Can Objective Loudspeaker Measurements Predict Subjective Preferences?

[Ossidian]

Your response is both arrogant and ignorant.

The Hirata method is based upon noting that musical waveforms often contain transient events which are brief, and asymmetric in shape. Recognising this, Hirata devised test waveforms which are asymmetric and quasi-transient, but which have a mean (average level) of zero – i.e. no d.c. component. Waveforms of this general type are very interesting since they can probe effects whereby the linearity of an amplifier (or other audio component) is sensitive to medium-term statistical or short-term asymmetries in the waveforms. Unfortunately, as with the Belcher method, Hirata’s proposal in the above article is based upon waveforms that have some related drawbacks. For example, the Hirata waveforms were essentially a group of rectangular pulses. The resulting spectra then depend upon the sharpness of the pulse edges, etc.

The purpose of what follows is to outline some of the amplifier effects which might need to be probed to identify non-linearities which affect musical reproduction. Then go on to consider an alternative method that might provide measured results more relevant to musical performance.

Measuring well and sound well are different things. Hirata tested many amps; some tested measured poorly like the Marantz 8B but sound fantastic while some amps that measured very well sounded bad on his testing method. There are many products that measure well and sound awful.

I won't get into the end user restrictions i.e rooms ect. Nobody lives and listens in an anechoic chamber.

 

[Bucknekked]

Your response is both arrogant and ignorant.

Measuring well and sound well are different things.
.

Well, there you have it. Succinctly expressed. I am both arrogant and ignorant.
As well, the conclusion that devices that measure well are different from devices that sound well.
Again, succinctly expressed and clearly done. All supported by a 38 year old research paper about a testing method.

I bow to your insight on my character and knowledge base. Perhaps your testing method, reading a few lines on a forum, should be codified and put in to a paper. There are a great number of people who would love to know how to do such a thing.

I think I've read enough to satisfy my curiosity though. Thank you for your insights.

 

[Pogre]

Your response is both arrogant and ignorant.

I read nothing in Buck's post that was ignorant or arrogant. What I can't believe is that you went back later just to edit this part into your post. Your earlier, more civil reply wasn't good enough? I don't know if "arrogant" is the right word for your response or the type of person who would bother to go back to add it in, but it does bring to mind a few other colorful adjectives...

 

[Alex2507]

I won't get into the end user restrictions

Thank you.

 

[Verdinut]

If anyone ever does and it turns out to be accurate, we'll never hear about it because, well, you know, women have a kind of telepathic network and they'll all know about it as soon as it's active. By "we'll never hear about it", I really mean "We'll never hear from the inventor".

In the mean time, there's this-

https://etherealmind.com/wp-content/uploads/2011/02/man-woman-knon.jpeg

Here are the facts: Male Brain and then Female Brain

Statistics have changed with regard to SEX in the last two generations: Women now play more with themselves than ever

 

[highfigh]

Here are the facts: Male Brain and then Female Brain

View attachment 30885

Statistics have changed with regard to SEX in the last two generations: Women now play more with themselves than ever

The brain stem needs to have "Mad at him because "I think you know!"."

 

[killdozzer]

Your response is both arrogant and ignorant.

The Hirata method is based upon noting that musical waveforms often contain transient events which are brief, and asymmetric in shape. Recognising this, Hirata devised test waveforms which are asymmetric and quasi-transient, but which have a mean (average level) of zero – i.e. no d.c. component. Waveforms of this general type are very interesting since they can probe effects whereby the linearity of an amplifier (or other audio component) is sensitive to medium-term statistical or short-term asymmetries in the waveforms. Unfortunately, as with the Belcher method, Hirata’s proposal in the above article is based upon waveforms that have some related drawbacks. For example, the Hirata waveforms were essentially a group of rectangular pulses. The resulting spectra then depend upon the sharpness of the pulse edges, etc.

The purpose of what follows is to outline some of the amplifier effects which might need to be probed to identify non-linearities which affect musical reproduction. Then go on to consider an alternative method that might provide measured results more relevant to musical performance.

Measuring well and sound well are different things. Hirata tested many amps; some tested measured poorly like the Marantz 8B but sound fantastic while some amps that measured very well sounded bad on his testing method. There are many products that measure well and sound awful.

I won't get into the end user restrictions i.e rooms ect. Nobody lives and listens in an anechoic chamber.

Man, you couldn't be further from the truth. @Bucknekked gave you such a calm, elaborate answer and took time to read your links, you can't say it is ignorant. Also, he didn't give you an opinion, he repeated end results of some serious people doing thorough experiments. You can't just say it's ignorant.

 

[Art Vandelay]

Ossidian

The basic premise of this thread is to pose the question "can measurement predict subjective performance?". That question has been thoroughly discussed, sliced n diced, and I think pretty well answered in the affirmative.
Measurements, done well, can predict from a group which speakers will sound better than others. Toole and others have demonstrated it to be true.
.

When it comes to amplifiers the situation isn't necessarily as clear cut though, and Bruce Candy, Pass and others have admitted that the correlation between subjective and objective measurements is far from definitive. Leaving aside for now the argument that a Halcro should sound identical to any other lower distortion amplifier regardless of price, when operated within voltage and current limits.


With loudspeakers there are established measurements that can provide a reasonably good guide to in-room sound quality but it's probably fair to say that 1. People read more into basic measurements than they really should, and 2. Measurements or specs from the manufacturer or other third party are still limited in various ways to not highlight deficiencies that can be audible in real world conditions. An example of that might be high IM or harmonic distortion with certain types of music that simple low power test tone sweeps are not going to reveal.

Ultimately the test is in the listening within your own specific listening environment, which is likely going to be far more compromised than the majority of well designed loudspeakers.

 

[Speedskater]

When it comes to amplifiers the situation isn't necessarily as clear cut though, and Bruce Candy, Pass and others have admitted that the correlation between subjective and objective measurements is far from definitive.

When it comes to amplifiers with euphonic colorations this is true. It would be hard to predict human preferences from measurements. And with different listeners, different loudspeakers or different music genres the preferences might be different.
But with accurate amplifiers there will be correlation (or the wrong measurements were made).

 

[Art Vandelay]

But with accurate amplifiers there will be correlation (or the wrong measurements were made).

Perhaps you mean to say "near-perfect amplifier", but it's more likely that the right measurements were not made, given that all measurements are valid. Typical sweeps and distortion measurements may not reveal some dynamic distortions which can be audible, which explains why two seemingly perfect amplifiers (from typical measurements) can sound different in some ways.

 

[colofan]

Why don't standard measurement suites test for dynamic compression at high volume levels? I haven't looked at a review in a while, and I'm just throwing this out there, but that seems to be a big limitation of many speakers, particularly smaller speakers. I'd like to see quantification of how much a speaker struggles to produce high dB when called for.

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.

 

[colofan]

With current technology I am going to explain a method I am using to get accurate objective data for a speaker system in a room.

LEAP by linearX did a pretty good job of mapping the bass response in a room. The simulation took into a number of room configurations though some assumptions were made.

Dr. Toole book and papers a superb in describing how human beings a test instruments is challenging. Hearing test on each of the listeners is need to under stand at least partially what each set of "ears" are hearing.

Let me return though to the transducers data collection.

Let me take the simple case of a cone driver. Most effective way of measuring said driver is measure the displacement of the cone while being driven with a laser and comparing the input waveform with the output waveform as a displacement display. This will give you an accurate way objectively to determine full modeling data of what is being actually produced independent of room. A full 3d model of the face of the driver will give the energy fielded by this driver.

Dome drivers and flat planer type drivers also could be measured directly.

With the energy field from the driver we could introduce then an fluid dynamic model into an target room to determine energy levels through the room. This can turn into a pretty complicated model however it is doable and doesn't require a super computer anymore.

The point is we now modeling methods to determine the what wave fields are going to be at the listeners location. Still needs to be correlated with what people hear but the tools we have are objective and reproducible.

 

[Edgar Betancourt]

Another timeless topic. The Harman method is not new at all, Consumer Reports has carried out basically the same method for decades. There is one small problem, no one listens to a speaker in an anechoic chamber! The best loudspeakers are tailored over and over to "average" environments in an effort to normalize real rather than mathematical environments. A lot of that tuning is done by careful crossover design trial and error. Chamber testing is crucial but just the starting point to a fine loudspeaker.

 

[Speedskater]

the early Consumer Reports loudspeaker tests had big problems!
but the Harman method, Spinorama anechoic chamber test is very good at predicting how a loudspeaker will do in a real listening room.
the Floyd E. Toole book "Sound Reproduction" 'Loudspeakers and Rooms' is an excellent source of speaker testing information.

 

[Edgar Betancourt]

the early Consumer Reports loudspeaker tests had big problems!
but the Harman method, Spinorama anechoic chamber test is very good at predicting how a loudspeaker will do in a real listening room.
the Floyd E. Toole book "Sound Reproduction" 'Loudspeakers and Rooms' is an excellent source of speaker testing information.

Their methodology was anechoic chamber with on and off axis curves. The Harman method is a modern refinement of that. You must remember there were no computers even nevermind modeling software
The only problem they ever had was with Bose when they correctly pointed out that the 901 was pure crap!

 

[Edgar Betancourt]

Bose Corp. v. Consumers Union of United States, Inc. - Wikipedia

en.m.wikipedia.org

.

 

[Jon AA]

Their methodology was anechoic chamber with on and off axis curves. The Harman method is a modern refinement of that.

If I remember correctly, they did that but they were reading it wrong. They were looking for flat sound power curves. Speakers that do that will sound way too bright for most people in most rooms. Harman actually found through their blind listening tests that peoples' preference was nearly the inverse of the CR rankings (speakers CR rated the highest, actual people rated the lowest).

So, more than just a little "refinement."

 

[Edgar Betancourt]

If I remember correctly, they did that but they were reading it wrong. They were looking for flat sound power curves. Speakers that do that will sound way too bright for most people in most rooms. Harman actually found through their blind listening tests that peoples' preference was nearly the inverse of the CR rankings (speakers CR rated the highest, actual people rated the lowest).

So, more than just a little "refinement."

In fact no. They did frequency response like anyone else. It wasnt rocket science and the test were designed by PhD's. What irritated audiophiles and their sponsors was that they frequently found that non audiophile brands were frequently as accurate as non audiophile ones! Its what happens when the testers don't rely on the manufacturers they are testing money! Check stereophile's review of the 901 they went to ridiculous amount of print not to outright say that they sucked! Finally they ended with a mah conclusion! What they should have said was simple, you can't make a high fidelity speaker with PA design, 9 midranges cannot accurately reproduce the entire music spectrum simple engineering!

 

[Jon AA]

In fact no. They did frequency response like anyone else. It wasnt rocket science and the test were designed by PhD's.

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.

 

[Beave]

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.