|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
I'd like to share yet another baffling audio experience. I figure many of you can probably relate to it.About a year ago I was in a high-end audio store (very nice place, well staffed, with some excellent demo rooms). I sat down to listen to a loudspeaker that had a truly impressive design criteria, fabulous measurements (among them: very flat frequency response, on and off axis, extremely low driver distortion) designed by one of the most knowledgeable teams one could hope for from a company.
On went a recording of Faure's "Requiem" (natch). The sound was everything promised from the hype: expansive, coherent, stunningly smooth throughout the midrange, clean from top to bottom and most of all an incredibly low "noise floor" (for lack of better word), giving the presentation an ease (lack of grit and grain) and clarity that was truly amazing. Must be those new ultra-low distortion drivers. It seemed as effortless to listen into the recording as it does in life at a concert.
But...if it sounded so real in so many ways, why was I not quite connecting to the music? I love the Requiem, but it wasn't grabbing me...something not quite right. Yet the presentation was otherwise so complete I couldn't put my finger on it. Every time I encounter a speaker this impressive and I'm not moved by the music I first start to blame myself - it's not the sound system, it's me: this sounds so good but I'm not in the right mood, I just can't relax today to enjoy music. Those kind of rationilizations.
More music: Jobim's "Girl From Ipanema," and other tracks featuring guitars. Then it started to dawn on me. I'm very sensitive to timbre and tonality - a high-end speaker has to sound, on some level, "right" with instruments in order for me to completely submerge in the experience. I play acoustic guitar, including fooling around on various models, classical and otherwise. I closed my eyes and performed my "rightness" test - Does that guitar on this recording sound like I could walk over, pick it up and play it myself?
Then it clicked. No. The guitar coming from these speakers did not actually sound like any real guitar I've encountered. Something about the tonality and quality was "off." The guitar sounded colder, whiter, slightly electronic and drained of the the organic quality found in real instruments - this was a guitar from a twilight zone episode, where you wake up to find that everyone looks that same, but there's something not quite right, something different....Or was it just me?
Sitting near these mega-speakers was a pair of Dunlavy SM-1 monitors.
Hmmm. "Do you mind if we play these tracks through the Dunlavy speakers?" - "Sure, no problem."Viola. There it was: organic sound. The vocalists sounded human. The guitar sounded like it was made out of all the right materials - wood, gut strings (or metal strings, depending on the guitar). I closed my eyes and listening to the Jobim track and it took no stretching of my imagination to feel like this was guitar that could be passed around the room for others to play. My mind said "aaahhh, this is more like it" and I felt compelled to drink in the music. Switching between the Dunlavy and the mega-speakers confirmed this effect, like switching between a tv whose picture was too blue, to one with properly balanced colors.
This kind of experience leaves me baffled. The mega-speaker, like the Dunlavy speaker, had fabulous measurements - very flat response, especially off-axis. The room was excellent - well dampened, lots of room for the speakers to "breathe." But one sounded "wrong" to my ears and the other sounded "right." Without hearing each speaker directly compared to real instruments I could never truly say which was right. However, I regularly compare real instruments to their reproduced counterparts on speakers that pass through my home, and almost always the ones that sounded "right" before the test also sound most accurate during the tests.
This is not to imply that *my* "right" is the same as, or better than, someone else's "right" sound. But these experiences remind me that there are subtleties buried in the measured performance of well engineered loudspeakers that still have an important impact on the sound. Frankly, I just don't know what the heck they are, and it's extremely hard to find out because asking anyone who builds loudspeakers inevitably results in a partisan view of which design criteria changes the sound from "wrong" to "right."
Any thoughts or comments? On the technical side: on a well designed product, what do you believe can go subtly wrong in a speaker to make it sound "off?" And on the consumer side: Do my fellow inmates have any reliable personal criteria for finding the "right" sound in a speaker?
Rich H.
**PS, I didn't mention the name of the mega-speaker because it's not the point of this post to bash that company, or denigrate a product anyone on this board owns - it was a fabulous product. And this isn't
a "plug" for Dunlavy. There just happened to be Dunlavy speakers sitting in the room as well. While I do think that Dunlavy speakers are some of the more tonally accurate speakers out there, there are quite a few speakers that I would also have chosen as sounding "right" in comparison with the mega-speaker.***PPS. One difference in the design between the Dunlavy speaker and the mega-speaker was phase/time coherence. The Dunlavy speakers are a "pulse coherent" design, while the mega speaker was not, being a high-order crossover design. Could phase/time coherence have contributed to the percieved "rightness" of the sound? The jury is still out on that one for me, and I sure wouldn't jump to conclusions based just on this demo. But, hey, who knows....?
Follow Ups:
This is a second submission on this thread. I've spent some time thinking about this and do not have a definitive answer, but I do have some pertinent second thoughts.I am of the opinion that the majority of differences perceived in hi-fi equipment is due to differences in frequency response, and/or dispersion characteristics. So here's a thought experiment:
Suppose the two speakers had a frequency response of +/- 1 dB from 30 Hz to 16 kHz. Further suppose that by some miracle that both had that response in the room where you listened. Unrealistic? Yes, but stay with me. Now suppose that speaker A had a response that was -1 dB at 30 Hz, rose to +1 dB at 36 Hz, and them sloped smoothly and continuously to -1 dB at 16 kHz. Next suppose that speaker B started at -1 dB at 30 Hz, rose to +1 dB at 36 Hz, fell smoothly to -1 dB at 100 Hz, rose smoothly and continuously to +1 dB at 13 kHz and then dropped to -1dB at 16 kHz. Would you hear a difference? Damn straight, you would. A would sound warm and rich, B would sound bright and a bit thin. And that's with the ridiculously tight band of +/- 1 dB. If you doubt me on this, get hold of a graphic equalizer and duplicate those curves and see what that small amount of change does to the coloration of your system. According to a paper by Fred Toole (Infinity website, under technical papers) even very small deviations over wide bandwidths are very audible. I think my system is well set up and like the sound a lot. Adding those slopes to mine made very significant changes in coloration.
If that's all it takes, no wonder the Dunlavy's sounded different than the megaspeaker, despite the impressive specs. It also means that the skill of the designer in "voicing" the speaker is more important than the specs.
It sure is a matter of taste. I recently listened to Watt/Puppy 6s hooked up to a top-of-the-line Mark Levinson system and a fully blown LP-12. I wasn't impressed. Then I got blown away by a pair of diminiutive Spendors - go figure.
I listened to Revel Salon + all-Levinson (separates) system recently. This system is regarded by many to have perfect synergy, not surprising since all components come from the same manufacturer.I could not find any flaw to the sound I was hearing. However, I also could not find anything special in its sound, nothing that pulled me into the music.
A much cheaper system (but NOT exactly cheap in absolute terms), Sonus Faber Guarneri Homage + Musical Fidelity Nu-Vista M3 + Musical Fidelity A3CD, gave me a much, much more involving musical experience. Right, the system's sound was a bit euphonic, and not quite as detailed as the Revel+Levinson, but the sound was simply fabulous.
And this is all because of personal taste. Somehow, my analytical mind is telling me that the Revel+Levinson system was the "more correct" system. But the artistic part of my brain tells me that the SF+MF system sounded "better".
They did everything right except allowing me to connect to the emotion of the music.
Me too, frankly.Rich H.
Please note I write this at work with some Dunlavy SM-1's about two feet from me. They are driven by a cheap AIWA system, but still sound good. If I crank it up early in the morning when no one else is at work and lean back a bit, the sound is quite enjoyable and there is some serious soundstaging going on.I agree with you in that the frequency response of a speaker doesn't tell the whole story. I have both a set of unbelievable Dunlavy Millenniums and a set of Paradigm Studio 100's. Even though the Studios have a respectable frequency response, they just don't hold a candle to the Millenniums. Part of it, I believe, is the decay of the speakers. Dunlavy's have very good waterfall plots. That is, if a tone is fed to a speaker and then suddenly cut off, how long does the speaker take to become silent? This effect gives the speaker a certain cleaness or quickness. Transients are cleaned up, overhang is gone, the speaker seems quicker and cleaner, you have to hear it for yourself.
You also mentioned the single pole crossover. This sure makes sense, that you want accurate responses to quick transient events. But I never heard before that single pole crossovers cause higher IM distortion. One would think single pole crossovers are necessary for good imaging but they aren't. However all single pole corssover speakers are notable for their imaging, Meadowlark, Vanderstein, Thiel and Dunlavy. So it must really help.
Brian,This is just the thing: the "mega speaker" has absolutely fantastic waterfall plots - Stereophile reviewed them and J. Atkinson remarked that they were perhaps the cleanest waterfall plots he'd ever seen.
And, in fact, if anything the sound of the mega speaker sounded a little cleaner and smoother than the Dunlavy (although I can't be quite sure). But the fact remains that through the Dunlavy speakers instrumental timbre seemed correct. "Familiar" might be the way to describe it for someone used to playing those instruments - everything seemed to be made of the right materials. Timbres through the mega speaker, while smooth and coherent as life itself, lacked a convincing organic signature. Weird.citroeniste,
The set-up for the mega speaker was fantastic - couldn't really ask for more, which is why it sounded so incredibly impressive. Still, every speaker has it's own timbre, or sonic signature. The mega speaker's voice didn't match my concept of "right."
Rich H.
The only thing left is the single pole crossovers. You said you heard other speakers that also got instruments right. Where these single pole crossovers too? If not we seem to be back where we started...If you wish, I would appreciate it if you would email me the name of that mega speaker. I am really curious. I won't divulge its idenity.
Brian,I have heard other loudspeakers that I felt portrayed instrumental timbres in a fundamentally correct way (nothing I've heard is perfect). And several of these used high-order crossovers. The Hales
Transcendence series (before they went under) amazed me. Paul Hales' first design goal was always correct timbre, and it was his view that high order crossovers achieved this more easily than low-order designs. I don't know that to be true, but I was amazed by the Transcendence 5/8s when I heard them. Excepting a *very* slight metallic aftertaste or glaze to the presentation, it was amazing how they managed to weave a sonic rainbow of correct sounding instruments. I confirmed their timbral authenticity at my own home, tested against my instruments and recordings.That said, I can't help but notice that virtually every phase/time aligned speaker design I've heard seems to make it to my list of "right" sounding speakers (and that, of course, is a very small number of manufacturers in the audio world).
When I listened to the Dunlavy Aletha a while back, I first listened to a track by saxophonist Joshua Redman. Normally as soon as I hear an instrument through a loudspeaker I quickly recognize that speaker's "take" on the sound - e.g. how it colors the sound of a saxophone. Through the Alethas the sax played and, well, I was sort of baffled. All I could think of was: "sh*t that sounds like a real sax!" I was taken aback at how hard it was pin pointing coloration to the sound. The Alethas did much the same for many other instruments.
It's interesting that REG of TAS also lauded the Aletha for outstanding timbral accuracy, and he is self-admittedly "obsessive" about instrumental tone/timbre. Not saying the Dunlavy speakers must therefore be perfect, of course. But they have been impressive in my brief encounters with them. Which makes the Cantata a very tempting product, I must say. (Sounded excellent at the last CES).But, back to the question. The fact that high-order speakers have also sounded correct to me does perhaps put us back at square one. I'm intellectually attracted to the idea of pulse-coherence, but I refuse to be seduced by theory only. I'm holding out for some form of proof (personal or otherwise) that it is meaningful. I wonder: if we could make well controlled, direct comparisons would the phase/time coherent designs show any advantage? Hmmm....
Rich H.
(I'll try and email you about the "mega speaker").
I share your enthusium with Dunlavy's. I agree they are accurate. That is why many don't like them. They call them clinical, unforgiving, or harsh. Actually they are simply letting you know what your're feeding them, as they should. Sure I have listened to some poor recordings on my speaks and noticed the problems, but then I thought, hey, despite these problems with the recording, doesn't this song sound better then ever? Can't you hear more of the music? The answere is yes and I would rather hear all of a favorite song, warts and all.I went to the factory and listend to the Cantatas. A remarkable speaker no doubt. My speakers are even better though, the Dunlavy Millenniums. Please give these a listen. The bass is to die for, plus all that extreme timbre accuracy.
Rich,Re crossovers and rightness:
I'm running a 21 year old set of KEF 104aB speakers and things sound right, especially voices these days. That "rightness" also seems associated with a "matter of factness" that doesn't leap out and grab you, so it's easy to pass that kind of sound by but it really holds up for years and years.
I'm going to buy new speakers when I finish work some time in the next 10 months or so, and the ones that are really high on my list prior to starting serious auditioning are Vandersteen 2CE sigs. They strike me as having a similar sense of rightness to the KEFs, and also sound matter of fact.
The KEFs aren't phase coherent and I think they have 4th order crossovers so I'm in the same boat as you on the crossover question,
David Aiken
I think there may be a simpler explanation. The Dunlavys were simply in better positions relative to where you were sitting. Perhaps the bigger speakers were set up such that their bass response sounded commensurately superior to the little Dunlavys, but overal timbral balance suffered for it. I've heard that many a time in hifi stores, even exceptional ones. Unless they're in their own exclusive demo room, the most expensive speakers are rarely the ones with the best set-up.A great speaker set-up just right will sound awesome. A great speaker set up mediocrely will sound ok. A great speaker set up badly, why waste your money unless you're going for their look or bragging rights?
NP: CNN
I own a pair of heavily modified Dahlquist DQ-10s. If you are not familiar with them, ask a local audio 'historian', dealer, or search the forum. They are speakers that legends and curses were (are) made of!!!FYI- They are laughably coloured in frequency response, relatively inefficient (about 84dB 1W/1M give or take a dB), and present a major pain in the tukus to the partnered amplifier (rumoured to dip to 2.7 ohms at 100hz!).
But when set up properly (away from room boundaries, toed-in, listeing chair well into the room as well) with sufficient power (manual rec: min 60W, real world double that into a 4 Ohm load!), they sound more musical than most every speaker within the price range of these pups (anywhere from $200 to $800 used!) and even go up against some speakers up to triple the price.
I should know!! I own a pair, and while I wouldn't use them to mix in my home studio (too finicky, and won't fit on top of the mixing board!!hehehe), I ALWAYS use them to judge the final mix or master before sending it out. Even with the 'flaws', if it doesn't sound musical there, back to the board!!!
"Listen for the emotion, not the commotion!!!"
- my grandfather (BTW, NOT and audiophile!!!)
Dman
Aucostic music in real life sound warm, focusing on the whole picture with details not so very forward. Some speakers do sound like this some don´t.
With correct timbre you lose some details or rather you have to listen for them to hear them. In a shop we tend to think this is wrong and we will buy the more detailed speaker thinking this is more correct. As fewer and fewer of us do visit a consert or hear real live aucoustic music fewer will know this is wrong.Some speaker manufactors know this and accept the market as it is. They will build very detailed, lean sounding speakers. Sell truckloads and live happily ever after.
Some do think that music is the reference and will not budge an inch from the warm sound it needs to sound real. They tend to build speaker that stay the same for a longer period of time, are "unsexy" and -in some cases - use poor marketing. But the sound is true. As a customer you don´t upgrade very often. A Harbeth user said: They are for people who like to listen to music for a long time, 30-40 years or so.
I do suspect that you are more in the latter category and the Mega speaker is in former?
cheers
Michael Y
The mega speaker to are closer to latter group
Rich,This is a great thread that I find very interesting. There could be a number of factors the Mega speaker did not suit you. The room may not have been condusive to this speaker. The components in the chain may have been ill suited to these speakers or lacked synergy. The speakers could be too resolving and uncovered flaws in the recording. But the thing that really sticks with me about audio as a whole is that everyone seems to hear things a little differently. What I mean is that everyone's tastes are different and when they hear a recording the nuances of the recording strike different chords in different people. Obviously, the Mega buck speaker did not strike you as pleasing but another person may come along and say that that was the best system that they have ever heard. As they say different strokes for different folks. All that matters is the music and if it pleases you. But a word of caution. You must audition the gear in your home, because ultimately that is where it counts!
hope this helps,
Johnny
I have a hunch that 6db x-overs are part of the answer. I have been searching for years for speakers that really satisfy me; I am about to go out and listen to Spendor SP 1/2s and Totem Winds, if I can fine a pair. One revealing episode came when I listened to B & W N 804s playing the final trio from Rosenkavalier (EMI Haitink version); sounded good, but not very moving; then listened to a pair of ML SL3s; top rounded off and all the usual complaints, but it made me sit down and really listen to the music. Now they have a 12 db x-over, but none above 250 H. I want to listen again to the ET LFT8a; that may be my answer. But I am convinced that high order x-overs in the voice range are a killer--though I will listen to those Spendors.
Chris,
I'm unsure what you're saying about crossovers and the correlation with your listening pleasure. You listen first and then later check out the crossover--6dB or 12dB--to see if that's responsible for your listening pleasure? Or you check out the specs first and then go to audition the speakers that specs lead you to believe will satisfy you??I ask this since I own ET's but have no idea what the crossover is like (I'm not particularly technical). I'm wondering what it is that makes you think LFT VIII's will satisfy?
Thanks,
Bruce
...almost an hour to write a lengthy reply to Rich… but then I had a power interruption and I lost my document. I do not do it again.Yes, the first order crossover is it something witch is indispensable for this type of music but there is lot of more to it . Probably it will be desiccated in future at:
http://www.audiooutcasts.com/wwwthreads/showthreaded.php?Cat=&Board=General&Number=90&page=0&view=collapsed&sb=5&o=
Regards,
Romy The caT
Romy, sorry to hear you lost your reply to me. I would really have enjoyed reading it (boy I hate when that happens, I've lost several long posts just before hitting "send." Kinda takes the wind out of you for a while).I really appreciate everyone's response to my post.
I'd like to point out that I'm quite aware of all the conundrums associated with high-end audio - accuracy vs musicality, the "absolute sound" vs the "absolute emotion" and on an on.
My post was not necessarily related to "how to buy a satisfying sound system." Clearly, tastes being different, some would have found satisfaction in the "mega speaker" over the Dunlavy.
Myself, I can generally enjoy music on almost any system, no matter how modest. However, hi-fi *is* short for "high-fidelity," and my post was actually more concerned with that aspect of the listening experience. I share with many audiophiles, past and present, a fascination with the sound of real instruments, and the reproduction of those instruments. Can we, using wooden boxes, vibrating cones of rubber, metal, paper, (or sheets of mylar), a bunch of magnets and wire, accurately reproduce the sound of real instruments? Sometimes it's amazing how close we get, and other times it's amazing how heroic efforts manage to fall entirely short.In my post, I was ruminating on the ability of my emotions, and acoustic memory in determining sonic accuracy of those loudspeakers.
One question I left the audition with was: was I right about the timbral accuracy of the Dunlavy vs the mega speaker? Is it possible that in blind testing against real instruments, I might have surprised myself and chosen the mega speaker as more accurate? Could be. But my experience in comparing speakers with real instruments so far implies that the feeling of "rightness" that I experience from a loudspeaker is a good indicator as to how accurately it will reproduce the sound of my real instruments (from my own recordings of those instruments). So if my impressions were correct, and the Dunlavy speakers were capturing something fundamentally right that the mega speakers missed, what was responsible for this? The frequency measurements for each speaker are among the most "accurate" you can find anywhere, with both going for a flat on-axis response. I find this to be a fascinating puzzle.
BTW, It's almost become fashionable to fear the quest for
sonic accuracy/realism. The sentiment is that, since absolute sonic accuracy is currently out of reach for us, attempting to reach that goal will simply drive you crazy; concentrating on the unattainable realism of the sound will distract you from the music. I happen to be quite sympathetic to this line of thinking, which is why I when I listen to music I tend to leave behind questions of sonic realism. However, I also LOVE the sound of real acoustic instruments, and I'm terrifically appreciative of those people who are striving to design gear that may some day reproduce those sounds in my home. They may not be there yet but bully for them, because ideally someday we'll have the choice of reaping the benefits of their work (*choice* as in: we can always stick to our less accurate equipment if we enjoy it more).So for me, I enjoy the music, but I can also derive pleasure from the engineering aspects of hi-fi, and the quest for accuracy. One line of thinking does not poison the other. Sometimes they feel like entirely separate goals, and other times they move closer together, actually becoming inseparable on some occasions. That's much of the fun of enjoying a hobby that intersects art/science/engineering.
My 2 cents, anyway..
Rich H.
*** I'm quite aware of all the conundrums associated with high-end audio - accuracy vs musicality, the "absolute sound" vs the "absolute emotion" and on an on.There are no conundrums about it. The accuracy, musicality, "absolute sound” and "absolute emotion" are the organic part of the very same pictures. You just have to know how to melt them together. For example: education and wisdom, age and maturity are not necessary come together but there are people for whom they are the conflict-less complimentary qualities…
*** speakers were capturing something fundamentally right that the mega speakers missed, what was responsible for this?
When I asked the same question to an individual, who believe that he know the answer, he replied that, yes, there are the things witch are responsible… but those things are not that we know.
*** It's almost become fashionable to fear the quest for sonic accuracy/realism.
The Sonic realism is the most bogus concept the Audio Cheerleaders ever deployed. When a listener stop to recognize the way how his “himself” reacts to music but begin to look for an abstractive “realism” then Music listener die and a zombie audiophile is born.
Without REALLY knowing what the recording sounded like when it was made, this isnt really a fair comparasion.
The more accurate speaker could be revealing the poor choice of microphones, placement, etc.
I feel there is no substitute for living with a speaker in your own home, and listening to many different types of music over a long period of time.
Remember, even a broken clock is correct twice a day!
As far as the phase goes, a good speaker can be designed a number of ways.
A first order crossover gives lots more IM distortion, and the dispersion pattern, especially off axis suffers.
The B&W 801 uses high order crossovers, and it is pretty much THE reference standard used by classical music recording studios.
It is important to note that a speaker with elevated response in the presence region can sound impressive.
Dont buy or do anything until you live with it for a week or so!
I think it is the crossover topology. I have PSB Stratus Gold/i's (4th order crossover) and like them very much, they sound very very good. But, I recently got a pair of Meadowlark Vireo's (1st order 2-way stand mounted) for my office and they also sound very very good but they have a sense of "sounds right" that is less there in the PSB's.
There may be a number of reasons the MEGA speakers didn't sound as good as the Dunlavys. Component mismatches that may have mucked things up, but I in my experience, many audiophiles and maybe even designers are "missing the forest for the trees".I would say the "audiophile's" list of the components of good sound reproduction would include descriptions like: soundstage, depth, timbre, PRaT, impact, decay, imaging, etc.
But, HOW DO WE LISTEN??? I think many audiophiles audition a speaker and consciously compartmentalize their audition by seperately evaluating each component of the sound. And if each component of the sound works, then the speaker is acceptable. I've made this mistake myself. I failed because I chose to purchase based "the sum of components" rather than the entire "musical experience".
Synergy is another key to building a good system, but that's already been discussed (but probably not practiced)
FWIW
Right On! Just about the whole audiophile cult is running off looking under the wrong rock.It takes a lot of time and patience to find the right components and interface them properly to get the system performance you want. I insist on musicality and naturalness above all. Over the years I have learned that more money or "better" specs is not the answer. (I just replaced a "highly reviewed", "product of the year" CD player with one that cost about 40% as much because in my system, it worked better)
Now, having said that, I'll get back to the original question: why did the Dunlavy sound more "live" than the "megaspeaker". If they were both driven off the same system, it is unlikely that it would be optimum for both. Rather than focus on crossover orders, or other specifications, I would first look at the overall system, including the listening room to see if the total system were optimum for both speakers. Second, speakers with similar spacifications often sound very different. Why? Because standardized measurements are not sufficient to characterize the sound we perceive in real rooms. Partly because something as simple as the height of the drivers above the floor can make a difference in how they sound in a real room. But the measuring microphone is usually aimed at the center of the cabinet or the center between the drivers or some other standard positioning--but my ears stay about 39" from the floor when seated.
But there are zillions of factors which could make the difference and without a whole heck of a lot more details and data, none of us are going to solve the problem in this forum. Personally, I have had enough trouble getting my system to sound "right".
THere was a second question about how do we tell if a system is right? I don't play guitar, so I can't use that, but I do use vocals. Male and female. And I close my eyes and see if I can visualize the vocalist or not, and if I can visualize their lips forming the syllables. If the system is "right" it happens easily and just about automatically, if the system if "off" just a little, it doesn't happen at all. Like the other respondents, I'm not sure what makes one system "right" and another "off", but smoothness and neutrality and a sense of "ease" are always associated with it.
I happen to be an engineer, and I have made many measurements of systems, and to date have not found the magic measurement. But ruler flat frequency response doesn't seem to be the answer.
This post is made possible by the generous support of people like you and our sponsors: