things up to the 3 rd floor. Besides these where in like new condition.

 

[AcuDefTechGuy]

If you are going with a sub, then cross it over around 130 and use a system with two 6.5" drivers. Yes I recommend a higher crossover for the simple reason that it will clean up your midrange and leave the little driver to do what they do best.

I believe you and Nathan recommend using a higher crossover point than most people. And I tend to agree because I set my XO @ 120Hz.

I will see if I can go to 130Hz to be like you.

 

[AcuDefTechGuy]

The easiest way might be to just find the best deal among the top speakers, buy them, then worry about "upgrading" only if I happen to hear something I like better.

Actually after you buy it, just enjoy all the music and movies and don't audition anymore speakers. Just enjoy life after that.

 

[Swerd]

I'll throw in my two cents…

So can you tell these things by looking at a graph?

Yes with a little training.

If the graph is basically a straight line. Look somewhere else. There is no speaker that has a frequency response that can be characterised as a straight line.

Do you see wiggles a squiggles? Good they are honest. It takes a peak or dip of at least 3 decibels over a quarter octave to be noticeable. An octave is a halving or a doubling of frequency. So 64 to 128 is an octave, 256 hertz to 512 hertz is an octave and so on. Most reputable graphs have the frequency scale on the bottom going left and right. Lower is on the right, higher is on the left. And sound pressure level vertically. You can figure out the vertical divisions by looking at the marker numbers. Generally they are in 5 decibel increments.

Our ears are most sensitive between 1000 and 4000 hertz in general terms. This range is also where all the cues to where something is in 3 dimensions are. So that is a critical region for your eyes to look over a speaker response graph.

Above 4000 hertz up to about 14 000 hertz is the limit to what most instrument produce with any degree of power. And most people don't really hear much above that in the first place.

Welcome to Audioholics! And thanks for your excellent post. As I read your comments and observations, I found myself nodding in agreement.

I especially liked your comments about what to look for in a frequency response graph.

You play the French Horn. I've noticed that this instrument often seems to play notes around the crossover frequency in 2-way speakers. If I play a recording with some prominent French horn, and the speaker seems to swallow that sound or make it seem distant, it suggests an audible dip occurs in the crossover range.

 

[Swerd]

herbu

If you're in Raleigh, NC, look up Selah Audio. Its nearby in Garner Selah Audio.

I'm not too familiar with their designs, but the owner has a reputation as a good speaker designer. It might be worth your time to visit there.

 

[AcuDefTechGuy]

...I explained to him a lot of guys talk on the internet and then people think its law, like "cd's are much better than mp3's" they may have better sq, but I can not hear it, and if it is its not enough to make me ignore how convenient the ipod is... So that was lesson one, then he said "I just don't have the money to buy a system like that", so I said "what do you think that cost?" he says, "I don't know, $10K"...

...I like the feeling I get knowing I didn't spend $15K on a system and it sounds so good to me...

I stream ALAC files on my iPad to my Denon. They are not much bigger than MP3 files.

Diminishing returns is the moral here when it comes to expensive speakers and gears?

I think the question is, is the diminishing return "significant" enough for the buyer to spend that money? For most people, the answer is "No". But for a few, it is "Yes".

 

[Irvrobinson]

I'll throw in my two cents.

I am not a speaker system or driver designer, but I have owned systems with many of attributes you've mentioned, both at the system and driver level. The only thing I have concluded is that as with so many different and unrelated products, outstanding engineering and execution is a more determining factor of ultimate utility and quality than generalizations about product architecture and component design choices. To take a completely unrelated example, high performance BMW brake systems used to violate just about every rule one could think of for building a great automotive brake system. Sliding calipers rather than fixed calipers, huge single pistons rather than a multiple smaller pistons, non-drilled rotors, aggressive pad compounds that dust like the Peanuts character Pig-Pen, and the pads aren't beveled so they squeal when cold. If you were judging them just by the components and the specs you'd call those brakes crude and cheap. On the other hand the BMW brakes stop as well as those fancy Brembos that Porsche uses, and they usually have better pedal feel, and they can be wonderfully progressive and subtle. BMW has been moving to fancy fixed-caliper designs with drilled rotors, but I'd guess only because brakes have become a fashion statement lately. I've often wondered, are metal cone drivers a similar fashion statement?

When I bought the Legacy Focus in 1996 I remember my serious speaker DIYer friend saying the Focus had no right to sound so good, and admitted that by just looking at them he would have never bothered to listen. Not that the original Focus is so awesome by today's standards, but it had rather remarkable price-performance in its day. Though I never did get them to throw an image to my satisfaction, I would have missed out on over a decade of fun if I had followed generalizations about drivers and design rather than just listening.

I think my point is that while I agree with a lot of what you posted, I think dwelling on individual aspects of design, especially things like driver materials, distracts buyers from what's important, the overall product execution and performance. I don't really care what the cone material is if the speaker sounds like live music.

 

[herbu]

If you think this is of any value ask questions on what is not making sense or on what I may have left out.

Ya'll's response(s) have identified a bunch of areas I need to research, and a lot of questions. Please let me start w/ drivers.
Are more drivers better?
Intuitively, I think yes. Especially when I see speakers like these:
RBH

, and Selah

and the ones you're currently having built.

But then I see highly recommended speakers like these:
Philharmonic

, Salk

and Ascend

.

So what is the advantage of more drivers? Simply SPL?

 

[GranteedEV]

Intuitively, I think yes. Especially when I see speakers like these:
RBH

, and Selah

and the ones you're currently having built.

But then I see highly recommended speakers like these:
Philharmonic

, Salk

and Ascend

.

So what is the advantage of more drivers? Simply SPL?

In the case of the RBH stuff, SPL is probably the main advantage. There are others, but of course there are gonna be tradeoffs to go with it.

In the case of the curved Selah CBT you posted, it's different. Yes, max SPL is very important, but you're also getting a very different pattern control - a very different room interaction. What does that mean? Well, the measurements show just how flat it really is IN-ROOM!

There is a smaller CBT, the CBT-36, which retails for about 6 or 7 grand (2 grand if you DIY ) made by Audio Artistry.

 

[AcuDefTechGuy]

So what is the advantage of more drivers? Simply SPL?

I think you're supposed to get increased efficiency and "unparalleled dynamic range and seemingly infinite soundstage".

To me, I get a little more details and resolution in SQ. Also soundstage and imaging. I don't play loud enough to test the dynamics.

But the proof is in the actual sound and speaker measurements.

 

[Swerd]

So what is the advantage of more drivers? Simply SPL?

Higher SPL AND lower distortion. Two drivers would have to move less to produce as much SPL as one driver. The lower the driver excursion, the lower the distortion.

Of course, there are limits to this generalization. The largest benefit probably comes from going from one driver to two, as in an MTM. More mid range or mid woofers add as much trouble as benefit.

Mark Kravchenko pointed out that:

I have found that a stack of three or four drivers and then a tweeter is a no go no matter what crossover voodoo has been done. They never quite sound right.

Same goes for a whole bunch of little midranges masquerading as woofers.

All this can very quickly go from a generalization that seems correct, to a case where details can make large differences.

 

[GranteedEV]

For tweeters. There are all kinds. Soft domes, hard domes, planars, ribbons. They all have their good points and bad points. For what it's worth here is my 30 plus years worth of listening.

In order of what I find most musical. Best is at the top.

Planar they are not all equal
soft dome some are better than others.
ribbon some are pretty good, but they all require a midrange to play awfully high up until they can take over.
metal dome don’t like them they tend to have a very distinct sound that I find irritating.

For midwoofers:

Paper
Poly Polypropylene
Metal Cone
Carbon Fiber
Kevlar

Simple rules of thumb. If it is absolutely stiff it will ring like a bell at some frequency. No matter what you read that is always the case and will ever be. You canna change the laws of physics.
Paper is not just paper and has not been for many years. The best cones are made of paper.
Kevlar is a weird one. It literally screams. and no matter what you do in the crossover it is still there.

Mark, I agree with pretty much everything you said and thank you for the awesome post - just not this specific part. I'm not denying your experiences, just suggesting that they're probably too isolated to be broad and general. I implore people to give things a try. I mean let's look at the woven-Kevlar mid from B&W. It's pretty much a piston right up until its breakup, and its breakup is as well-damped as you could possibly get. It displays better behaviour even at breakup than most paper drivers. Does that mean you should use it that high? Course not (they do though). But it certainly doesn't "scream". Likewise with metal dome tweeters - I'm sure some do, but just as much some don't, especially not the modern aluminum / magnesium / ceramic stuff like the Transducer labs units.

 

[Dennis Murphy]

Mark, I agree with pretty much everything you said and thank you for the awesome post - just not this specific part. I'm not denying your experiences, just suggesting that they're probably too isolated to be broad and general. I implore people to give things a try. I mean let's look at the woven-Kevlar mid from B&W. It's pretty much a piston right up until its breakup, and its breakup is as well-damped as you could possibly get. It displays better behaviour even at breakup than most paper drivers. Does that mean you should use it that high? Course not (they do though). But it certainly doesn't "scream". Likewise with metal dome tweeters - I'm sure some do, but just as much some don't, especially not the modern aluminum / magnesium / ceramic stuff like the Transducer labs units.

Also, there are a number of ribbon tweeters that can be crossed well within the comfortable operating range of a midrange. Fountek and RAAL are two. As for woofers, my least favorite is pure poly. I've done controlled tests with the most expensive poly units available, and they just don't have the detail of many other materials. But I agree with EV that woven composite poly cones can perform very well, as can metal domes and cones. I've never had any trouble designing crossovers that effectively suppressed their resonant modes.

 

[mwmkravchenko]

You are correct Denis.

And I know you make some great speaker systems. And your opinions hold weight with me.

I didn't make a post to be taken as the be all and the end all of audio knowledge. It's there as a bit of a guide. A bit of a stirring of the pot to get things in motion.

As has been stated good engineering is the first and foremost part of speaker design.

I have built and produced metal coned woofers that were extremely well received by some designers that love metal coned drivers. And converted the opinions of a few designers that hated them previously. It took a year of design a prototyping to get it right. So I am not categorically against metal coned drivers. My driver has a break up mode as do all others. It's what is done to mitigate that breakup that matters. Work done on the driver design side is very valuable in terms of what can be done after, on the crossover side.

Anyone that has done work using both a passive crossover system and duplicating it with an active system knows what I am talking about. Crossover components do colour the sound. Cross over components bleed away efficiency. And they restrict the designer capabilities because the more you want to correct and contour, the more you tend to colour the sound. Going active mitigates all of that.

I have listened at length to RAAL's. I'm indifferent. They can sound pretty good when done right. They fragile, and very revealing of crossover work. And they can sound awefull in the wrong hands.

And I have heard metal dome tweeters that I liked. Just not so many of them.

A case in point is the Neo8 driver. I listened to a pair every day for 5 years. They have their strengths and their weaknesses. And they have a nasty resonance that is hard to tame an one I can hear everytime I go back to them, even with a notch filter. But I enjoyed listening to them for a long time. They are very musical sounding drivers. And that resonance is high enough up in frequency that you don't usually hear it.

In the pictures Herbu posted there are some excellent examples of loudspeaker engineering that use multiples of smaller drivers. There is even an example of how not to mount your drivers. (RBH)

Regarding Kevlar drivers in particular the one B&W uses. I have been to the factory that makes the cones for them, I use cones from that vendor for a few designs I make. And in this matter I stand firm. Kevlar cones have some of the nastiest breakup modes I have ever listened to. And the idea that you can bury them in a crossover work only goes so far. It is almost impossible to do so in a two way system. That I hear in every small B&W speaker I have listened to. And it is difficult but not impossible in a three way system to tame the little beasts.

If any of you are attending the RMAF drop in on the Funk Audio room. I'll be there and it will be interesting to discuss this in person.

 

[Dennis Murphy]

You are correct Denis.

And I know you make some great speaker systems. And your opinions hold weight with me.

I didn't make a post to be taken as the be all and the end all of audio knowledge. It's there as a bit of a guide. A bit of a stirring of the pot to get things in motion.

As has been stated good engineering is the first and foremost part of speaker design.

I have built and produced metal coned woofers that were extremely well received by some designers that love metal coned drivers. And converted the opinions of a few designers that hated them previously. It took a year of design a prototyping to get it right. So I am not categorically against metal coned drivers. My driver has a break up mode as do all others. It's what is done to mitigate that breakup that matters. Work done on the driver design side is very valuable in terms of what can be done after, on the crossover side.

Anyone that has done work using both a passive crossover system and duplicating it with an active system knows what I am talking about. Crossover components do colour the sound. Cross over components bleed away efficiency. And they restrict the designer capabilities because the more you want to correct and contour, the more you tend to colour the sound. Going active mitigates all of that.

I have listened at length to RAAL's. I'm indifferent. They can sound pretty good when done right. They fragile, and very revealing of crossover work. And they can sound awefull in the wrong hands.

And I have heard metal dome tweeters that I liked. Just not so many of them.

A case in point is the Neo8 driver. I listened to a pair every day for 5 years. They have their strengths and their weaknesses. And they have a nasty resonance that is hard to tame an one I can hear everytime I go back to them, even with a notch filter. But I enjoyed listening to them for a long time. They are very musical sounding drivers. And that resonance is high enough up in frequency that you don't usually hear it.

In the pictures Herbu posted there are some excellent examples of loudspeaker engineering that use multiples of smaller drivers. There is even an example of how not to mount your drivers. (RBH)

Regarding Kevlar drivers in particular the one B&W uses. I have been to the factory that makes the cones for them, I use cones from that vendor for a few designs I make. And in this matter I stand firm. Kevlar cones have some of the nastiest breakup modes I have ever listened to. And the idea that you can bury them in a crossover work only goes so far. It is almost impossible to do so in a two way system. That I hear in every small B&W speaker I have listened to. And it is difficult but not impossible in a three way system to tame the little beasts.

If any of you are attending the RMAF drop in on the Funk Audio room. I'll be there and it will be interesting to discuss this in person.

I think we'll have to agree to disagree on this one, although I do respect your opinion. I just haven't encountered the problems you're referencing. For example, the peak on the Neo 8 is at 15 kHz, and it disappears rapidly as the listening axis moves up, as it would in a 3-way that uses it as a mid. Further, if you cross the Neo around 3 kHz with 4th order acoustic slopes, the peak is a non-issue. I just sense you're over-generalizing about drivers. I dislike some of the B&W Kevlar designs as much as you do. But I've had 3 people bring me their B&W's, and I've had no problem eliminating any brightness issues simply by lowering the crossover point and using appropriate acoustic slopes. The key is the crossover--and I think resistors, inductors, and capacitors introduce a lot less coloration than you're claiming. The real gremlin is an inadequate crossover that can't tame the raw response of the driver-not a properly designed, and possibly complex, filter.

 

[mwmkravchenko]

I think we'll have to agree to disagree on this one, although I do respect your opinion. I just haven't encountered the problems you're referencing. For example, the peak on the Neo 8 is at 15 kHz, and it disappears rapidly as the listening axis moves up, as it would in a 3-way that uses it as a mid.

Good to hear BG have moved up the resonance frequency. I asked Igor about that possibility. I was working with it in 2001 to 2005. And it was in a two way. So your application is a good example of proper engineering. Since we both have the same experience And you have tamed the problems I think we actually agree. And I have learned something new to boot. Thank you.

I just sense you're over-generalizing about drivers.

Yes I am. This is not a place to pick the greatest drivers and say that anything made with them is going to sound good. Because that simply is not the case. Your comments about the B&W are a case to prove that true. They have a particular sound. And I agree that it can be corrected. What baffles me is why they leave the B&W factory the way they sound? A good driver does not guarantee that you will have a good sounding loudspeaker. It's the voicing and crossover design work that make up a loudspeaker.

The key is the crossover--and I think resistors, inductors, and capacitors introduce a lot less coloration than you're claiming.

Yes they do. I'm not directly claiming that they introduce a great amount of colouration. I don't really believe that they do if properly chosen. What they do introduce is losses in the system. And what I really failed to say is that you cannot boost or easily contour with a passive system as you can with an active system. Nor is it easy to setup delays. Yes it is possible. But it is not the same thing as dialing in a delay and taking a measurement on an active system.

The real gremlin is an inadequate crossover that can't tame the raw response of the driver-not a properly designed, and possibly complex, filter.

Bingo!

P.S. Nothing in my posts is there to denigrate your work. I know people whose ears I trust who own and enjoy your products immensely. And you have solid logical design and execution of your products. When I have mentioned anything about them I have spoken about the NEO 8. And now it turns out my experience with the NEO 8 is dated and incorrect. So now I stand corrected and I thank you for this.

 

[GranteedEV]

Regarding Kevlar drivers in particular the one B&W uses. I have been to the factory that makes the cones for them, I use cones from that vendor for a few designs I make. And in this matter I stand firm. Kevlar cones have some of the nastiest breakup modes I have ever listened to. And the idea that you can bury them in a crossover work only goes so far. It is almost impossible to do so in a two way system. That I hear in every small B&W speaker I have listened to. And it is difficult but not impossible in a three way system to tame the little beasts.

Keep in mind, the Kevlar breakup of the 6" B&W speaker is centered around 3.5 to 4khz.

Now normally, you want this kind of breakup to be well below the level of the main sound. Even getting it 24db down in level on-axis can suffice off-axis since the driver is beaming.

You mention the B&W speakers sound horrible.

Well - they should. Their crossover point is above 4khz - and I believe in phase quadrature at that. So basically they are at most 3db down IF there's a notch on the breakup.

So of course they sound bad in that system. The system is stupid. It excites the breakup because they adhere to some wierd philosophies about keeping crossovers out of some magical frequency range and keeping them shallow etc.

I don't know anything about the system Dennis worked on with the driver, but this is what I do know:

The subharmonics don't even excite the break-up appreciably. Cross it anywhere between above 1khz to just above 2khz and you're probably home free. Plenty of beastly tweeters these days are happy in that range too crossed steep-ish - SB Acoustics Satori, Transducer Lab, Beyma TPL, Scanspeak Illuminator, Seas Crescendo, the larger high-end RAALs. That driver you're working on seems to have the excursion for it too, the KAXBLTWT.

Yes, there's limits to how low you can cross a given tweeter for a given goal - of course there are. But that doesn't mean you just skirt around it and cross it three octaves above its Fs.

This brings us to your own point:

Yes I am. This is not a place to pick the greatest drivers and say that anything made with them is going to sound good. Because that simply is not the case. Your comments about the B&W are a case to prove that true. They have a particular sound. And I agree that it can be corrected. What baffles me is why they leave the B&W factory the way they sound? A good driver does not guarantee that you will have a good sounding loudspeaker. It's the voicing and crossover design work that make up a loudspeaker.

+1

Drivers are just that - parts in a system and even companies that engineer good drivers do not necessarily engineer good systems. Part of this is almost because they design those excellent drivers to support their dated design philosophies / dogma. That those drivers work with modern philosophies is a coincidence! The engineers were chasing the wrong goal, but that doesn't mean the work they put in was done poorly. They just picked the wrong goal from the get-go.

It's really the case with a lot of high-end speakers. Some claim time-coherence with fancy stepped baffles. Others claim cabinets so dead a bomb could go off inside. Nothing's necessarily wrong with those goals - until you sacrifice or ignore the basics.

 

[GIEGAR]

Sirs, this discussion is excellent. *salute*

 

[mwmkravchenko]

@ GranteedEV​

​

I should explain my experience with a manifold of cones and their sonic attributes.

Every cone material suffers from flexure and the attendant frequency peaks and dips.

I got my first real taste of this while working with Phil Jones on a full range driver designed to compete with Lowther's offerings. We tried many a cone. And the worst was Kevlar. In the size format we worked with, a nominal 6.5 inch driver we ended up using a paper cones sanded to a few thousands of see through thickness.

That was in 1997. And yes there are different binder formulations available and such. B&W's are thermo pressed in a vacuum form. Quite an interesting process to watch. No video available of course. There are things that have to stay proprietary.

I'm talking about a wideband for the midrange because ideally that is what you want for a midrange. You want a driver that has good out of band response. It is much easier to work with and it does sound cleaner.

A crossover can tone down nasty resonances, but as you turn up the volume you also turn up the entire suite of response that your driver emits. If your driver has the cone area and swept volume to keep up and deliver addition SPL with additional power input you are fine. It is the small midrange drivers that can run into some problems. They usually have a smaller voice coil and that limits their power handling. A hot coil will drift in it's Re and the more power you pour in, the hotter it get's and you do not get more SPL. This is a very basic explanation of power compression. Not a desirable thing in a quality loudspeaker.

We listen to music where most of what is happening in the music is in the midrange band. Even the lower midrange octave of roughly 246 to 512 hertz there is considerable power developed.

You mention my tweeter. It is an attempt to better a good design. I think I have accomplished that. This build house is state of the art in it's build quality. I have worked with them for almost two years and the product has been rock solid. With exceptionally tight tolerances.

It is in production, as are a few other interesting offerings that are going first to OEM's and then in different variations to the DIY community.