It’s All a Preference

[Phelonious Ponk]

Please read my previous exchange of posts with Amir in this thread. No time to repeat it again as the Aida are arriving

I read them the first time. Just re- read them. I think Amir answered them, and with more Harman research, except for the one about the MLs not being a good representation of dipoles. I agree. Was there something somewhere that said the MLs were included to "represent dipoles?" What were the Revels an B&Ws representing?
Tim

 

[rbbert]

In any case, I bet those Aida's will sound good by anyone's criteria!

 

[Phelonious Ponk]

One of the mods needs to put a sticky on this one.

 

[amirm]

One of the mods needs to put a sticky on this one.

Since I am involved in it, it would be self-serving . I will leave it to the rest of the management to decide to do it.

 

[steve williams]

Done

 

[terryj]

My take on things is that people who love the ML Summit love the Summit for it's midrange and are not bothered by the bass bump.... Personally, I can't stand that bump that all ML hybrids have, but some people like it, ignore it, or are willing to make the tradeoff. They have been listening to that sound for 20+ years and really like it. Just like sitting in the same comfortable chair for 20 years, you will know that someone switched the chair on you and that the midrange of the $300 infinity speaker will not better that of a Martin Logan even if you blind them and stick pins under their finger

I don't get it.

THIS is the entire point of the thread/discussion...THIS is exactly what harmon have found 'does not tend to be true'.

That blinded to identity, people do NOT tend to like poor reproduction. It was also in the link I gave from Dave Moulton, which no doubt was not read.

Can people simply not look and see how these findings might apply? Instead, as always, we have automatic dismissal cause (natch) 'we audiophiles know better'.

Again I will ask...is there zero curiosity from any of you to-even on a hypothetical level-participate in a test like this??

Myles is not interested, but we can understand that he has a vested interest as a reviewer to maintain whatever lofty status he thinks that bestows upon him (not game to test his MLs tho, a sign of integrity and bravery).

Why is it IMPOSSIBLE that *you* too would mark down the MLs when you did not know what you are listening to????

Can you not admit that if you are certain you would not mark them down, then there has to be an element of brand prejudice (pure defintion of the word) in play? A priori you have made a decision based on your knowledge of the speaker.

Which ironically, IS the entire premise being denied.

 

[mep]

I for one would gladly participate in one of Harman’s tests given the opportunity. There is nothing to fear. I would also like to go through Harman’s training to see what that entails. I just wonder if they “teach the test” so to speak. Apparently Harman “trains” you for what they want you to listen for and then you are supposed to identify the traits they trained you to hear. I don’t know how far this training goes with regards to the music they picked out. They have like four different songs they play over and over again and I don’t know if they train you what to listen for in each song or if this just some generalized training on how well designed speakers should sound.

 

[Phelonious Ponk]

Since I am involved in it, it would be self-serving . I will leave it to the rest of the management to decide to do it.

It's your stuff that needs to stick, brother.

Tim

 

[Phelonious Ponk]

I just wonder if they “teach the test” so to speak. Apparently Harman “trains” you for what they want you to listen for and then you are supposed to identify the traits they trained you to hear.

Nope. That's not what they are trying to accomplish and, unsurprisingly, Harman has tested their training thoroughly, and probably made adjustments along the way. What they've found - and I believe Amir has already posted this - is that their trained listeners don't come to different conclusions than untrained listeners. The just get there faster.

It's about efficiency; it's a business.

Oh, and by the way, you can take the training: http://harmanhowtolisten.blogspot.com/

Tim

 

[amirm]

I for one would gladly participate in one of Harman’s tests given the opportunity. There is nothing to fear. I would also like to go through Harman’s training to see what that entails. I just wonder if they “teach the test” so to speak. Apparently Harman “trains” you for what they want you to listen for and then you are supposed to identify the traits they trained you to hear. I don’t know how far this training goes with regards to the music they picked out. They have like four different songs they play over and over again and I don’t know if they train you what to listen for in each song or if this just some generalized training on how well designed speakers should sound.

I think there is some confusion. Trained listeners are trained on detecting small differences. They are not taught what to like and not like about music. They are being used as distortion analyzers. As for taking the test, you sit there and listen to music and score it. You control what you hear (blindly). You can go back and re-listen, compare as many times as you want and then vote. I had no training whatsoever when I took it the couple of times. I just listened and wondered which of the speakers sounded more "normal" to me. Exactly what I would do if I were listening to music sighted.

 

[amirm]

Oh, and by the way, you can take the training: http://harmanhowtolisten.blogspot.com/

Tim

I have taken this test twice with Sean Olive driving and also giving answers. He would EQ one of the bands (blindly) and ask us which one is changed. We would give the answer. At the beginning I would give the right answer and he would confirm *before* revealing the answer. I.e. he knew it as well as I did. As the difficulty went up, I was probably the only one left still voting right relative to the group of people we were with (AV dealers). As it got harder, I would cheat by giving two numbers: "2 or 3" . In every case, Sean would then fine tune that to the only right answer. Clearly I was one step behind him in ability to do this. His hit rate was nearly perfect, mine maybe 70% and then we would stop before the test would get ridiculously hard.

 

[Gregadd]

That blinded to identity, people do NOT tend to like poor reproduction. It was also in the link I gave from Dave Moulton, which no doubt was not read.

This is the question that intriques me. Especially as at applys to to Amir. That someone who claims to have loved the ML previoulsy and had always heard it sound good "sighted"but when listening blind found it to be so bad he questioned ML "highend" credetials. Given somoen who is a dealer for Revel and JBL one would think he would be intimately famialiar with those speakers. For me that is an epiphany of epic proporions.

One obvious suggestion is that he has trained himself to prefer a different set of standards than ML customers. if we are talking about preferences to certain "sound" that preference would not change just because the demo is "blind." It may well be tyhat ML customers, a sizable group are more bothereed by errors in the time domain than an errors in the frequency domain.

 

[Phelonious Ponk]

A question for those in this thread vigorously defending MLs from this non-attack --

What we have here is solid data, gathered through very carefully designed and implemented methodologies by people intending to bet their livelihoods on the results. They're not fooling around. They're not trying to prove anything to audiophles. They develop the products they take to market on the results of this research so, really, their intent here is not to make you look bad.

The measurements say that the ML hybrids in question have very uneven off-axis response and not a terribly inaccurate response on-axis. The listening tests say that people didn't rank them highly compared to some other speakers.

What, exactly, are you disputing? Do you think the measurements are inaccurate? Do you think the opinions of the listeners are not what they really felt about the speakers? Or do you simply disagree, because that's cool. If you like the way ML hybrids sound, regardless of how they measure, regardless of how they test in careful blind listening test, you really don't need to invent a bunch of reasons why the tests are wrong. You don't need page after page of "but, but, but...." You simply need to like your Martin Logans.

These tests, these measurements, these people do not have to be declared wrong, for you to enjoy listening to ML hybrids. But I've heard most of the last 2 generations of ML hybrids, well set-up and in a really good room. I can understand why they might be somebody's favorites. But on the other hand, if what you're trying to say is that they do not have a midbass hump and that they have good off-axis response, hell, I've got to wonder if you've ever actually heard them. Martin Logan themselves would tell you they don't have good off-axis response. They use the rapid fall-off of their panels off axis as a marketing point.

It may well be tyhat ML customers, a sizable group are more bothereed by errors in the time domain than an errors in the frequency domain.

Maybe. Let's go with that. Where is the research that shows they're better in the time domain?

Tim

 

[Gregadd]

Of course they have gone beyond the question of ML quality. There is nothing new here. It's the same old argument -Do better neasurements equal better sound? If so what measurements? The fact thaIMO sounds better but measures in some areas bothers all. The big suggestion is once again listener bias. Eradicated when testing blind. Nothing new there.

 

[Phelonious Ponk]

Of course they have gone beyond the question of ML quality. There is nothing new here. It's the same old argument -Do better neasurements equal better sound? If so what measurements? The fact thaIMO sounds better but measures in some areas bothers all. The big suggestion is once again listener bias. Eradicated when testing blind. Nothing new there.

Are you saying they have good off- axis FR? That they have no rise in the mid bass? That the measurements are wrong? Or are you just insisting that the people in the listening test should have liked the off-axis response and the rise in mid-bass and agreed with you?

 

[RUR]

Hell, maybe we shouldn't be listening to CD or dipoles at all. Maybe we should be listening to multi-poles, otherwise described as "improved Bose 901's".

http://home.provide.net/~djcarlst/SLReport10.05.pdf

and background: http://home.provide.net/~djcarlst/20101118.htm

 

[amirm]

This is the question that intriques me. Especially as at applys to to Amir. That someone who claims to have loved the ML previoulsy and had always heard it sound good "sighted"but when listening blind found it to be so bad he questioned ML "highend" credetials. Given somoen who is a dealer for Revel and JBL one would think he would be intimately famialiar with those speakers. For me that is an epiphany of epic proporions.

You seem to think there is a big conflict here but there really isn't. I had never, until this test, heard the ML play the same music as another speaker. For all you know, and I know, I would have had the same conclusion sighted had been able to compare this way.

Is it odd that I would find demos of ML speakers to sound and the speaker performing so under par in this test? Yes. I grant you that. But this is beside the point. The two situations are not comparable. What is comparable is them being in that Harman test. Under that controlled condition, the ML did not perform well. And then it happens that its measurements said the same thing.

On me knowing the sound of Harman speakers, I had never heard the JBL speaker that was in the test that I partook. And in no time did I recognize its sound other than it sounding the most normal and natural.

Prior to taking the test again, we were introduced to the new line of Revel performa series. I thought these had killer sound. I forget why but I think someone implied that it would be in this double blind test. I am going to be honest with you. I had butterflies in my stomach that the speaker that I was scoring down was that unit! A sigh of relief when the curtains were opened and the speaker was not in the test, and the one I picked again as my favorite was the JBL.

One obvious suggestion is that he has trained himself to prefer a different set of standards than ML customers.

I don't know why you speak as if I am not here! Please ask me these questions before suggesting them. As I have noted, the ML sounded so different and poor that you could not help but score it down. And it was not just me. More than a dozen people in two occasions did the same. Is your theory that we are all master speaker testers and have memorized this sound? I bet if I put you in the test seat you could not tell your speaker from others blind! I certainly could not. At all times, I was trying guess but I could not. At the end, I just voted as I heard them.

if we are talking about preferences to certain "sound" that preference would not change just because the demo is "blind." It may well be tyhat ML customers, a sizable group are more bothereed by errors in the time domain than an errors in the frequency domain.

Are you trying to be an engineer now? The only domain that I would think matters is "which one sounds better." Or which one I would want to take home. Many people have sat through these tests and that ML speaker was not one that majority of them wanted to take home. Whatever conspiracy theories there are,they are a very weak ones.

If you are looking for plausible theories, here is one: maybe the clips that Harman has picked are very revealing of speaker weaknesses. That when we demo MLs sighted, we hear them with material that they have found to sound good with them. As for owners, they probably gravitate to what sounds good on them too. I know this was clearly the case with our planar speakers. I had a list of stuff that sounded good on them and that is what I demoed.

 

[amirm]

There is nothing new here. It's the same old argument -Do better neasurements equal better sound?

They do but not any old measurements. Certainly not what you see in the marketing material for speakers.

If so what measurements?

They take 70 anechoic measurements in both horizontal and vertical axis. Then they apply weighting to them based on their importance. This data with some 90% confidence predicts how a speaker does in a listening test.

The fact thaIMO sounds better but measures in some areas bothers all. The big suggestion is once again listener bias. Eradicated when testing blind. Nothing new there.

You have to explain this as I don't understand your English .

 

[terryj]

This is the question that intriques me. Especially as at applys to to Amir. That someone who claims to have loved the ML previoulsy and had always heard it sound good "sighted"but when listening blind found it to be so bad he questioned ML "highend" credetials. Given somoen who is a dealer for Revel and JBL one would think he would be intimately famialiar with .

Yeah, it is interesting ain't it.

But when you think for a little bit, it can make sense you know. Firstly, forget for now sighted and previous bias/conclusions etc, we only ever hear things in isolation. What I mean is, when something is taken alone with NO counter reference, it can be just accepted and we have nothing to compare it to (well duh)

BUT, suddenly have a counter reference we can immediately compare it to, then things change dramatically. All of a sudden, that is when peaks etc jump out at you. (think the colour chart comparison, take one alone...but have two side by side and a different ball game)

Same deal with the Dave Moulton, dunno if you or anyone bothered to read it (he is a very well respected figure in audio btw) he was absolutely and completely convinced he and his partners were on to a winner with their speaker. How many times and in how many different situations did he hear them??

Same deal tho, when he was suddenly able to juxtapose them with other sound references (blind of course) and be able to directly compare them, he too had the audio epiphany you spoke of with amir.

Look, it is kinda like habituation. It is a relatively recent phenomenon but flat bass etc is a good exa,ple of that. The vast majority oif people have and live with lumpy poor bass (tis only in the last decade say that some have begun to measure in their room at home, prior to that it woulda cost thousands just for the program let alone the technician to go with it!).

Yet the majority of them would be completely unaware of that, and listen to that sound as if there was nothing up. (we still have arguments on forums about bass, subs and measurements, there are many audiophiles out there still completely stuck in the dark ages).

Now, suddenly flip that switch and have the bass sorted and smooth, WOW! And it is not just the bass, the perception of the entire frequency range changes completely.

If you have ever experienced that demo, then you'd have an idea of what I am talking about. The point is however, that for *years* we have been completely happy with the sound we had, and it is only by that process of side by side comparison that gives us the reality.

 

[amirm]

Same deal with the Dave Moulton, dunno if you or anyone bothered to read it.

I not only read your quote but went to his site and read the rest. It really resonated with me as you say.

So thanks for posting it. I had not read that story before.