The ASR "objectivists tribe" ;) has to read the Keynotes at the AES160th by Lars Risbo

[Frank Dernie]

Many people in here drive expensive cars, not Priuses

I drive both partly because the Prius hybrid system is a brilliant bit of engineering I admire, and yes, in my sports cars I don't use their full performance 90% of the time but I still enjoy them.

 

[kyuu]

This is reminding me strongly and uncomfortably of the whole MQA debacle with some people getting their hackles raised that anyone would dare question someone of Bob Stuart's stature. I guess I'll cease being critical in order to avoid driving other members away.

 

[TimVG]

Interesting notes. Some thoughts:

With regards to comments about standard listener preference studies, they seem to overlook several established methodological facts from Toole and Olive's research.

To start, the argument that listeners merely picked the "least-bad" room interaction due to uncontrolled low-frequency modes ignores some details. The double-blind tests used a pneumatic shuffler to move each speaker into the exact same physical location. Now, to be fair, even in the identical spot, different speakers will still excite room modes slightly differently due to variations in low-frequency extension and physical driver placement (such as woofer height). Furthermore, room size is a massive factor in how we perceive low-frequency transients in the first place, as smaller domestic spaces inevitably dominate the time domain with long modal ringing. We haven't even touched the subject of multiple low frequency sources at this point (I'm thinking of the acoustic envelopment thread), but I believe room size is the determining factor in the end with regards to perceivability of transients and direction. Let us also not forget, much of the music, movies and video games some of us enjoy are still primarily created in rooms which feature the same limitations. So in that respect, testing in rooms is certainly valid, as it is where the material is created and "consumed" (for lack of a better word).

There is also the psychoacoustic ability to "listen through the room". Humans naturally adapt to a room's stationary acoustic signature and mentally separate it from the source. This adaptation is exactly why the preference for a smooth direct frequency response and directivity (the foundation of the CTA-2034 standard) remains completely unimodal across widely different rooms and listener demographics. Granted, most of the testing was also done using direct radiating cone/dome systems. So I'm unsure if everything applies 1:1 to systems with different behaviour (constant directivity, cardioid response systems, in-wall, .. ) From a test here on ASR which featured a constant DI system (Grimani) Amir ended up tilting the slope to get a natural sound balance. Which is in line with thoughts by Kimmo Saunisto (Vituixcad). Smooth seems to at least be the common denominator.

Speaking of Kimmo, when it comes to the time domain, he has recently shared some tests on the audibility of crossover delays. While phase distortion and delays from typical crossovers are notoriously difficult to actually hear in the mid frequencies, his work highlights that they can become quite audible in the low range, in room.

To reconcile this with standard literature, we have to separate the ringing caused by the room from the delay caused by the speaker itself.

Regarding the room: Toole certainly doesn't disagree that timing plays a role in the lower register, but it depends heavily on the size of the room. The critical takeaway is that low-frequency room resonances behave as minimum-phase phenomena. Because of this, if you correct the amplitude versus frequency characteristic (e.g., EQing down a modal peak), you simultaneously correct the phase response, which effectively eliminates the temporal ringing and overhang caused by the room. However, fixing the room doesn't fix the crossover. The low-frequency crossover delays Kimmo is testing introduce excess phase (group delay) into the direct sound of the speaker itself. Room EQ cannot fix this.

As for overall transient accuracy, it remains true that chasing perfect time alignment across the board is secondary; a demonstrably smooth frequency response is vastly preferred. Our ears are simply much more sensitive to amplitude deviations than they are to time-domain smearing. But group delay naturally spikes at low frequencies, and if a steep crossover introduces enough of it between the sub and mains, the bass will audibly lag. It is one of the few areas where phase distortion genuinely crosses the threshold of audibility.

These are just my ideas, but if they are wrong - I'd be equally happy to learn so.

 

[jareinertson]

For me the real take-aways were:



I do believe in both tribes there are folks that are curious but in the end are not willing to switch tribes.




This is demonstrated by demanding 'proper blind test results' in situations where it is really hard to do so ...properly.
Audibility thresholds (of each individual) is what it is all about.


O.K. it is time for another Dutch audio-wisdom-tile.

View attachment 536524

Cause if ya ain’t Dutch, ya ain’t much! How do I know? Because my wife is Dutch!

 

[OCA]

Physical objectivity is an illusion because the universe is fundamentally a Markov chain of shifting data states.

 

[Mort]

Physical objectivity is an illusion because the universe is fundamentally a Markov chain of shifting data states.

And yet I can confidently and correctly claim that 'There are no unicorns on earth'.
Meaning is contextually dependent, isn't it.

 

[Haskil]

Physical objectivity is an illusion because the universe is fundamentally a Markov chain of shifting data states.

Is there any benefit in seeing correction software work at resolutions higher than 96 kHz, or even 48 kHz? And do we have AES publications demonstrating the listening superiority of HR formats over 16/44.1 and between 24/48 and higher?
And congratulations for everything you do in this field !

 

[OCA]

And yet I can confidently and correctly claim that 'There are no unicorns on earth'.
Meaning is contextually dependent, isn't it.

Your confidence in your claim is justified because human language has evolved to navigate the interface we live in, not the raw code beneath it

 

[Curvature]

He is also correct to point out that the masking studies were done with simultaneous tones. Are the tones still masked if they are temporally separated? And if they become unmasked, what is the temporal threshold? Because, music does not just consist of tones, they also have transients. What then, if transient smearing separates the tones? Once again, nobody knows the answer because the studies have not been done!

There are many studies about pre- and post-masking thresholds. What there aren't studies about is whether or not normal speaker driver nonlinearity has a meaningful temporal component whose masking can't be approximated by simultaneous masking thresholds.

 

[Jazigo]

Lars is right about many things, but I'll make a few contrary comments:

• He is right that psychoacoustics is an incomplete field. Not only very complex, but very underfunded.
• LF localization has very unclear boundaries. Most psychoacoustical experiments stop at 500Hz, and the available data below that is scant. Localization at 30Hz is not clearly supported by evidence, at least in terms of the kind of precise direction sensing normal at high frequencies. At VLF, as far as I've read and experienced, the most that happens is imaging effects rather than any kind of precision. The fact that externalization is occurring (when you hear sound outside of your head) is proof that there is some spatial information, but it's a stretch to call that localization. These are separate effects, a simple fact being that you can localize sound even without externalization, with headphones. Either way, the research isn't very clear cut here, not the way it is at higher frequencies. Not enough work has been done.
• He has made the claims about hysteresis distortion many times. As far as I know there is no experimental proof for this kind of haze in speaker drivers. Hysteresis isn't only magnetic (Lars has talked about "grains" in the magnet structure affecting sound in a long interview a few years ago, which I can't take seriously); slot-based cardioid speakers have acoustic hysteresis distortion which is clearly measurable (in the D&D 8c prominently). I think he usually leans on his experience with the clean sound of his drivers when asked what he means. I think the correct psychoacoustic mechanism at play is called roughness, which occurs when tones are close together. As with other things, research is scant.
• In terms of masking, it has always had spectral, temporal and spatial components: sounds are harder to hear when coming from a similar direction, with a similar bandwidth and frequency distribution and when occurring closely in time. His example doesn't make sense: the fact that you can hear a quiet tone means that it has been unmasked because some of the occlusive factors have moved out of the way, in time, space or spectrum. THD plots as we know them only show the spectral component. There's a lot more to study if we want to meaningfully disambiguate effects.
• The work to find distortion mechanisms and push the numbers lower is always meaningful. It doesn't need to be supported by audiophile consumerism (although it is, by lower-number chasers) and doesn't need to lead to subjective effects to be worth the time.

Overall he took the opportunity to speak loosely and expansively. I think he would have done better to focus on his work whenever he had the impulse to reach for something bigger. Would have been more interesting.

Personally, I think Lars is hitting it home with more tests needs to be in the time domain. I have usually wrapped things over in the frequency domain and notched the frequencies and calculated their power envelope. While trying to build a null-test (the golden test) in the time-domain, things starts to get very rough fast.

 

[PristineSound]

The point on how many measurements are a snapshot which leaves temporal related data out. That reminds me of Iconoclast cables, which Amir bench tested.

Without getting into the details, I have my share with dealing with some of the best doctors (not for myself), and often times, if they have a hunch, they will never say it outright publicly, if there aren't proven science to back it.

I applaud Lars, it's bold as a man of science to express hunch with such conviction without having any substantial science backing it (or maybe there is and Purifi just doesn't want to share it). Now, if Lars can provide such science, then we have advancement in this hobby. Until then, one tribe is still not convinced.

 

[AndyBell]

Surely if any of this were true then properly controlled double blind tests would verify it.

I am totally open to the idea that double blind tests COULD reveal something not yet measured/measurable.

In the absence of these or measurements showing something is different then it’s just conjecture.

 

[olieb]

I have a question about something in the text.

The dropped coordinate was time.

The proof is sitting in our own toolbox — and it is almost funny. The exponential sine sweep, the standard method we use to generate THD plots, works by spreading harmonics out across time. The harmonics of the sweep arrive at the listener’s ear at different times.

Temporally separated like that, people routinely detect distortion products even at minus 80 decibels. By ear and without effort.

I just don't understand what he is saying.
How do the harmonics get temporally separated from the fundamental (and each other?) during a sweep?
To me, he says that for a 1kHz tone (at the respective moment in the sweep) the 2, 3, 4 kHz harmonics will not reach the ear at the same moment as the fundamental. What is he talking about?
Are my ears broken, as I seem to hear the fundamental and harmonics all at the same time?

EDIT:
Nobody, who can explain Risbo's statement? After all this is an AES keynote, there should be substance. Genuinely confused and curious.

 

[Buckchester]

It would be very sad to see you go.

I believe that Lars Risbo is correct. He is essentially arguing that the field of psychoacoustics is underdeveloped, and many of our questions have not been answered. I too am FAR from a psychoacoustician, but I have read the papers. What I saw were a whole bunch of studies that try their best, but ultimately do not answer relevant questions. One example: we can not extrapolate studies done with headphones to listening rooms, yet we routinely do that. Not because we are stupid, but because we have no choice. There may be NO relevant studies done in listening rooms, as Dr. Risbo points out.

He is also correct to point out that the masking studies were done with simultaneous tones. Are the tones still masked if they are temporally separated? And if they become unmasked, what is the temporal threshold? Because, music does not just consist of tones, they also have transients. What then, if transient smearing separates the tones? Once again, nobody knows the answer because the studies have not been done!

My take-aways from that keynote:

1. Stop being so bloody dogmatic! Be more open-minded about what may be audible, and what may not be audible.
2. Be careful about what the studies say, and more importantly: what they don't say. Do not place too much weight on secondary endpoints. The study was not designed to look at those, and drawing conclusions from it may be misleading.

These are all points I wholeheartedly agree with. Dr. Risbo might call himself an engineer, but he is no mere engineer. He has proven that he understands science more than any of those people on ASR who are complaining about his talk. Let alone some of you who have clearly not understood what he said or skimmed the article before bloviating your half-baked thoughts on ASR. You are an embarrassment - when I see that, I sometimes feel like youngho and wonder why I am part of this community (the answer is, there is nowhere else to go ... otherwise I would have gone there ages ago!).

Hmmm, kinda sounds like you’re engaging in some “bloviating” of your own.

 

[Bjorn]

The part about acoustics and preferences and the weakness of a research, whether it's Harman or somewhere else, is obvious and something I've discussed here several times.

The result of such test only gives answers to that particular room with its acoustics and with the type of music is used.

When the speaker is far away from the side wall and there is also maybe absorption on the rear wall that will influence the result. With very close proximity to side walls and for instance proper diffusion on rear wall, you'll have a different result.

When someone claims objective truth about directivity preferences based on a test under mainly one acoustic circumstance without paying attention to the time of arrival of side wall reflections etc. it's clear it's not much science.

It does not help much either if the narrower speakers tested don't have broadband directivity. Or cleae abberations in the directivity like for instance JBL M2 has.

The way the the so called science of preferences are operating today, there will continue to be a major split between this community and a large part of the audiophile community.

What disappoints me the most is the lack of humility with the obvious weaknesses in the tests and the so quick attitude to claim objectiveness.

This probably doesn't reach many of you, but I personally have had speakers with broadband wide directivity and broadband narrow directivity for years and tested them against each other in several rooms with many positions and music genres .

For critical listening I generally prefer narrow broadband directivity speakers, though there are certain things you can achieve with a wide directivity speaker with especially the use of diffusion (phase altering) on opposite side walls. And clearly things change for a backround speaker in a very wide room. Music genres also plays a role.

 

[voodooless]

I applaud Lars, it's bold as a man of science to express hunch with such conviction without having any substantial science backing it (or maybe there is and Purifi just doesn't want to share it). Now, if Lars can provide such science, then we have advancement in this hobby. Until then, one tribe is still not convinced.

Just remember that, next to a scientist and an engineer, he is also a businessman.

 

[Sokel]

The irony is that what looks like an excommunication here, is really a family affair .

Yes, Purifi is a business and as such it needs a story, there's no way around that for survival.
Thing is though, that Lars is one of the people who absolutely belongs at this very community (who now seems in distress) and has built gear that this community pray to.

What's our problem? The words?
I haven't seen trashing Kef about Muon for example, even though is the pinnacle of the "old" way.

There's a smell of old, sheltered/settled people who refuse to even doubt, and that's a shame either scientifically or humanly.

(if this forum wants a real scientific tag to it, it must ban one single word: "my" .
This "my" is the fountain of all evils.

 

[Hayabusa]

"One mechanism we focused on was magnetic hysteresis distortion in loudspeaker motors. You may have heard this in a system without knowing what it was. A granular texture in the sound."

Please explain what is wrong with this? I don't get it..

 

[whereisbrian]

I also want to point out that the thread title is whimsical and well-chosen from a marketing perspective.
The objectivist will be provoked and will automatically dive into the topic.
The subjectivist will be delighted and will use it as a new argument.
Perfect!

 

[voodooless]

What's our problem? The words?

Words matter.. Practice what you preach.

I haven't seen trashing Kef about Muon for example, even though is the pinnacle of the "old" way.

Is it? They are very much engineering-driven, optimizing their speakers in all kinds of details that might or might not be audible in practice. They don't seem to imply that we're missing critical parts in our measurement systems. Likewise, Topping isn't claiming that their 125 SINAD DAC sounds much better than their previous 124 SINAD DAC.

Please explain what is wrong with this? I don't get it..

The problem is that the facts are not in evidence. There is no research indicating that magnetic hysteresis distortion leads to a granular texture in the sound.