Tip of the Day: Don't Be too Focused on Loudspeaker Specs & Measurements
Measurements and specifications are important tools in the decision making process for purchasing audio equipment. They can help those with the knowledge to interpret them identify potential performance issues. Further, objective data is useful for making comparisons between items. Nonetheless, it is important not to get too wrapped up in graphs and specs. In the end, your ears are the final arbiter of sound quality, not your eyes. Not all measurements and specifications are created equally.
It is perhaps important to reiterate one part of the above: specifications and measurements are useful to those with the knowledge to interpret them.
For example, on the face of it, a sensitivity specification seems
fairly straightforward: a speaker with a specified sensitivity of 93dB
at 1 meter with a 2.83 volt input should have an output advantage over a
speaker with a specified sensitivity of 91dB at 1 meter with a 2.83 volt
input. However, how sensitivity was rated, (ie in room or anechoic, impedance of the speaker, bandwidth tested), will have considerable impact
upon what those numbers mean. As Dr. Floyd Toole once said:
"You often get more useful information on the sidewall of a tire than you do from a loudspeaker manufacturer's website."
We also warn the consumer that it is unwise to focus on one particular aspect of performance at the expense of all others. For example, if your focus is on-axis frequency response, you may miss issues with respect to distortion, compression, resonance, etc. In the end, what matters most is whether or not you are pleased with the sound emanating from your speakers in your listening space. No speaker is perfect, so it's up to you to find the best balance of compromises for your ears—and while measurements can be helpful, they're no substitute for actually auditioning equipment.
Source: Forum member Steve81
Confused about what AV Gear to buy or how to set it up? Join our Exclusive Audioholics E-Book Membership Program!
Recent Forum Posts:
Could you offer some advice, please? I am in the market for some speakers, generally I don't pay a lot of attention to specs but to sound quality. I am considering: Klipsch RF7II @ $2600/pr., incl. tax, a steal, I think, KEF R500 @ $1950/pr., and JBL LS 60 @ $2200/pr. The Klipsch according to specs go the lowest and are the most efficient. I do like speakers with a bit of bite. I tend to listen to classical, opera, jazz and vocals. My equipment consists of a Shanling A3000 tube amp and Shanling CD3000 player. Which of the speakers would you place your money? Thanks.
Pottscb, post: 936018
Paul Barton (of PSB) has logged as many hours in an anechoic chamber as any person alive, yet he designs his top of the line speakers to measure incorrectly in the anechoic chamber …
What? I thought the Synchrony One is one of the flattest & most accurately (anechoic) measured speakers out there?
Pottscb, post: 936018
…but maybe 20% better is worth that extra $19500 to some.
Why stop at 20%?
If they think the $20K amps sound 100% better (night & day better), it's totally worth it.
We buy what we can afford.
Yes, if I were a trillionaire I would buy $400K of amps. But I would still choose to believe that my $400K amps won't sound any better than Crown & Emotiva.
That's just my belief.
And some people who can only afford $400 amps can choose to believe that all amps sound significantly different.
We respect all beliefs, religions, and politics here as long as they do no harm.
This is an interesting article. As you may or may not know, you (and Audioholics) have quite the following of internet groupies who use your review articles to proclaim “if it ain't in the measurements, its not audible” and they then extrapolate this into a $500 100 watt (measured) ht receiver will sound identical to a $20,000 100 watt (measured) class A monoblock just because they put out identical power at identical impedance, or some such nonsense. Anyway, I hope this new article gives them pause…I especially like the sentiment “if it sounds better to you, its probably worth it.” I don't think owners of $20K monoblocks really think they sound 4000% better than $500 receivers, but maybe 20% better is worth that extra $19500 to some. I agree that measurements are misleading and this is why many publications have ceased to use many of them. Paul Barton (of PSB) has logged as many hours in an anechoic chamber as any person alive, yet he designs his top of the line speakers to measure incorrectly in the anechoic chamber to take into account the “floor bounce” of sound that will inevitably occur in your living room when you get it home. A prime example of measurements not bearing out the actual quality of a speaker…
gene, post: 935442
Try measuring the crossovers. Many of them are higher than 1st order despite the claims. 1st order crossovers are rarely a good idea when designing a speaker to produce high output levels with minimum distortion and compression.
I must say, I have also never seen any worthy proof that a speaker with a first order crossover exists. And let's make something clear here, we are not talking about electrical 6dB, we are talking about a speaker which hits true 6dB acoustical slopes. Would sure love to hear one!
It has been blind tested many times, shallower slopes do indeed sound better. But… 6dB is impossible. And 12dB is extremely hard to pull off.
First, 99.99999% of all drivers are manufactured in a way that they literally design themselves into LR 24dB slopes with 2 parts in the crossover. Using a slightly asymmetrical (shallower on the woofer) LR4, one can easily design a nearly perfect speaker, which is time aligned and minimum phase. The system should have lower distortion because of the steep slopes and better power handling. Drawback = group delay.
So, since shallower slopes tend to sound better, how to design them? Out of hundreds of drivers, there might be a handful which can be used. Shallower slopes = higher distortion and lower power handling. Let's say the system will never be pushed hard… With shallower slopes, it is impossible to time align a speaker using asymmetrical slopes. A.) The baffle has to be severely slanted, which causes other problems as now one will be listening to ALL drivers off-axis. B.) Do an ugly, expensive, and inconvenient modular design, where the tweeter can be physically offset from the midrange/midwoofer. C.) Use a lattice network to delay the tweeter. A lattice network (check out Dynaudio) is extremely complex and adds the same amount of group delay as an LR4 filter, negating most of the benefits LR2 had to begin with.
Forget about designing a 12dB crossover. But there is a place where one can hear what it sounds like. Seek out the very expensive Dynaudio speakers. The most expensive Focus and up mostly use 12dB acoustical slopes with lattice networks. The Esotar tweeters and midwoofers are absolutely state of the art and are specifically designed to work with shallower slopes. The speakers mostly sound terrific, but Dynaudio is only one of few companies in the world who can hit 12dB acoustical slopes.
Every company brought up and discussed here uses LR4. They include: Revel, TAD, Infinity, JBL, Klipsch, NHT, Paradigm, B&W, Magico, YG, Goldmund, Polk Audio, PSB, Axiom, Aperion, RBH, Ascend, Salk, Philharmonic, Focal, etc.
Dynaudio is the only company that I know who makes LR2.
Thiel and Vandesteen is one giant question mark. They for sure don't hit 6dB acoustical slopes. The slopes are shallower than a tradition LR4, but it's a mess if you ask me. Thiel has a slanted baffle (off-axis problems galore). Vandersteen physically offsets the tweeter, but the speakers measure bizarre…