MC-1000: Best Speaker in the World?

[amirm]

Gosh, reviews of these thrift store finds whilst amusing, seem a waste of your valuable talents and resources Amir.

The purpose was to find a "low anchor" for speaker measurements. Seems like even budget speakers are designed well enough that they did not serve that purpose. We now know that our "Olive" scores can indeed dip negative which was important to establish.

Until we see a flawed speaker, it is hard to know what is not.

 

[tuga]

The purpose was to find a "low anchor" for speaker measurements. Seems like even budget speakers are designed well enough that they did not serve that purpose. We now know that our "Olive" scores can indeed dip negative which was important to establish.

Until we see a flawed speaker, it is hard to know what is not.

What does it take for a speaker to be flawed?
Isn't this particular speaker flawed?

In my experience even the most comprehensive set of measurements cannot fully characterise the performance of a speaker.
This is why an observational listening assessment is an indispensable complement to a technical evaluation.

What does the "Olive" score score besides frequency response and dispersion characteristics?
As far as I know there are different "schools" regarding directivity, as there are different target curves for room response.
How accurate is the "Olive" bias?

Shouldn't we also be looking at the spectral decay of the tweeter and top of the mid/midwoofer's passband?
Nowadays most speakers are bi-wireable so it's easy to measure each driver separately:

 

[amirm]

Shouldn't we also be looking at the spectral decay of the tweeter and top of the mid/midwoofer's passband?

Not if you understand how misleading those graphs can be and the countless hidden variables behind them.

 

[amirm]

What does the "Olive" score score besides frequency response and dispersion characteristics?

I suggest reading the two papers, or Dr. Toole's book for a summary. Alternatively have had extensive discussions about it in the forum.

 

[amirm]

In my experience even the most comprehensive set of measurements cannot fully characterise the performance of a speaker.

I don't know what experience that is since you are posting under an alias. Over four decades of "experience" and research is behind the set of graphs I present and how to interpret them. Much of this is peer reviewed.

Once you step out of the best research, then yes, everyone has an opinion based on their experience. That randomness of opinion has led to thousand and thousands of different speakers and brands. Our goal here is stick to science that has proven very good correlation between objective anechoic measurements and controlled listening tests.

You can follow that or chase what the woofer is made out of, what bracing is in the speaker, what resonances show up in CSD charts, etc.

To be clear, the measurement data gets you close. Without it, you have no compass to know the direction let alone whether you are near your destination.

 

[Xulonn]

Gosh, reviews of these thrift store finds whilst amusing, seem a waste of your valuable talents and resources Amir.

Not really, because such reviews create perspective, a sense of history, and help to comprehend the way our minds relate to the hobby of audio. It appears that Amir does not put nearly as much effort into illuminating the past as he does when wrestling with the controversies of modern speakers and the universe of myths that surround them. And that's a good thing, IMHO.

I believe that one can comprehend and appreciate the latest technolgy better if one understands its roots in the past.

 

[Hiten]

Though I understand the purpose of vintage speaker measurements (Low anchor point for measurements, as mentioned earlier) and also knowing that no original data is available can anyone guess how the age (20-30 years) of speaker can deteriote and departs from original specs/measurements ? One newbie guess from me is as surround (depending on material) and cone hardens over age., the resonances and frequencies increase ? Caps loose value and esr the crossover point and attenuation changes ? It would be interesting to see measurements of classics(Which have been measured/reported long back) and how they measure now.
Thanks and regards.

 

[tuga]

I don't know what experience that is since you are posting under an alias. Over four decades of "experience" and research is behind the set of graphs I present and how to interpret them. Much of this is peer reviewed.

I'm new here and there's a lot to catch up with. Is there a topic dedicated to how you interpret measurements?

Once you step out of the best research, then yes, everyone has an opinion based on their experience. That randomness of opinion has led to thousand and thousands of different speakers and brands. Our goal here is stick to science that has proven very good correlation between objective anechoic measurements and controlled listening tests.

Observation is science. As far as I know, part of Toole's work resulted from analysing observation data.

Are you saying that the research developed by the BBC Research Department, by Linkwitz, by Geddes and many others is "randomness of opinion"?

You can follow that or chase what the woofer is made out of, what bracing is in the speaker, what resonances show up in CSD charts, etc.

To be clear, the measurement data gets you close. Without it, you have no compass to know the direction let alone whether you are near your destination.

You mentioned that "even budget speakers are designed well enough", or so say your measurements.

Perhaps measuring more aspects of performance like cabinet and driver resonances will point out the differences that sort the wheat from the chaff.

Does the "best research" dismiss their significance? If it does, then perhaps it's not the best...

 

[Deleted member 12179]

I think there are unfortunately many factors that can not be measured well, after all brain is a mystery, until today and for a long time coming will continue to be a mystery. Certainly in say 20 to 500 years they say they will be able to record and store dreams... not yet though.

I think if you are looking for a transparent, perfect, sharp, fully, totally and completely reproducing a source file as accurately as the source is, yes, all these measurements are awesome, amazing, really appreciated, and I can better purchase a dac, speakers and what not.

Now, when imperfections show up, those imperfections can and will make the experience worse, or better. Obviously. I repeat; worse, or better. Many of those "imperfections" are extremely pleasant. I´ve had 80's speakers that were clearly not transparent, not fair to the source, clearly reverbering, but boy, when you played dire straits, or pink floyd on them, loud or mid loud, they really felt amazing. For modern sharp electronic music sounded really awful, slow bad, dirty and old fashioned. Thing is, good luck measuring if those imperfections are pleasant or not, and how much and why and for who. Best of luck. And even though it can not be measured, boy you can be sure, it works like that. You can't measure everything, fortunately.

So these JBL are highly imperfect, and that is now proven with data, cool, awesome, thanks so much fo the effort and work required to do that. We already knew that, but thanks cause now it's measured. Now, question would be, so what does that imperfection does to my brain with this or that type of music???? Ahhhh, so you can not measure that, ahhhhhh... ok. I see. Because that is in this case the most important question, and stays not answered. The answered question was another, and it was a much less relevant question for these JBL, no?

 

[beefkabob]

Dudes! Trust in the science.

But seriously, I like my JBL 104 even though they measure poorly. Why? They're good for $64 powered speakers, though. I mean, they're a lot better than my laptops built-in speakers and the other cheap, nasty plug in speakers I've used before.

 

[amirm]

It all depends on conditions.

Observation is science. As far as I know, part of Toole's work resulted from analysing observation data.

That data is based on controlled testing. Not personal observations as you stated you had.

Are you saying that the research developed by the BBC Research Department, by Linkwitz, by Geddes and many others is "randomness of opinion"?

No. I said your observations posting as an alias on a forum is. It has no weight. If you want to post references to your sources of experience, then we can talk. Until then there is no there, there.

 

[tuga]

It all depends on conditions.

That data is based on controlled testing. Not personal observations as you stated you had.


No. I said your observations posting as an alias on a forum is. It has no weight. If you want to post references to your sources of experience, then we can talk. Until then there is no there, there.

Please ignore my listening observations.

But I would like to raise the question of an ideal or target anechoic performance taking into account that the most prominent designers don't agree on what the best dispersion pattern should be, particularly due to the fact that, if I'm not mistaken, Toole's research is focused on loudspeaker response preference.
No matter how controlled or scientific the method, the object of the study was personal taste.
In other words, Toole's ideal response resulty from a scientific study but the result is not based on science but personal preference.
Had this study been performed in Germany, where speakers generally show somewhat exaggerated treble, and the results might have been different. Or in the UK where wide dispersion doesn't work so well in the typical 11x12ft domestically furnished sitting room. And room construction also plays an important part on how much room gain one gets.
Also, speakers (i.e. Dali) which are flat off-axis and should be listened with no toe-in produce a lot more interaction with the side-walls and this is particularly unfriendly to smaller listening rooms.

I wish I could could substantiate my point in a more articulated manner.
What am trying to say is that care should be made when commenting on the results of the measurements as they could be easily be misleading; I think that it is wrong to believe that there is one size fits all solution for doemstic purposes.


You forgot to reply to my last question.

 

[RayDunzl]

The negative score throws off my ranking system chart:

I should think the lowest score would be "0", for a speaker that doesn't speak.

Though maybe silence could sound better than one that is really bad.

Good luck on your scoring project.

 

[beefkabob]

Please ignore my listening observations.

But I would like to raise the question of an ideal or target anechoic performance taking into account that the most prominent designers don't agree on what the best dispersion pattern should be, particularly due to the fact that, if I'm not mistaken, Toole's research is focused on loudspeaker response preference.
No matter how controlled or scientific the method, the object of the study was personal taste.
In other words, Toole's ideal response resulty from a scientific study but the result is not based on science but personal preference.
Had this study been performed in Germany, where speakers generally show somewhat exaggerated treble, and the results might have been different. Or in the UK where wide dispersion doesn't work so well in the typical 11x12ft domestically furnished sitting room. And room construction also plays an important part on how much room gain one gets.
Also, speakers (i.e. Dali) which are flat off-axis and should be listened with no toe-in produce a lot more interaction with the side-walls and this is particularly unfriendly to smaller listening rooms.

I wish I could could substantiate my point in a more articulated manner.
What am trying to say is that care should be made when commenting on the results of the measurements as they could be easily be misleading; I think that it is wrong to believe that there is one size fits all solution for doemstic purposes.


You forgot to reply to my last question.

Amir never said one size fits all. Crap is, however, crap. There is a high correlation between test results and what people prefer to listen to. To scientifically determine human preferences and how they connect to measurements, double blind and level-matching ensure no bias.

But you, at home, without level matching and without a double blind setup, cannot determine your real preference. Psychoacoustics play too large a role. This is why Amir's subjective preferences are also largely useless. Even if he doesn't know which one he's listening to, he's read the measurements and knows what the components are. His lack of total blindness absolutely affects his preferences.

 

[restorer-john]

RadioShack sold Genuine Shure cartridge stylus replacements(incl. V15s) under their own brand at much reduced prices. Don't get excited, the secret is long out of the box.

You should have seen how many RS/Shures I scored for peanuts when I took over a Tandy store and found boxes of discontinued and written off styli and cartridges in the late 1980s...

Basically, Tandy/RS marked down 50% (inventory cost loss) every 3 months, once the clearance markdown period was done. The goal was to encourage managers to get rid of dead/clearance stock, but many were lazy and just didn't count the old annoying stuff at stocktake and it "disappeared" into a box out the back. Not stolen, just written off. Items would fall off the live inventory when they went below 1c. Many styli were sitting around 7c-49c. The odd expensive ones were a bit more, maybe up to $7.49.

I was able to buy whatever I wanted at current inventory value less 10% staff discount!

 

[Wombat]

You should have seen how many RS/Shures I scored for peanuts when I took over a Tandy store and found boxes of discontinued and written off styli and cartridges in the late 1980s...

Basically, Tandy/RS marked down 50% (inventory cost loss) every 3 months, once the clearance markdown period was done. The goal was to encourage managers to get rid of dead/clearance stock, but many were lazy and just didn't count the old annoying stuff at stocktake and it "disappeared" into a box out the back. Not stolen, just written off. Items would fall off the live inventory when they went below 1c. Many styli were sitting around 7c-49c. The odd expensive ones were a bit more, maybe up to $7.49.

I was able to buy whatever I wanted at current inventory value less 10% staff discount!

I'll give you $AU7.49 for a V15 IV stylus.

 

[tuga]

There is a high correlation between test results and what people prefer to listen to. To scientifically determine human preferences and how they connect to measurements, double blind and level-matching ensure no bias.

Again, if the data was taken from a regional sample then perhaps the preference refers to a local preference only. Perhaps, we don't know.

And don't you think that people's actual preference (the speakers people buy) should also be taken into account?
Harman/Toole has published a study which seems to indicate that people prefer speakers with a flat horizontal response on-axis and a smooth off-axis response and wide horizontal dispersion.
Yet after more than ten years of participating in webforums from different countries (British, American, French, Spanish, Portuguese) I have observed that most people actually don't buy speakers with such characteristics.

Would a "Master Preference Ratings for Loudspeakers" table affect the buying trend?

Would the widespread promotion of the study (telling audiophiles that people prefer speakers with a flat horizontal response on-axis and a smooth off-axis response and wide horizontal dispersion) and making all speaker manufacturers publish CEA2034 change the buying trend?

Besides different room characteristics have a huge impact on sound and North American domestic listening rooms are very different from European rooms. This should be taken into account in a "Master Preference Ratings for Loudspeakers" table (I've mentioned Dali's horizontally flat off-axis response).


I find measurements extremely useful for shortlisting equipment worth listening to (predict tonal balance, room interaction, amplifier suitability) and to track down possible causes of problems that I have identified through listening.

I would rather have access to a comprehensive set of measurements (including CSD) than to rely on a rating system which resulted from someone else's interpretation of measurements.
"Master Preference Ratings for Loudspeakers" table is more effective than a simplistic What Hi-Fi? style of star rating but it relies on the assumption that people's actual preferences coincide with those which resulted from the Harman/Toole study. But I believe that if one's are able to correlate preference with measured performance then examining the measurements instead of relying on a prejudged interpretation is far more effective.
And why aren't driver resonances taken into account but only tonal balance and room interaction?

 

[tuga]

Crap is, however, crap.

Could you elaborate on this?

 

[thewas]

And why aren't driver resonances taken into account but only tonal balance and room interaction?

Driver resonances are indirectly taken into account as they appear usually as wiggles in the FR plots and get punished as deviations.

Agree though that especially in the modal region the interaction of the loudspeaker and listening room is very important and for example often a loudspeaker with a falling bass can have a more linear result at the listener position than one that goes linearly very low, but that cannot of course be taken into account in a metric that relies just on loudspeaker measurements.

Thus I think its obvious for the most here that the "Master Preference Ratings for Loudspeakers" being a limited single number has significant limitations and can only be a first quick guide for a newbie who would be probably be better off by choosing a loudspeaker with a "5" instead of a "2". Advanced users like many here will rather look at all measurements and do their own weighting on them, for example I don't care as much about as on axis linearity as I care about smooth directivity and low distortions as I anyway use an equaliser.

 

[Juhazi]

http://seanolive.blogspot.com/2010/07/are-there-cross-cultural-preferences-in.html
Conclusions

Very little research has been done in cross-cultural preferences in the sound quality of reproduced sound. What we know is that differe Preliminary investigations by the author in preferred spectral balance of music reproduced through loudspeakers have not revealed any significant differences in cross-cultural preferences to date. If cross-cultural preferences exist, the music and audio industries have largely ignored catering to them, instead distributing products that are optimized for a single universal audience.

Finally, an important question is whether audio companies should even be catering to these cross-cultural preferences if research eventually finds that they indeed exist? If the audio industry takes an “audio science in the service of art” philosophy where the goal is to faithfully and accurately reproduce the art as the artist intended, the question of cross-cultural preferences becomes moot. If certain cultures don’t like the sound of the art, then that becomes an issue between the artist and the recording producer/record executive - not the audio manufacturer.

http://seanolive.blogspot.com/2014/10/at-recent-137th-convention-of-audio.html

---
Makes me think that individual preferences and listening environments have more variance than national averages. You could have guessed!