Importance of Audio Measurment Revisited

I enjoy reading the reviews, can't make sense of the charts most of the time.

Measurements are important in understanding some of the perceptions and claims for what the ear hears. For example when I've read reviews of the Messa Barron tube amp or Atma-Sphere tube amps the connection to impedance swings and speaker loads becomes more understandable to my non-engineer mindset.

http://www.presentationmagazine.com/humour.htm

My answers would be; depends, depends, depends, and depends. Depends on what? Whether the piece sounds good to me. If it does and the manufacturer provides some explanation that makes sense within the laws of physics, I'm pretty happy to attribute the good sound to the supplied measurements.

I figure every manufacturer needs to distinguish his/her products somehow and every buyer needs to make his short list of pieces to audition somehow. I might put a piece on the audition list based on a measurement. The piece still has to pass the audition.

If I've got two pieces that sound either identical or one of those cases where they both sound good, just different, I might buy one and reassure myself with the notion that it measured better in some aspect.

It gets complicated sometimes. You get the case of the manufacturer who publishes measurements and the enthusiast who claims the piece sounds better because of the measurements. It doesn't quite fit with the laws of physics. You point that out and the enthusiast reverts to a subjectivist position that measurements don't matter. Okay, but uh....if they don't, why is the manufacturer staking his deal on those measurements and if that doesn't hold water, now what?

No matter what, your head is connected to your ears.

It can make you hear things that aren't there.

And it can make you fail to hear things that are there.

If we wish to be honest, ears and audio memory are not that reliable, brains are not that reliable, manufacturer's claims are not that reliable, word of mouth is not that reliable.

But, you still have to wade through it all and do the best you can.

And hopefully enjoy the results.

Great stuff, makes a lot of sense. I've often thought the phenomenon of perception changing with arousal might be due to evolutionary advantages.

May be a good survival skill on some level, but doesn't help with audio.

I've also suspected that this phenomenon may have something to do with the nature of consumer audio reviews.

If you read some of those audio review sites where average consumers review gear, it is very interesting.

Mid level, bang for the buck stuff gets five out of five stars.

When you get to more expensive gear, suddenly you read guys saying that this gear not only isn't to their liking, it's so bad it hurts their ears.

My guess is it does hurt their ears.

I suspect they are auditioning gear out of their price range, the price tag is making them nervous, their perception changes due to arousal and it becomes painful.

I would bet that there is something in their changed perception that is adaptive in a pre-historic environment.

The ears change to become more acute in some areas in order to hear approaching predators while becoming less sensitive to other sounds.

Terrible way to listen to music.

Ray Kimber once told me that they [Kimber] had been able to document changes in cables occurring during the "break-in" period. I was never, unfortunately, able to get Ray to disclose this "proprietary" data.

Click to expand...

I think most of us would have been very interested to see his data on that one

The cynical angel on my left shoulder wants to joke that the change Kimber noted was the customer's check clearing.

He's a baaaaad angel!

I think stress has to play some part in the break-in phenomenon.

You research like a madman, you audition, you lay out the dough, you hope to heck you made the right decision.

If you bought new, you know the minute you broke the seal and opened the box, the resale value of your purchase
just plummeted like a penny thrown off the top of the Empire State Building.

You might be experiencing a little stress.

A few weeks later, you've made peace with the fact that the money is gone, you've had enough time to relax, holy mother of god, that thing has finally broken in!

I have to wonder if the question regarding the value of audio measurements was a serious one or whether it was posed simply to stimulate discussion to a question that one hopes has an unequivocal answer. I am a physician. Medicine is described as a practice that is in part science based, and in part, an art. Sound reproduction, whether it is home audio or audio reinforcement for live performance is similar in that there will always be a technical side as well as an artistic side. Who could argue that certain fundamental technical specifications of the equipment are worthy of knowing? Just to use the most obvious of examples, two of the most basic properties of a speaker are its frequency response and its distortion. I think most users would have at least some interest in understanding these parameters, which can be most helpful in guiding listening or purchasing decisions.

Wouldn't you want to know that your speaker is of a design that has a certain suck out in the 100-200 Hz range due to floor interactions, as many do, so that you can understand if that design is going to work well in your room? Wouldn't you want to know if it has a "polite" dip at 2.5KHz as many from the "British School" do? Wouldn't you like to know if the frequency response is attentuated at the top end, or if it is enhanced, especially if you have no plans on using EQ and the speaker has no tweeter level adjustment? And wouldn't you want to know if your speaker has minimal distortion or 10% distortion at the volume level you normally prefer? These, it would seem to me, are all obvious. Yet the paradox is that this information is often easier to find for many inexpensive speakers, as opposed to those in the categorical "best" class. Want to know this information for a pair of PSB Alpha bookshelves? No problem. Want that information on Wilson Alexandria's, Magico M5's or Gryphon Poseidon's or any other "heavy weight" contender, and the data is mostly non-existent. I can appreciate the logic that its hard for independent firms to test these massive speakers, but I think the manufacturer should make this most basic and fundamental data available to the end user. Frequency response and distortion. Is that too much to ask? I guess so. This does not dismiss the importance and value of listening in any way. But the omission of such basic data is difficult to applaud. OK, enough for one post!

Marty makes excellent points. I think we can all agree that if measurements show a fault, no subjective evaluation can get around that fault. In that sense, the scientific measurement clearly trumps any other evaluation.

Conversely, in the areas where measurements are inadequate, then using them to prove something is just as bad. To use the same example in reverse, if the spec for a digital transport is simple frequency response, that alone won't tell you anything interesting about its performance.

Therefore, it is critical that one understands the nature of what is measured and whether that measurement is of high or low value. If it is high, it should be end of discussion.

The biggest horror here is if a designer avoided all measurements in creating his product. I can't think of any audio product that would not benefit from some measurements. I was once asking an acoustician what he thought of one of the speaker brands discussed in high-end circles. He paused and said, "they build high quality gear but I am not sure they know what they are doing." Surprised, I asked why. He said he had a high-end project and called the designer and CEO of this high-end speaker company to potentially use his product for it. On the call, he asks for the "RadPat" data. The designer says sure, we will get that to you. Then there is a pause and question: "what is RadPat?" He answers, "you know, the radiation pattern." To his shock and amazement, the designer had no idea what a radiation pattern measurement was! He literally had to explain to the designer of speakers that costs tens of thousands of dollars that radiation pattern measurements are necessary in figuring out if a speaker is suitable for the specific space it is going in. The fact that someone is designing speakers without even rudimentary knowledge of measurements for it, is kind of scary.

Now, per above, it is not a given that measurements given you everything. No company believes in measurements (and blind testing) more than Harman. Revel speakers are liked by many, but not adopted by all. So some of the tricks Marty talks about, might be necessary to move boxes.

marty said:

Frequency response and distortion. Is that too much to ask? I guess so. This does not dismiss the importance and value of listening in any way. But the omission of such basic data is difficult to applaud. OK, enough for one post!

Click to expand...

I'm in the John Dunlavy camp here. I would much rather see anechoic impulse response than frequency response since FR can be retrieved from an impulse (through FFT) but not the other way around. Additionally, step response, phase response, polar response would also be helpful.

I believe there are two key reasons why we will never see these (1) many speaker companies are small and do not have the resources to either build or rent a full scale anechoic chamber to do these measurements correctly and (2) they don't want you to know.

When I was in the business of installing SigTech room correction systems, I saw some VERY strange impulse response from some staggeringly expensive and well knows speakers. Tweeter ringing, drivers purposely wired out of phase with other drivers, horrible room response, etc. If I were a manufacturer who was having some success selling my speakers without providing good measurement data, why would I want to show them how really bad my speaker measured and mess up my business.


Just one man's opinion !!!

Hi

Lately Audiophiles have been quite defensive on the merits of Measurements. I believe this is due in good part to the usual / current set not relating very well to ALL that we hear. The mistake we, audiophiles, should not make, is that measurements are ALL utterly useless. They are NOT. A quick reading of a speaker FR can tell us quite a bit about its behavior in certain frequency regions as Marty so clearly posted. I have often posted that what we all are enjoying come from Scientific studies, be they tubes SS, Analog or Digital and the results are definitely measurable, maybe not measured but MEASURABLE.
There are several very eloquent posts in this threads, amongst them those from rsbeck. not verbatim but it is true that our ears are connected to our heads and we often hear what is NOT there. Once one has been subjected to a blind test , one often comes out very humbled, often even humiliated. Measurements keep things in perspective, a solid fundamental ground on which to anchor our perceptions that have been proven extraordinarily unreliable (notice I did not say imprecise).

I don't think I 'll use rsbeck gracious euphemism about a "bad angel". I would say to Ray Kimber that if he had been able to measure what break-in does to his cables to prove it. This is happening too often in High End Audio one makes a wild assertion and hides behind proprietary

We need to develop new sets of measurements. I like what someone like Earl Geddes has worked on. I do not subscribe to all that he declares but he has some valid points on our perception of Harmonic Distortion for example or what we need our speakers to do. We need to see similar studies, coming from the our sector.
I will not invoke the excuse of speakers manufacturers being too small to measure. If one were to look on the plethora of DIY forums, one would see tons of sometimes surprisingly good measurements and measuring techniques. let's not fool ourselves any Audio designer HAS to measure something. They make a commercial decision, often if we were to pay attention to some of their measurements we might ask questions, some are really and truly and sound that way too.. We force our mind to accept these for a while then we change, we call it the audiophile itch to change but it is simply disappointment .. I understand the choices designers make but some product are flawed and the flaws touted as choices.. ... Was I off topic ?

IN conclusion , measurements are important. We need study to develop new ones and to correlate these to our perception. In the absence of that we must use measurements , our sense and our logic. if a speaker system is down 10 dB per octave from 8 KHz to 20 KHz please don't tell me it has a glorious midrange .. because that all that's all it is playing ...

Marty, Great post..

Frantz

A great thread, lots of good information.

I'm one of the people who need to be educated on what all those charts and graphs mean. However, in the end, I'd rather hear it for myself and let my ears decide.

kach22i said:

A great thread, lots of good information.

I'm one of the people who need to be educated on what all those charts and graphs mean. However, in the end, I'd rather hear it for myself and let my ears decide.

Click to expand...

When I have more time, I can start down the path of explaining them one by one. Shall we create a new thread for it?

With your expertise I vote yes

amirm said:

When I have more time, I can start down the path of explaining them one by one. Shall we create a new thread for it?

Click to expand...

Yes, good idea.

I think that I once took the effort to learn the graphs and charts (long ago), but if you don't use that skill often the tendency is to forget it, like a second language.

OK will do. I will probably create a thread on each top separately so that discussions can occur after each one as to value, questions about the description, etc.