Can Objective Loudspeaker Measurements Predict Subjective Preferences?

[Edgar Betancourt]

OK, don't take my word for it. https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=15&ved=2ahUKEwjp2O3PornnAhXEHDQIHdqHDYY4ChAWMAR6BAgCEAE&url=http://digitool.library.mcgill.ca/webclient/DeliveryManager?pid=114011&custom_att_2=direct&usg=AOvVaw1RV37Jnwf9YO3TewFC2cKo

Turn to page 47 (pdf, page 30 of the actual paper) and start reading.

Didnt read anything about CR there....

 

[Edgar Betancourt]

Didnt read anything about CR there....

On another forum, Dr. Toole has talked about the Consumer Reports measurements many times.

To keep it short, his recollection of events is not consistent with what Edgar is posting.

Im not advocating CR's measurements. Simply pointing out that their methodology was exactly the biggest factors in the new method are exactly the same, near field frequency response on and off axis! Also the same thing that stereophile does too! If he poopooed the CR method anywhere he was full of crap!
Regardless as I pointed previously a flat frequency response regardless of axis is no guarantee of excellent sound! Ear tuning after measured parameters is the final crucial process! Even after all that personal preferences make pleasing even all critical ears an essentially impossible task. Even crappy 901's made excellent 802 PA speakers!

 

[Beave]

CR's original ranking methodology was flawed (I think John AA has it right - they were ranking based on flat sound power when they should have used flat on axis response).

Dr. Toole (and others?) let CR know that their results were flawed.

CR brought in Dr. Toole (and others) to help them fix their methodology.

That's how I remember Dr. Toole's posts, at least.

How does one tune by ear? With what music? In what room? At what listening distance? At what playback levels?

 

[Jon AA]

Didnt read anything about CR there....

Apparently you didn't read anything in there. I even gave you the page number.

 

[Speedskater]

the above linked Dr. Sean Olive (Harman International) paper was written in 2007.
it does not address how Consumer Union/Consumer Reports id testing loudspeakers now.

 

[Edgar Betancourt]

the above linked Dr. Sean Olive (Harman International) paper was written in 2007.
it does not address how Consumer Union/Consumer Reports id testing loudspeakers now.

I dont know how they do it now either their audio testing sort of fell off significantly in the digital age. My guess is that it was too expensive for the target audience. Now its mostly headphones and that type of thing. By the way I did read the pages you indexed I wasn't going to read the 150 plus and it was all about audio reflections and dispersal in a room.

 

[Beave]

the above linked Dr. Sean Olive (Harman International) paper was written in 2007.
it does not address how Consumer Union/Consumer Reports id testing loudspeakers now.

They stopped testing speakers - probably for a number of reasons (the Bose lawsuit, the fact they were doing it wrong for a number of years, etc).

 

[Edgar Betancourt]

They stopped testing speakers - probably for a number of reasons (the Bose lawsuit, the fact they were doing it wrong for a number of years, etc).

They won the lawsuit all the way to the supreme court.

 

[Alex2507]

Apparently you didn't read anything in there. I even gave you the page number.

Now I'm gonna have to go read it.

 

[Edgar Betancourt]

Now I'm gonna have to go read it.

Sorry I read page 47 instead of 30! However here is an important excerpt from it
First, it used finer frequency resolution measurements than the Consumer's Union model (1/20-octave versus 1/3-octave). Second, the new model considered several different target functions other than the "flat" sound-power target preferred by the Consumer's Union model (Consumer's Union, 1973). The new model considered the slope, the smoothness, and the average narrowband deviations within the sound-power curve as possible criteria for predicting the perceived sound quality of the loudspeaker.
As I pointed out both methods were anechoic miking on and off axis. However the harman method includes more data points and allows additional modeling criteria. In effect its a refinement on the original testing model.
Like weather forecasting 30 years of expirience and additional data points make a noticeable increase in model prediction. It didnt make the previous one wrong just makes the new one better. When introduced the CR model was state of the art.

 

[Beave]

They won the lawsuit all the way to the supreme court.

But at what cost? And would they want to go through something like that again?

"Won the lawsuit" isn't really a proper depiction of what happens when an entity gets sued, has to fight for their reputation for years, spends countless $$ on lawyers, and eventually gets out of it without paying the plaintiff.

It's like a war - even the winner takes losses.

 

[Alex2507]

It's like a war - even the winner takes losses.

But in this war, the loser actually won because people still think Bose is the best in the business.

 

[Edgar Betancourt]

But in this war, the loser actually won because people still think Bose is the best in the business.

They barely make speakers anymore. They make some fairly decent headphones though.

 

[Matthew J Poes]

There is AES papers and standards have this descibed in spades. Pro audio makes this mandatory to know and use every day. You have to be a member or know someone who is to get the infromation.

The Industry standard for compression testing in consumer audio is actually CEA-2034a. Various other standards exist and are in use in the pro audio industry with the primary difference being around the test stimulus used.

We will be publishing something on all of this in the coming months. I'd love to do compression testing on all speakers that come for review but there are potential risks of harm and we don't feel comfortable asking manufacturers to allow us to do that. It also brings new demands on our testing that may not be feasable. Things like sufficiently large and powerful amplifiers and an ability to play annoying tones at really loud levels for extended periods of time outside. My neighbors are likely already at their limits. I can't get far enough out into a field to totally address this problem. Still you never know, we might start doing this on some products.

 

[Edgar Betancourt]

They barely make speakers anymore. They make some fairly decent headphones though.

Dont take me wrong my first stereo was a pair of 301's and a yamaha 25w receiver. I even bought a pair of 901's! The worked fine for DJ purposes but sucked at home!

 

[Edgar Betancourt]

The Industry standard for compression testing in consumer audio is actually CEA-2034a. Various other standards exist and are in use in the pro audio industry with the primary difference being around the test stimulus used.

We will be publishing something on all of this in the coming months. I'd love to do compression testing on all speakers that come for review but there are potential risks of harm and we don't feel comfortable asking manufacturers to allow us to do that. It also brings new demands on our testing that may not be feasable. Things like sufficiently large and powerful amplifiers and an ability to play annoying tones at really loud levels for extended periods of time outside. My neighbors are likely already at their limits. I can't get far enough out into a field to totally address this problem. Still you never know, we might start doing this on some products.

Buy a pair 1.25 K Mcintoshes. It will excercise any speakers and barely get warm

 

[Edgar Betancourt]

Buy a pair 1.25 K Mcintoshes. It will excercise any speakers and barely get warm

A bargain at 25k for a pair.....

 

[Jon AA]

As I pointed out both methods were anechoic miking on and off axis.

They both measured speakers with a microphone! They're exactly the SAME!

If you can actually read the above paper and not see how it directly contradicts your original claim, I can't help you. I'm out of spoons.

 

[Edgar Betancourt]

They both measured speakers with a microphone! They're exactly the SAME!

If you can actually read the above paper and not see how it directly contradicts your original claim, I can't help you. I'm out of spoons.

Please, again you are comparing methodologies separated by 30 years! Its like trying to compare an IBM punch card mainframe with a modern computer processor. In 1973 you would use an oscilloscope, write down the result and pull out a slide rule. In medical terms its like comparing an m mode sonogram with a modern doppler (one beam vs hundreds of beams. The M mode was not bad or wrong it was less sophisticated.

 

[Jon AA]

Please, again you are comparing methodologies separated by 30 years!

No, you were. I really don't have time to waste quoting you to yourself. If you are unable to understand the conceptual differences between the methods and why they will have drastically different results or are too proud to acknowledge it, I can't help you any further.