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

[tengiz]

For those preferring TL;DR version of the intended spirit of the essay:

Whether it's true to this logos, it's another question entirely.

 

[Sal1950]

Distortion is (depending on how you look at it) defined as any change to the signal OTHER than noise. In a colloquial sense you could call noise distortion, but in a technical / engineering sense they're two exclusive categories.

I'll take your word on that but still hang onto my opinion there.
Just as the SINAD grading is a single reference number that can be used to determine how badly the output does or doesn't reflect the input.

 

[MattHooper]

It's valid if you are writing fiction, perhaps on a site like "Audio Fiction Review dot com"

No, it’s valid in the empirical investigation and reasoning domain… science included.

Science can work from observation first, and then an explanation of the observation. It can also work from proposing the existence of a phenomenon, and if it exists describing the type of characteristics it should have. And then we go looking for it. If a proposed phenomenon is hypothetical if anything that makes the description more important so you know what you’re looking for!

And I’ve given examples for why the idea that you cannot use any descriptive tools at our disposal, especially analogy to already established phenomenon, it’s just obviously wrong. We constantly use an analogy to established phenomena to help characterize new or proposed possible phenomena.

Before anybody ever produced an actual set of stereo loudspeakers, it would’ve been completely valid to describe certain predicted properties by borrowing from other established sense domains - eg analogy to sight: the concept of producing “focused” sonic “images” within the stereo field.

And I pointed out several examples which includes the fact that engineers/specialists/scientists defined and described A.I. by analogy to human intelligence, long before we even had ChatGPT… and on our way, possibly to artificial general intelligence. In fact characterizing it that way is what gave them a direction to shoot for, and also led to analogous structural tracks like “ neural networks” …. Some of the AI we used today are direct descendants of such biological analogy.

Anyway… yes I understand you and some people are sceptical about Purifi’s claims regarding audible hysteresis distortion. I am too. But we shouldn’t let our scepticism misfire.

There are two questions we need to keep separate here:

1. Does the phenomenon exist? (audible distortion from Hysteresis).

2. If it exists, what would it’s characteristics be?

Suppose Purifi says: “Hysteresis produces a grainy character.”

We skeptics might reasonably challenge:
“Show me evidence that listeners can reliably detect this effect.”

That’s question #1.

But it would be strange to say:

“Until you’ve proven the effect exists, you’re not allowed to describe what you’re claiming listeners hear.”

Because if you’re proposing a phenomenon, you pretty much have to describe its expected characteristics. Otherwise nobody knows what observation they’re supposed to look for!

Right?

@voodooless do you agree? (I ask because it looks like you earlier supported RexrothPigeon’s objection).

Cheers.

 

[Sokel]

The problem is that we take this keynote as if it was a white paper or something.

The experts's comments on Facebook draw the true picture and the normal scientific way.

And what's this way? How it works, even for the most groundbreaking stuff?
We allow assumptions. We grant assumptions.
Even the greatest was granted with such.

And what's the way that normal questions and curiosity turn to (somehow) valid assumptions usually other than following others work?
Plain, old, rigorous, careful observation, trial and error, etc.

Anyone following MIT's latest research without having a clue (as I for example) will be very surprised about the WILD assumptions about consciousness for example.

So unless someone found the fundamental all-truth about everything and close all arguments in existence, pleas allow the rest of us having fun playing with this world.
Dogma is boring, to say the least.

 

[youngho]

I was intrigued when he proposed a bi- or mulitmodal distribution of listener preference (in the case he was referring to, speaker performance), because I'm pretty sure Toole has written about that very thing with regard to related research on preference: for 'envelopment' vs 'detail', with the general consumer tending to prefer the former, while audio production engineers (and IIRC, musicians? I forget the details) the latter. (I take that as a proxy too for mid/far field listening vs near field. It makes intuitive sense.)

I can't tell whom you're referring to by "he," but if @Floyd Toole , he has written in multiple editions of Sound Reproduction text like "Ando et al. (2000) found that musicians judge refl ections to be about seven times greater than ordinary listeners, meaning that they derive a satisfying amount of spaciousness from refl ections at a much lower sound level than ordinary folk: “Musicians prefer weaker amplitudes than listeners do.” It is logical to think that this might apply to recording professionals as well, perhaps even more so, because they create artificial reflections electronically and manipulate them at will while listening to the effects. There can be no better opportunity for training and/or adaptation. In fact, it is entirely reasonable to think that acousticians who spend much of their lives moving around in rooms while listening to revealing test signals can become sensitized to aspects of sound fields that ordinary listeners blithely ignore. This is a caution to all of us who work in the field of audio and acoustics. Our preferences may reflect accumulated biases and therefore may not be the same as those of our customers" (this is from the second edition). Two groups

Also Toole: "In Fig. 14 the author has modified the original data to separately show the result of evaluations by trained and untrained listeners. This is compared to the small room prediction from Fig. 13(a). The “all listeners” average curve is close to the predicted target, except at low frequencies where it is apparent that the strongly expressed preferences of inexperienced listeners significantly elevated the average curve. In fact, the target variations at both ends of the spectrum are substantial, with untrained listeners simply choosing “more of everything.” An unanswered question is whether this was related to overall loudness—more research is needed. However, most of us have seen evidence of such more-bass, more-treble listener preferences in the “as found” tone control settings in numerous rental and loaner cars. More data would be enlightening, but this amount is sufficient to indicate that a single target curve is not likely to satisfy all listeners..."
http://www.aes.org/e-lib/browse.cfm?elib=17839. Two groups that they smooshed/averaged into one

If @Sean Olive , https://www.audiosciencereview.com/...er-preference-curve.50602/page-4#post-1845069. You can see some of my issues with what's commonly quoted as "research" in this forum at https://www.audiosciencereview.com/...er-preference-curve.50602/page-3#post-1819700 and https://www.audiosciencereview.com/...er-preference-curve.50602/page-5#post-1848097. Three groups? I'm still skeptical about the way this is presented

If @j_j , I'm sorry, I'm super-foggy on the details, but I seem to recall that he's written about certain groups of listeners being more sensitive to pre-echo or other artifacts of encoding. I can try to find it at a later date.

Distinct groups of listener preference go all the way back to concert hall acoustics: https://www.audiosciencereview.com/...cert-hall-acoustics-links-and-excerpts.51487/

 

[krabapple]

Well, EQ equipment was very popular in the 80s, every halfway ambitioned system seemed to have one, so nothing new under the sun

And bass/mid/treble tone control knobs were around before graphic EQ. Who said anything about EQ being new?

I had a Technics one with like 20 sliders or so, mostly used to decrease human voice frequencies so we could karaoke (the EQ box had two microphone inputs). At that age, many parties were celebrated that way at my place!

Doing it based on hopefully accurate measurements is what is awesomely new. And of course I have done it, but I didn't adjust it the way Dirac wanted entirely because I wanted to preserve the "character" of my system, not slavishly bow over. The measurements confirmed my listening impressions and allowed me to do sub adjustments (volume, position). An "ideal neutral system" is not my preference, although I am probably >85% there.

If you're; using Dirac at all, you're doing 'a lot of EQ', compared to the old days and certainly compared to the purist audiophile/high enders who eschew EQ completely.

 

[RexrothPigeon]

Suppose Purifi says: “Hysteresis produces a grainy character.”

We skeptics might reasonably challenge:
“Show me evidence that listeners can reliably detect this effect.”

That’s question #1.

But it would be strange to say:

“Until you’ve proven the effect exists, you’re not allowed to describe what you’re claiming listeners hear.”

Nobody's telling anybody else what they're allowed to do. You write thousands of words a week here and nobody stops you.

Anyway when we start theorizing about the strength and speed of bigfoot, we are either presupposing the existence of bigfoot or we are writing fiction. So while yes of course you are allowed to write whatever nonsense you want (as you always do, at great length), if you put the cart before the horse you should accept that people are going to notice it and say so and that nobody serious will accept that baseless theorizing.

 

[pablolie]

... If you're; using Dirac at all, you're doing 'a lot of EQ', compared to the old days and certainly compared to the purist audiophile/high enders who eschew EQ completely.

I don't use EQ at all in my reference setup. I used Dirac to set up my sub(s) where it belongs best, and to not be totally unused (as would be my natural preference, honestly).

I use PEQ a little in my headphone DAC, but I don't go even remotely go anywhere near the radical corrections that are suggested with my headphone choice because I chose them for their out of the box natural sound, but a little finetuning within 1dB plus or minus here and there seemed even nicer.

 

[youngho]

hi, thanks yes i agree with your criticism. It is not as simple as i put it and I am probably to used to look at the deconvolved response. however, the deconvolution is simply a filtering operation which averages the chirp it self out save for one pulse. the harmonics and their time encoding is in the signal we hear.

Cheers

Lars

Thank you for your considered responses to @KSTR (and vice versa) and mutual discourse. This should be considered ASR interactions at the finest/highest level, in a similar line with @Floyd Toole , @Sean Olive , and @j_j 's responses to Sean Olive's post on Facebook

https://www.facebook.com/sean.olive/posts/pfbid0rrnkEfkkRsVdsn4GcS3wZF5bDRJFZdWtc9caq3autARHv4FR34JfcNg2UwcTNdWYl

. No snark, no "winning". Just a straightforward exchange of ideas and opinions.

 

[Keith_W]

@youngho it's people like you who helps make ASR worthwhile. Your summaries of studies along with links has been incredibly helpful.

I don't consider Dr. Toole's writing to be reference material, rather I consider it to be a summary of available evidence. OTOH his studies are reference material. The real value in his book is the reference section, where I can look up the original studies and decide for myself if Dr. Toole has presented their findings adequately or not. This is especially important in controversial areas of audio where the evidence is weak. That happens a lot in psychoacoustics! In many cases, because of limitations of length, Dr. Toole is forced to simplify the core message of those studies, sometimes to one line or less. This is not a criticism of Dr. Toole, it is an acknowledgement that textbooks have limitations. By necessity, a lot of nuance has to be left out.

And yes, I don't care about "winning" either. I will happily lose an argument if it means that I have learnt something and it has moved my understanding closer to the truth. I'll even tell that person that they are right, and I thank them. But ... if I think that someone is full of snark, or not engaging in the spirit of open minded discussion, I ignore them. There is no need to reply.

So ... please stay. You have your fans and people who appreciate you here.

 

[MattHooper]

Nobody's telling anybody else what they're allowed to do.

You know that point was referring to this (my emphasis):

When trying to explain something whose existence is strongly contested AND has not been quantified or proven in any way, you simply can't analogize with something whose existence is universally accepted AND is easily quantified and proven.

That’s a very strong, confident claim… which i’ve argued is incorrect for all the reasons I’ve given.

It could have been more fruitful if you’d addressed points made against your claim, rather than repeating the same claim and suggesting the counter arguments are just nonsense.

Anyway when we start theorizing about the strength and speed of bigfoot, we are either presupposing the existence of bigfoot or we are writing fiction. So while yes of course you are allowed to write whatever nonsense you want (as you always do, at great length),

I’m discussing this in good faith. I take the time to carefully respond to what you wrote, which I think is a way of paying respect rather than just brushing it off lazily. I would be happy to be shown where I have written something incorrect.

If you don’t want to respond in kind…ok. But could you please refrain from substituting ad hominem?

if you put the cart before the horse you should accept that people are going to notice it and say so and that nobody serious will accept that baseless theorizing.

By the same token, if you post here a bold claim, declaring a principle relating to science or engineering or empirical reasoning that is fallacious, you should expect someone will point it out, right? That’s pretty much what we do here . It shouldn’t be an occasion for ad hominem.

You made a wide-ranging, generalized claim about empirical reasoning that is clearly untenable and inconsistent with science. Slogans like putting carts before horses don’t address the specific errors pointed out in your claim.

But we will leave it there and see if somebody else wants to defend your claim.

 

[voodooless]

@voodooless do you agree? (I ask because it looks like you earlier supported RexrothPigeon’s objection).

You can both be right, can't you? I agree with you that you can qualify something you experience without having it measured. But at the same time, it's just your qualification.

If Lars talks about granular texture sound, it doesn't mean he talks about the same phenomenon you talked about. Similar to there being lots of different types of noise you can apply to images. Just saying that the image is noisy really doesn't cut it. You'll have to be very specific with what you see to deduce the actual type of noise. And even if you don't know the noise source, it's still helpful to qualify it as accurately as possible. "granular texture", and some flowery prose around it, is not nearly exact enough.

 

[Sokel]

I'm going to politely disagree. One problem is that there are a number of members who write things like "Research shows" or "As I recall from studies" without any references or actual representation of actual findings. Another is that many members oversimplify matters to write things like "Listeners can listen through rooms" or "The Harman curve is preferred" or "Bass below 80 Hz is not localizable" (and suggest that there is no contribution of non-mono bass below 80 Hz to listeners' experience and enjoyment, BTW) without consideration to the study design, study population, and ability to extrapolate the study findings. For example, it would be trivial to do a study of a few dozen listeners and conclude that the ability to identify pitch without external reference is not verifiable, especially if you just average the results of all listeners. Then people can summarize as "Listeners can't identify pitch" or "Pitch is not identifiable." Maybe they'd go even further ato state the matter as "Objective research shows that listeners do not prefer absolute pitch."

In my opinion, too many posters like @kyuu and @Mort here are trying really hard to not allow the others to have "fun playing with the world." I'm not even playing with the world but rather trying to explore and understand what's been written/published and posted here and elsewhere, adding references and direct links whenever possible. This forum is called AUDIO SCIENCE REVIEW. In science, review articles include references and nuanced summaries. It seems to me that for users like @kyuu and @Mort, this is more like OBJECTIVIST DOGMA REGURGITATED with as much snark and as many one-liners as possible. It's like they're determined to win a race to the bottom.

Anyway, I got a notification from @Keith_W referencing me that pulled me back in momentarily. Please don't feel any compelling need to include my username in replies. Rather, feel free to quote my text without identifying me specifically, and I concede that You Win!!! Congrats! I'm humbled by your snark.

[EDIT]
Oh wait, I forgot to delete the space in "refl ections." Probably @kyuu and @Mort won't able to read that ("No more trouble than writing (or translating) coherently with punctuation" and "Your copy-pasta job (or whatever that is) is extremely difficult to read, so what nuance may be there isn't likely to get picked up on."). My apologies, "ref lections" means "reflections." I hope that makes it more coherent and easier to read, so that you can detect the nuance.

Also "ato" was supposed to be "as to." Apologies again, and I hope that clarifies things

You're of the ones I appreciate the most, you research for references at the thread about Bass for example is monumental.

The way to survive in places like this is filtering and identifying intentions.
That's the way i do it, and sometimes I agree, members may appear harsh and dismissive. Humor is the way to deal with it.

Thanks for your work!

 

[voodooless]

I'm going to politely disagree. One problem is that there are a number of members who write things like "Research shows" or "As I recall from studies" without any references or actual representation of actual findings. Another is that many members oversimplify matters to write things like "Listeners can listen through rooms" or "The Harman curve is preferred" or "Bass below 80 Hz is not localizable" (and suggest that there is no contribution of non-mono bass below 80 Hz to listeners' experience and enjoyment, BTW) without consideration to the study design, study population, and ability to extrapolate the study findings. For example, it would be trivial to do a study of a few dozen listeners and conclude that the ability to identify pitch without external reference is not verifiable, especially if you just average the results of all listeners. Then people can summarize as "Listeners can't identify pitch" or "Pitch is not identifiable." Maybe they'd go even further ato state the matter as "Objective research shows that listeners do not prefer absolute pitch."

Isn't this exactly what Lars is doing in his keynote, though?

 

[Bjorn]

Dr Toole built the speaker shuffler room at Harman for the purpose of comparative evaluation of loudspeakers by listeners, and for not rooms nor for room treatments. Well, yes, room geometry, furnishings, and acoustic treatments will all greatly affect the quality of the reproduced sounds in the room. If rooms make no difference, people wouldn't be spending substantial amounts of time and money on concert hall acoustic research.

The preparation study Dr Toole did was to show, for the purpose of evaluating relative sound quality of loudspeakers, rooms had no significant confounding influence. Good loudspeakers ranked higher in listening tests than poor loudspeakers regardless of the rooms in which the comparisons were conducted (as long as they were compared in the same room), and therefore one speaker shuffler room was sufficient for the job.

In the same preparation work, Dr Toole also looked at the differences of rooms. Since blind, fast physical switching between rooms wasn't possible, he used binaural recordings in the comparisons. The results were the following, and they showed the effects from the rooms overwhelmed those from the loudspeakers. These results therefore showed that, even with rooms of very different sound quality, when listeners were given the little bit of time to adapt to the room, they could "listen through the room", and were able to rate the sound quality of loudspeakers with high consistency.

View attachment 536646

Thank you. But this implies something else the way I see it. Adaption and sound quality isn't the same.

In my initial post I talked about the weakness of a preference study of speaker directivity. Wide vs narrow, the influence of placement, arrival of side wall reflections and how different types of treatment can influence the result.

You used the argument against this that the room and acoustics basically don't matter. However, that would actually imply that directivity in a room doesn't matter either. The whole study on preferences for directivity would be meaningless if the room didn't matter at all. Can you hear a difference between a narrow broadband speaker and a wide one? Or between an omni and narrow dispersion speaker? Yes, the difference is major.

When the single speaker was tested at Harman, the distance was far enough to side walls to avoid early arriving reflections, and to my knowledge the rear wall in Harman rooms always have absorption. Early arriving reflections from side wall is highly audible and considered detrimental. And absorption for late arriving reflections (rear wall) has always been considered psycoacoustically inferior and gives a different result compared to deep phase grating diffusion. Studies of this were conducted in the development of LEDE and we have other studies on the negative effects of early arrival reflections. In other words, the directivity of the speaker would mean a lot when placed close to the side wall with no treatment.

As for treatment, Toole hasn't done any studies on deep phase grating diffusion vs absorption etc. So we can't dismiss different types of acoustic treatment and the influence on direcitivty preference based on his studies. A deeper phase grating diffuser placed for example on the opposite side wall reflections has a huge impact on spaciousness. The study of spaciousness would be altered with such treatment.

Another weakness in the study are the speakers being tested. A broadband constant and narrow directivity speaker of good quality was never tested in his studies to my knowledge. This type of speaker would largely avoid the nearest side wall reflection, while a speaker ála Revel wouldn't. But the studies haven't been conducted with very close proximity to side wall, despite that most have the speakers placed that way.

 

[Lars Risbo]

thanks @Bjorn, exactly the kind of discussion I was hoping for!

 

[amirm]

Early arriving reflections from side wall is highly audible and considered detrimental.

No it is not. You are repeating a common myth in pro circles and promoted by vendors to sell acoustic products.

And no, it is not "highly audible." But audible in the way it widens the perceived location of the speaker -- a good thing. This is extensively research by Dr. Toole and crew and also by others such as Clark.

 

[olieb]

The whole study on preferences for directivity would be meaningless if the room didn't matter at all. Can you hear a difference between a narrow broadband speaker and a wide one? Or between an omni and narrow dispersion speaker? Yes, the difference is major.

I fully agree to the last sentence.
What I not fully understand is, what you mean by “study on preferences for directivity“. As far as I know there has not been a specific study investigating directivity. And there is no “directivity target“ as a result either. As I see it the main results are “smooth and flat on-axis (no resonances)“ and “smooth of axis sound“.

Then there is the important question in how far the details of the test have an impact on the result on relative preference and what the limitations of the results are. (And Toole did investigate that in respect to many factors.)
However, it is much easier to construct conditions and constellations that the research did not fully account for, than to do the research that takes all these variations into consideration.

So it would be great to have more information on the below (including the limitations of these studies, too).

Studies of this were conducted in the development of LEDE and we have other studies on the negative effects of early arrival reflections.

Before we can arrive at such “results“ as below. (Emphasis by me)

Early arriving reflections from side wall is highly audible and considered detrimental. And absorption for late arriving reflections (rear wall) has always been considered psycoacoustically inferior and gives a different result compared to deep phase grating diffusion.

 

[amirm]

Another weakness in the study are the speakers being tested. A broadband constant and narrow directivity speaker of good quality was never tested in his studies to my knowledge. This type of speaker would largely avoid the nearest side wall reflection, while a speaker ála Revel wouldn't.

An informal blind listening test was conducted between M2 and Salon 2 and the latter won.

 

[amirm]

You used the argument against this that the room and acoustics basically don't matter. However, that would actually imply that directivity in a room doesn't matter either. The whole study on preferences for directivity would be meaningless if the room didn't matter at all.

You misunderstood NTK's post and seem to not be familiar with the research. He said and I repeat, for the purposes of evaluating speaker preference, the room effect is dialed out as long as listeners are allowed to acclimate to it.

This does not at all mean that the room has no effect on what you perceive. Dr. Toole and crew have done a ton of research in this area and huge sections of his book are dedicated to every reflective surface.

Where this nets out currently is that you must get the on-axis frequency response right. And the right sum of off-axis. Once there, then it is an open debate as to preference for wide vs narrow directivity. It is this research that is unfinished. Not that it impacts preference for speakers.