Is Jay talking about Amir?

[kotmj]

Stopped at 4 min. By then, it is quite clear he does not have basic knowledge on the subject he is tutoring - electroacoustics, measurement technique, signals and systems.

I gave this video a chance. There are many yt-presenters out there doing audio/hifi related content, of which most are at best right out misleading. They pose as experts, but lack basic knowledge about the technology and physics behind - basic knowledge that has nothing to do with preference or subjective-vs-objective or opinion.

It is not wrong to do a video about your opinion on a product and its sound, like you experience it. But it is wrong to present technical information that is incorrect according to established science.

Science is not opinion. It is not democratic. You can not vote to find the answer. Science can be challenged, yes, but then you must start by learning what is already known, and then you can put down a hypothesis and prove though theory and controlled experiments that your new discovery is correct.

Well said!

 

[kotmj]

That's absolutely not what "Toole says". (he said and still says many interesting things obviously, but this is about the part you are alluding to)

What Toole did (very heavily summarized)

1) collected preference ratings.
2) made different objective measurements.
3) derived a model that is the "best" (for the dataset available) at predicting preference ratings by giving weights to measurements. (see https://patents.google.com/patent/US20050195982A1/en)

-
from a plurality of candidate independent variables indicative of loudspeaker sound quality, selecting a set of independent variables X1-Xn determined as maximizing the ability of a loudspeaker preference variable Y1 to predict a loudspeaker preference rating;
performing a multiple regression analysis to determine respective weights b1-bn for the selected independent variables X1-Xn; and
arranging the weighted independent variables into a linear relationship on which the loudspeaker preference variable depends according to:

Y 1 =b 0 +b 1 X 1 +b 2 X 2 +b 3 X 3 + . . . b n X n,
where n is the number of selected independent variables.
-

There are a lot of reasonable angles to attack the research if one is so inclined. If the field really mattered, you'd see people improving the predictor by 0.x% every few months or so, either by collecting more preference ratings, selecting different variables, using different types of regression or, other methods. But the field isn't really essential to mankind's survival and Toole's predictor is the best we have right now.

In fairness, the blind generalization and simplification of what "Toole says" isn't found exclusively in the subjective/anti-Toole crowds...

Thank you for the explanation. I suppose I shall have to buy his book. From a lecture he gave, he really didn't do justice to his ideas.

 

[Music1969]

Is that the "research" where the various listeners were told to do what they like with the bass and treble controls?

Question for @Sean Olive ?

 

[redshift]

Sorry guys...

It's probably just me, but I agreed with just about everything Jay said. It's about speakers... not electronics.
In short.. he says measurements matter, they should be done properly and for speakers those measurements are really hard to interpret aside from those that know them well. It is also hard to predict how they will sound in your room to your ears.
Jay states measurements are a tool that can narrow down a long list of options and that's certainly true.

The 'tape measure thing' he states is a bit weird at first glance but in the end what he means is that once you have speakers they must be positioned correctly and use measurements when experimenting so you can pick the optimal position and after the experiments you can use the tape measure and plots you made to put it in that optimal position.

Jay is not talking about Amir imo. He talks about speakers, measurements, positioning and listening tests in general and that these all matter.

Yes, rooms “ruin” speakers designed to perform well in anaechoic chambers. That is for sure.

The need of heavy EQ’ing and room “treatment” is a telltale sign that something has gone wrong with speaker design for quite some time.

 

[amirm]

The need of heavy EQ’ing and room “treatment” is a telltale sign that something has gone wrong with speaker design for quite some time.

At low frequencies the room is in charge. You can put any speaker in there and you still need to EQ. To not would be to ignore elementary physics.

 

[redshift]

At low frequencies the room is in charge. You can put any speaker in there and you still need to EQ. To not would be to ignore elementary physics.

I’m thinking furniture, walls, floors and ceilings dampen/reflect/refract/comb filter plenty of the audio signal in normal rooms?

Try turning your speakers 180° directly into the (back) wall. I’m quite sure you’ll hear the mids and highs nonetheless, however attenuated.

ABX test it?

 

[solderdude]

It is not wrong to do a video about your opinion on a product and its sound, like you experience it. But it is wrong to present technical information that is incorrect according to established science.

Science is not opinion. It is not democratic. You can not vote to find the answer. Science can be challenged, yes, but then you must start by learning what is already known, and then you can put down a hypothesis and prove though theory and controlled experiments that your new discovery is correct.

Are we talking about the same video ?

The video was not about an opinion of a product nor its sound. There was no technical info presented.
He also did not say science is an opinion and values measurements. He questions how people interpret measurements and how the measurements come about.

 

[kimmosto]

If the field really mattered, you'd see people improving the predictor by 0.x% every few months or so, either by collecting more preference ratings, selecting different variables, using different types of regression or, other methods. But the field isn't really essential to mankind's survival and Toole's predictor is the best we have right now.

This motivates me ask one more (last) time why also authors on ASR are willing to maintain situation where few magnitude responses at some constant level are almost the only variables measured? Spinorama and Preference rating include just magnitude responses at moderate level. That part should be okay of course, but it's not enough at least for me as potential client and speaker designer.
Measurement set limited to magnitude responses and THD could also prevent development; using old data and upgrading to more inclusive objective ratings in the future. Current situation could also bias some manufacturers to design speakers which behave well in spinorama, but otherwise are weak. That wouldn't be "Emissionsgate" though product is optimized to get high score in certain test. Just enough to convince members on ASR and more people to buy.

 

[Mart68]

This motivates me ask one more (last) time why also authors on ASR are willing to maintain situation where few magnitude responses at some constant level are almost the only variables measured? Spinorama and Preference rating include just magnitude responses at moderate level. That part should be okay of course, but it's not enough at least for me as potential client and speaker designer.
Measurement set limited to magnitude responses and THD could also prevent development; using old data and upgrading to more inclusive objective ratings in the future. Current situation could also bias some manufacturers to design speakers which behave well in spinorama, but otherwise are weak. That wouldn't be "Emissionsgate" though product is optimized to get high score in certain test. Just enough to convince members on ASR and more people to buy.

If the speaker has good measurements it's not going to be 'weak'. It may not be what a specific individual is looking for but that's not the same thing.

Measurements are an excellent way to narrow down the field. Otherwise there's so many options you could spend several lifetimes trying all of them.

Personally I buy second hand, based on measurements, live with it for a while, if it's not for me I sell it on. That approach has (eventually) got me to exactly what I wanted without having to write-off a small fortune and without wasting any time trying out junk.

 

[redshift]

This motivates me ask one more (last) time why also authors on ASR are willing to maintain situation where few magnitude responses at some constant level are almost the only variables measured? Spinorama and Preference rating include just magnitude responses at moderate level. That part should be okay of course, but it's not enough at least for me as potential client and speaker designer.
Measurement set limited to magnitude responses and THD could also prevent development; using old data and upgrading to more inclusive objective ratings in the future. Current situation could also bias some manufacturers to design speakers which behave well in spinorama, but otherwise are weak. That wouldn't be "Emissionsgate" though product is optimized to get high score in certain test. Just enough to convince members on ASR and more people to buy.

I’d like to see measurements from “normal” rooms at a rather wide listening (floor) area. It shall sound good for Bart, Lisa, Homer, Maggie and Marge in the Simpsons sofa.

I guess the causal Hifi enthusiast (with attachments) won’t accept the obnoxities of “sweet spot” listening.

Anaechoic measurements is in the realm of supportive measurement because that isn’t the situation in which people find themselves listening to music. For sure it is useful to weed out some obviously rancid stinkers from the ensemble.

 

[kimmosto]

You can put any speaker in there and you still need to EQ.

Not necessarily. This depends on room, speaker concept, how listening setup i.e. speakers and listener locations are adjusted and of course personal tolerance; how high peaks sound unnatural and bass notes too difficult to follow. For example EQ at any frequency has not been mandatory in our house, but possible locations for traditional boxed speakers are quite limited.

 

[kimmosto]

If the speaker has good measurements it's not going to be 'weak'. It may not be what a specific individual is looking for but that's not the same thing.

This depends on how you specify "measurements". Just magnitude responses and THD allow speaker to be weak and lame and sometimes also other unacceptable (imo) features.

 

[Mart68]

This depends on how you specify "measurements". Just magnitude responses and THD allow speaker to be weak and lame and sometimes also other unacceptable (imo) features.

If the speaker has low distortion and a reasonably flat anechoic response then it's probably worth trying. If it doesn't, then disregard it. No need to make it any more complicated than that IMO.

There's a lot of complete rubbish out there often at high prices, it's easy to avoid them just based on those two metrics alone.

 

[redshift]

Not necessarily. This depends on room, speaker concept, how listening setup i.e. speakers and listener locations are adjusted and of course personal tolerance; how high peaks sound unnatural and bass notes too difficult to follow. For example EQ at any frequency has not been mandatory in our house, but possible locations for traditional boxed speakers are quite limited.

Yep, I consider the room as an integrated component of the playback system and not a necessary evil. There is a reason people prefer the acoustics of the room than listening in an anaechoic chamber.

Overly “treating” and EQ/DSP:ing for sure will make that a worse experience than a properly designed speaker that isn’t afraid of singing with the room that in which it is located.

I’m running my rig with some loudness only for psychoacoustic compensation. BTW; my sweet spot:

 

[PierreV]

Just enough to convince members on ASR and more people to buy.

Well, I imagine that a commercial entity supporting/funding research about "what most people prefer" is obviously interested in the "what most people would buy" secondary outcome . In order to maximize sales, maximize the probability your speaker will receive a high average preference score among the target audience. The research can be and was developed further by age groups and sex I believe.

To take a medical analogy, very good survival (and even better survival with good quality of life) models can be developed if one takes physical exercise, smoking, alcohol consumption and, social contacts as parameters. Those are very valid statistically, on a much bigger scale and with a higher confidence degree than speaker preference models. That doesn't mean that you will not find many individual outliers.

The same happens with speakers: outliers, in terms of preferences, abound. There's nothing intrinsically wrong with that.

Where, imho, some ASR enthusiasts jump the gun is when they enthusiastically treat a "best fit" model between a population of listeners and a population of speakers as a strong predictor of individual preference. As quite a few others have noted here before, the value and the confidence levels of the model and derived scores as an individual predictor of preference are quite weak.

One can hardly blame Toole, Olive and, others for that over-enthusiastic extension.

I think that, just as there are subjectively inclined people willing to embrace any nice story because it is nice, there are objectively inclined people willing to trust hard numbers simply because they are numbers

 

[jae]

The video was not about an opinion of a product nor its sound. There was no technical info presented.
He also did not say science is an opinion and values measurements. He questions how people interpret measurements and how the measurements come about.

To what end? He isn't challenging any scientific methodology, he isn't even giving any examples of poor measurement or showing what measurements may matter to him or a viewer. He isn't explaining why they don't matter in any practical or usable way. Is there a reason he does not show how to conduct proper measurements, or exclude ones that don't matter? Why does supposed neuroscientist not even conduct a proper blind test (and dismiss it as "for fun" when called out on it)? Or even link to a source doing an using measurements correctly?

He implies that in the absence of certain measurements he is able to make acoustic conclusions on gear that measurements cannot, since that is effectively what every video on his channel is about despite maybe not directly doing so in this video. Appearing to be an impartial or educated scientist and just "questioning" it is a poor facade aimed to sedate or silence those that question his influencer status and "experience growing up in a studio and listening to thousands of speakers", while simultaneously appearing credible to his supporters. It's a lot of effort and a long video to say absolutely nothing of value.

 

[redshift]

Well, I imagine that a commercial entity supporting/funding research about "what most people prefer" is obviously interested in the "what most people would buy" secondary outcome . In order to maximize sales, maximize the probability your speaker will receive a high average preference score among the target audience. The research can be and was developed further by age groups and sex I believe.

To take a medical analogy, very good survival (and even better survival with good quality of life) models can be developed if one takes physical exercise, smoking, alcohol consumption and, social contacts as parameters. Those are very valid statistically, on a much bigger scale and with a higher confidence degree than speaker preference models. That doesn't mean that you will not find many individual outliers.

The same happens with speakers: outliers, in terms of preferences, abound. There's nothing intrinsically wrong with that.

Where, imho, some ASR enthusiasts jump the gun is when they enthusiastically treat a "best fit" model between a population of listeners and a population of speakers as a strong predictor of individual preference. As quite a few others have noted here before, the value and the confidence levels of the model and derived scores as an individual predictor of preference are quite weak.

One can hardly blame Toole, Olive and, others for that over-enthusiastic extension.

I think that, just as there are subjectively inclined people willing to embrace any nice story because it is nice, there are objectively inclined people willing to trust hard numbers simply because they are numbers

Right; it is possible to tweak and twiddle with EQ/DSP and room treatments until “perfection” arises in the minuscule “sweet spot”, conveniently forgetting that room acoustics is a fundamental thing that makes listening to music enjoyable.

Absolute playback fidelity:
Headphones > near fielders > far fielders

Relative subjective quality:
Far fielders > near fielders > headphones

IMO.

 

[solderdude]

To what end? He isn't challenging any scientific methodology, he isn't even giving any examples of poor measurement or showing what measurements may matter to him or a viewer. He isn't explaining why they don't matter in any practical or usable way. Is there a reason he does not show how to conduct proper measurements, or exclude ones that don't matter? Why does supposed neuroscientist not even conduct a proper blind test (and dismiss it as "for fun" when called out on it)? Or even link to a source doing an using measurements correctly?

I am not judging the person nor what else he does and has done. I am merely reacting to the video itself.
And no he isn't questioning properly conducted measurements. Nor does he give examples.

It is just a video addressing that speaker measurements have to be done properly. You can do this in various ways. There are good and less good and improper ways.

He indeed isn't explaining why measurements don't matter to him because he says properly made measurements matter. He also says that (obviously for acoustical measurements) they don't tell the whole story. Certainly not to the general public and measurements made by the general public say nothing unless those are accompanied with essential info.

Why should he show a blind test in a video that broadly addresses why measurements matter and that on the web you can find good and poor ones.

I have not looked at his many other videos and frankly I don't care because I don't watch any (watched Amir's and a few others to see what they say)
Only adressing this particular video about this particular subject where he tries to convey this isn't an easy task.

 

[redshift]

I am not judging the person nor what else he does and has done. I am merely reacting to the video itself.
And no he isn't questioning properly conducted measurements. Nor does he give examples.

It is just a video addressing that speaker measurements have to be done properly. You can do this in various ways. There are good and less good and improper ways.

He indeed isn't explaining why measurements don't matter to him because he says properly made measurements matter. He also says that (obviously for acoustical measurements) they don't tell the whole story. Certainly not to the general public and measurements made by the general public say nothing unless those are accompanied with essential info.

Why should he show a blind test in a video that broadly addresses why measurements matter and that on the web you can find good and poor ones.

I have not looked at his many other videos and frankly I don't care because I don't watch any (watched Amir's and a few others to see what they say)
Only adressing this particular video about this particular subject where he tries to convey this isn't an easy task.

I’d say that measurements is all that matters in the room in which the speakers are placed, however not only the direct path, add in the first, second and third reflection. After all, the hearing and biological DSP seem fully capable of doing some pretty sophisticated 3D-mapping, EQ:ing and spatiotemporal correction seemingly effortless and by default.

For the most ardent audiophile Hifi Puritan, the measurement 3D volume(s) should be exactly where your ears are located in the room.

For the casual Hifi enthusiast (suspect?), perhaps 25 to 50% of the room’s volume, or at least cover the fave sofa or couch with some margin to spare.

Now that is a formidable endeavor to do measurements on and to map those to what makes a speaker sound “good” (or bad).

I’d say that anaechoic measurements are perfectly fine for determining which speakers could be expected to perform somewhat reasonably well given the acoustic “chaos” in a normal living room.

IMO.

 

[sq225917]

I turned off half way through his first straw man.

However I'd disagree strongly with the previous poster, all that matters is that listener likes the sound, how that is arrived at, by measurement of or sheer luck, is moot. The pragmatic amongst us will take a measurement lead approach, others may like to feel like they've had their hand held by their dealer as they undertook a fantastical voyage of discovery leading ulimately to a beautiful climax where they opened their wallet, most likely, for whatever the dealer had in stock...

It's just hifi