Take a look at these measurements:
http://www.soundstagemagazine.com/measurements/reference3a_mmdecapo/
I've heard local and internet audiophiles raving about this speaker, but looking at the measurements... 😕 I mean seriously, WTF? To me this looks awful. Anyone heard these?
http://www.soundstagemagazine.com/measurements/reference3a_mmdecapo/
I've heard local and internet audiophiles raving about this speaker, but looking at the measurements... 😕 I mean seriously, WTF? To me this looks awful. Anyone heard these?
There's a cool article out there somewhere by Floyd Toole which postulates that a ruler-flat response curve isn't nearly as important as taming the deep resonances which affect the speaker's overall timbre.
While this may be true, I'm not sure it applies to this speaker in particular. My understanding is that they run the woofer with no crossover and a first order high pass on the tweeter. Perhaps they are taming these resonances mechanically though? Or perhaps I misunderstand how such resonances would be tamed. From this plot:
it looks to my untrained eye like there is a big resonance at 1k. (More measurements here: http://www.stereophile.com/loudspeakerreviews/1203reference3a/index4.html )
Certainly there are worse measuring speakers out there, but the FR looks a bit ragged here, and the stereophile measurements seem to point to other problems with ringing. I don't have a lot of experience interpreting measurements, so perhaps there's something I'm missing about them, or perhaps I'm making a mountain of a molehill? For comparison, many of these similar but much less expensive speakers seem to measure as well or better to me. I know these measurements aren't the final word (no impulse response, waterfalls, etc), but I was under the impression that a relatively flat frequency response, on and off axis, and low distortion, were two of the main attributes of good speakers.
http://www.soundstagemagazine.com/measurements/ascend_cbm170/
http://www.soundstagemagazine.com/measurements/paradigm_monitor5/
http://www.soundstagemagazine.com/measurements/energy_connoisseur_c3/
I guess I just don't understand what makes this speaker so great, especially for $2500 US. Do the benefits of this type of minimalist crossover really outweigh the drawbacks? For that matter, audiophile platitudes aside, from an engineering POV, what are the benefits of this type of design?
PS All the measurements I posted come from www.speakermeasurements.com, a very useful site.
I wasn't defending any speaker in particular, just the notion that the timbre of loudspeakers can't always be undersood from the bode plot. I can think of a few "highly regarded" speakers that don't have great looking FR graphs.
I'm not trying to be controversial - it's just the old problem of trying to match a subjective opinion to a measured result. Sometimes it works, sometimes it doesn't.
Is box office sales a measureable indicator of a good film?
I'm not trying to be controversial - it's just the old problem of trying to match a subjective opinion to a measured result. Sometimes it works, sometimes it doesn't.
Is box office sales a measureable indicator of a good film?
Understood and agreed. And I wasn't arguing with your point in general, just pointing out that in this case, it may not apply.
😕 This analogy doesnt work at all IMHO. You could equate box office sales to speaker sales, or quantifiable visual/audio quality (ie resolution, bitrates, etc) of a movie to quantifiable quality (measurements) of a speaker. But this is an apples/watermellons comparison.
Like always, it comes down to listening to the speakers.
I too am under the impression that a flat response is an attribute of a good loudspeaker. FR Response, phase coherence, Low distortions. Those are at the top of my list. Oh, and a large and wide soundstage.
I too am under the impression that a flat response is an attribute of a good loudspeaker. FR Response, phase coherence, Low distortions. Those are at the top of my list. Oh, and a large and wide soundstage.
I agree all the way around.
I used the film analogy just for fun to point out that popularity and good taste aren't always the same.😎
When I was younger, I worked for a notorious self-proclaimed golden-eared audio guru. While testing a new speaker, I told him we had made an adjustment in the XO. He listened for a few minutes and said something like 'Wow, much wider soundstage!'
Of course, I hadn't changed anything.
I'm sure he had good ears (as well as a keen sense of marketing), but I learned a valuable lesson that day about the power of suggestion and valuing the opinions of others at the expense of my own.
Trust, but verify.
Mike
I used the film analogy just for fun to point out that popularity and good taste aren't always the same.😎
When I was younger, I worked for a notorious self-proclaimed golden-eared audio guru. While testing a new speaker, I told him we had made an adjustment in the XO. He listened for a few minutes and said something like 'Wow, much wider soundstage!'
Of course, I hadn't changed anything.
I'm sure he had good ears (as well as a keen sense of marketing), but I learned a valuable lesson that day about the power of suggestion and valuing the opinions of others at the expense of my own.
Trust, but verify.
Mike
I've never heard the speakers first mentioned, but I don't feel any more that a flat FR is that big a deal. Years ago I had a set of Linn Kans, and whilst they had lots of issues, especially if you measured them, they were one of the most FUN speakers I'd ever listened to for the music I liked.
Many expensive speakers don't measure well. Even the Wilson WATT/Puppies are pretty mediocre, and worse than that off axis:
http://www.soundstagemagazine.com/measurements/wilson_wattpuppy7/
There are some "high end" speaker designers who don't appear to own any measurement equipment at all (e.g. nOrh). Other small boutique outfits just haven't spent the money for a decent measurement room. Past $2000, for many vendors it's no longer about performance anyway.
IMHO, you tend to get the best measured performance in speakers in the $500-1500 range. Also, companies that do their design in Canada or have a historical connection to Canada's National Research Council (PSB, Paradigm, Energy, Infinity, Axiom, etc.) generally seem to place a great emphasis on measurements, and this trickles down to their lower end models. There are some exceptions, e.g. Totem.
Measurements aren't everything though, as speaker measurements are generally woefully inadequate in capturing their sound. For example, rarely do we see harmonic distortion spectra for loudspeakers.
Still, it's been my experience that measurements do correlate to some degree to how a speaker sounds.
http://www.soundstagemagazine.com/measurements/wilson_wattpuppy7/
There are some "high end" speaker designers who don't appear to own any measurement equipment at all (e.g. nOrh). Other small boutique outfits just haven't spent the money for a decent measurement room. Past $2000, for many vendors it's no longer about performance anyway.
IMHO, you tend to get the best measured performance in speakers in the $500-1500 range. Also, companies that do their design in Canada or have a historical connection to Canada's National Research Council (PSB, Paradigm, Energy, Infinity, Axiom, etc.) generally seem to place a great emphasis on measurements, and this trickles down to their lower end models. There are some exceptions, e.g. Totem.
Measurements aren't everything though, as speaker measurements are generally woefully inadequate in capturing their sound. For example, rarely do we see harmonic distortion spectra for loudspeakers.
Still, it's been my experience that measurements do correlate to some degree to how a speaker sounds.
I find it interesting that the two examples provided have some striking similarities. They both have a big bump in the "drum impact range" around 80 Hz, and an elevation of the midrange roughly between 500-1000 Hz. It doesn't surprise me at all that a bass bump is designed in, so many people find that pleasing, but I wonder what the elevated midrange is about. All I can think of is that midrange-forward designs sound immediately different from most mass-market stuff, which tend to be a bit scooped in order to reduce mud.
Could designers of these high-end speakers really be designing in order to sound "different" and "special" and play to the strengths of their expensive drivers, rather than trying for the most balanced sound? That seems odd. And I believe Dave Wilson was a recording engineer before getting into the speaker biz, one would think that he in particular would try for flat speakers.
</tin foil hat>
Could designers of these high-end speakers really be designing in order to sound "different" and "special" and play to the strengths of their expensive drivers, rather than trying for the most balanced sound? That seems odd. And I believe Dave Wilson was a recording engineer before getting into the speaker biz, one would think that he in particular would try for flat speakers.
</tin foil hat>
Does it matter? Chasing some ever receding horizon about 'perfect reproduction' is a big WOFTAM as far as I'm concerned (now). Not that I don't find designing and building stuff enjoyable, but you'rre never going to get there, so relax and enjoy the music.HeatMiser said:
The Kans (and the Sara's too) I used as an example a few posts above are IMO and example of an excellent speaker for a music lover, because they were fun and enjoyable to listen to for the great majority of music, the great majority of the time. They also got the most positive comments from non-'phile listeners that heard them for the same reason. A speaker (or other component) that makes you want to listen to music is better than one that's technically superior, yet doesn't create that desire to spin some discs.
Do you have one of these too? Ever since my brother George went all goggle-eyed and wandered off into the back paddock to be sucked up into a beam of light, I worn mine all the time and none of them eetees have spoke to me agin
Of course you are right, I am just surprised how far these examples deviate from the theoretical "ideal" and find myself wondering if there is some other motivation besides making it sound good. In my more cynical moments I can imagine sounding "different" selling speakers almost as well as sounding "good". In some segments of the market.Brett said:
I haven't heard either of the speakers in question, so feel free to ignore my ramblings for that reason, or just on general principle. 😉
😀 Funny link.
Sound vs Measurement
Take a look at two graphs of my new line arrays:
one a smoothed, 1/3 octave, semi anechoic, and the other a listening position, high resolution (one bucket per 5 hz), in room.
look at the bottom of this page
I've been building and measuring and tweeking for several years now - and I believe that attributes of dynamic range combined with responsiveness make a huge difference compared to say, a 3 db rise in FR. As far as I can see the art of design is finding the balance among tradeoffs. I guess these people made their choice.
So: the two points I would make are: 1) that speaker could sound amazing (but the problems are fairly severe) - lower order crossovers always sound much better to me all things being equal, 2) Its an honest measurement - not trying to look better with smoothing or y axis expansion.
Here's the better question: With all the brains in this business why can't we invent a measure that will tell us how a speaker sounds? .... or am I missing something ....? It isn't all subjective. Perhaps this has been worked out and we just don't know the answer.
Take a look at two graphs of my new line arrays:
one a smoothed, 1/3 octave, semi anechoic, and the other a listening position, high resolution (one bucket per 5 hz), in room.
look at the bottom of this page
I've been building and measuring and tweeking for several years now - and I believe that attributes of dynamic range combined with responsiveness make a huge difference compared to say, a 3 db rise in FR. As far as I can see the art of design is finding the balance among tradeoffs. I guess these people made their choice.
So: the two points I would make are: 1) that speaker could sound amazing (but the problems are fairly severe) - lower order crossovers always sound much better to me all things being equal, 2) Its an honest measurement - not trying to look better with smoothing or y axis expansion.
Here's the better question: With all the brains in this business why can't we invent a measure that will tell us how a speaker sounds? .... or am I missing something ....? It isn't all subjective. Perhaps this has been worked out and we just don't know the answer.
Re: Sound vs Measurement
That's what a lot of the NRC research in the 70's was trying to do. I haven't read any of those papers, but it seems that ex-NRC guys repeat the same basic mantras: a flatter response sounds better to most people, an even off-axis rolloff is critically important especially in acoustically "live" rooms, simpler crossovers are better (not all the NRC guys agree on this last one; Toole insists two-way designs are inherently suboptimal). It would be interesting to hear what kind of research is going on these days.
TheoM said:
That's what a lot of the NRC research in the 70's was trying to do. I haven't read any of those papers, but it seems that ex-NRC guys repeat the same basic mantras: a flatter response sounds better to most people, an even off-axis rolloff is critically important especially in acoustically "live" rooms, simpler crossovers are better (not all the NRC guys agree on this last one; Toole insists two-way designs are inherently suboptimal). It would be interesting to hear what kind of research is going on these days.
Brett said:
Oh Man! You mean all I've got to do is wear one of those. I'd better get wrappin!
You'd think being up here in Canada that those NRC guys would passed that one along.
Sound quality enigma
Sound quality is truely very perspective. The Idea of perfect reproduction can only be achieved in a single way. Everyone uses the exact same speakers in the exact same listening environment. Say for example, your favorite band is coming out with a new cd. They produce it in the studio using studio monitors. They do everything they can possibly do to make it Exactly how they want it to sound. The audiophile wishes to reproduce that exact sound that the musicians heard in the studio on the final edit. The only way to do this truely would be to use the exact same components. This becomes basically impossible considering you cant have a different stereo and room for every CD you have, but what an audiophile desires is to posess a music system that can reproduce all of his or her music in the most enjoyable way, preserving the most possible detail from the recording. I havent heard the words transient response used yet in this thread. I wonder why? I always thought transient response had a huge role in the sound quality of a speaker. How well the speaker can respond and reproduce the signal is in part dictated by transient response. I am quite bass obsessed, and it has made me realize something about speakers. You not only need transient response, but you need transient response throughout excursion. If a speaker can play a 20hz tone and move 3 inches peak-peak while playing a 80hz tone flawlessly, then it really can do something. I see mids and tweeters the same way. Its not only about frequency response. Those plots are probably measured from a sine sweep going through the speaker, not some pink or white noise. FR changes insanely with listening position, and with the addition of stereo or surround sound channels. In my opinion FR plots DO NOT show the sound quality of the driver in ANY WAY. They simply demonstrate the usable frequency range of the speaker.
Sound quality is truely very perspective. The Idea of perfect reproduction can only be achieved in a single way. Everyone uses the exact same speakers in the exact same listening environment. Say for example, your favorite band is coming out with a new cd. They produce it in the studio using studio monitors. They do everything they can possibly do to make it Exactly how they want it to sound. The audiophile wishes to reproduce that exact sound that the musicians heard in the studio on the final edit. The only way to do this truely would be to use the exact same components. This becomes basically impossible considering you cant have a different stereo and room for every CD you have, but what an audiophile desires is to posess a music system that can reproduce all of his or her music in the most enjoyable way, preserving the most possible detail from the recording. I havent heard the words transient response used yet in this thread. I wonder why? I always thought transient response had a huge role in the sound quality of a speaker. How well the speaker can respond and reproduce the signal is in part dictated by transient response. I am quite bass obsessed, and it has made me realize something about speakers. You not only need transient response, but you need transient response throughout excursion. If a speaker can play a 20hz tone and move 3 inches peak-peak while playing a 80hz tone flawlessly, then it really can do something. I see mids and tweeters the same way. Its not only about frequency response. Those plots are probably measured from a sine sweep going through the speaker, not some pink or white noise. FR changes insanely with listening position, and with the addition of stereo or surround sound channels. In my opinion FR plots DO NOT show the sound quality of the driver in ANY WAY. They simply demonstrate the usable frequency range of the speaker.
There are many goals that should be met in speaker design. Some of them are not easily combined. A typical trade-off has to be made between response linearity and phase coherence.
I don't know the speaker in question, but it's step response (and thus phase coherence to a certain degree) is better than 99.9 % of all the speakers out there !!
Regards
Charles
Edit: The perfect speaker exists only on high-gloss brochures and in some salespeoples heads !!!!
Re: Sound vs Measurement
Earl Geddes thinks he has done this, to some extent. See:
http://www.gedlee.com/distortion_perception.htm
TheoM said:
Earl Geddes thinks he has done this, to some extent. See:
http://www.gedlee.com/distortion_perception.htm
Hi
I personally think that distorsion is really important in a speaker.
Some things cannot be measured like the soundstage capabilities because it is very room-dependant.
I believe that some ripples in freq response is not that important, but the general shape of the response is more important (rising freq response, relaxed mids, high rolloff, etc...).
F
I personally think that distorsion is really important in a speaker.
Some things cannot be measured like the soundstage capabilities because it is very room-dependant.
I believe that some ripples in freq response is not that important, but the general shape of the response is more important (rising freq response, relaxed mids, high rolloff, etc...).
F
- Status
- Not open for further replies.