Beyond Linearity: Why Speaker Dispersion Matters Far More Than People Expect

[SSS]

My experience with narrow beam speakers (especially horn systems) and a almost acoustically "dead" room the soundstage is between the speakers and one can hear the recording probably like the master engineer did. But for me this is to some extent not naturally. In a live concert in a normal room there are always sound reflections from side walls, ceiling and rear wall which is natural to me. The opposite is a open-air live concert without reverb and reflections, so this like in the "dead" room. All comes from the front. Although it may not be perfect, I prefer the reflections from may listening room which gives me a more natural feeling.

 

[ahofer]

My experience since learning about spinorama measurements and comparing speakers with measurements: I definitely prefer wider directivity. I had a similar experience to OP auditioning KEFs-not that they were bad at all, but felt more confined compared to a linear speaker with wider directivity. As I just said in another thread, for me Revel>KEF.

 

[chervokas]

At the heart of this thread lies one key question:
What kind of early reflections do people actually prefer?

Here's the thing about preferences, it really doesn't matter and isn't terribly useful datum for any individual know what other people, or the mean, or the majority prefer, because and individual might have a difference preference. For example, most people like chocolate, but if you don't like the taste of chocolate, it doesn't matter in anyway to you or your experience of chocolate that most people like it. C'est la vie. If you're building products to sell -- and Toole worked for Harman -- it's valuable and useful knowledge. But for any one of us, it's kind of useless info.

From where I stand, the evidence points toward early reflections that include high-frequency content being more pleasing. After all, no real-world instrument—or anything our ancestors heard—had narrowly focused dispersion aimed solely at a sweet spot. Environments with suppressed reflections are more akin to dry studio rooms or headphone listening. Some people like that, and that’s totally fine. But for me, it's clear: early reflections with rich frequency content feel more natural and engaging.

I don't think the comparison with musical instruments quite holds up. First, our playback gear are not musical instruments, they are playing back spatial cues and information about the sonic dispersion of the original sound in its original acoustic space. If that information is not on the recording, it doesn't really matter what's happening at playback.

Second, instruments do focus their sound -- think of what a clarinet or a trumpet sounds like if you listen to them right in front of the bell or to the side of the bell. Sound power and harmonic spectrum vary because the instruments don't generate a flat frequency response signal in an omnidirectional pattern.

But, as to the opinion part of this all, I think Floyd Toole may have come up with some useful information for product developers building products that more people will like than not to be use in common area living space type of acoustic situations and doing mostly casual type of listening -- social listening, walking around, the kind of stuff were people are commonly listening off axis.

But I never do that kind of listening. I only listen to music when I can just sit down and do nothing else but focus on the music. I hate headphone and find using them a totally unnatural and cognitively dissonant experience, but I also hate room effects, I find them, together with noise, to be among the most disruptive and distracting -- the opposite of natural and engaging -- things in hifi playback. They seem to always be calling attention to the fact that I'm listening to a canned experience, not a real one, they muck up the frequency response with boundary effects, they much up the ability to hear into the recorded soundstage by layering on their own local acoustic ambient cues that create ambiguity between the recorded "ambience" and the local one.

Like, you might not want to record a sound in a highly damped room. You likely wouldn't get a "natural" sound on the recording, but when you play it back you don't want to add a bunch of additional, cognitively confusing local ambient. Ironically, of course, that's exactly what happens with most popular music. You stick the instrumentalists is a damped, isolated space, you close mic (or you even take the sound from an electronic source direct into the board) so there's little to no ambience in the recording to begin with, maybe you add a bunch of artificial ambience as a special effect during mixing, and then people play it back in their living rooms where they season to taste with local ambience.

I suppose it can be somewhat effective if you're trying to get a "they are here" experience out of playback instead of a "you are there" kind of experience of being transported into the acoustic space of the recording (presuming there is one, and I'm mostly a jazz and classical listener so often there is).

But most home living spaces kindy sound crappy -- they're full of bright flutter echo and hard reflective surfaces in parallel, they have uneven decay times across the frequency spectrum, below 200 Hz response is totally swamped by room modal effects, room dimension are such that you there are all kinds of deep and high boundary interference effects, at high SPLs and low frequencies you have all kinds of things rattling, etc. You can stick audio gear in there that delivers flat frequency response, high spls without distortion, wide dispersion if you want (hell, if you like wide dispersion, go get yourself some omnis, they certainly can produce a seductive "realism" effect) -- but you still wind up to a large degree listening to the room, and the room inevitably is mucking things up.

It may or may not be true that most people have a particular shared preference -- we have 8 billion people in the world, I think Toole's and researchers who have followed haven't quite got to a wide enough, cross cultural enough, big enough set of subjects for use to definitively say that and I think there's enough variety in both use -- such as the music we listen to and the way we listen -- and preference that, per that Evans et al. paper that @Keith_W cited, what "most people" prefer is inconclusive, and what any one of us prefer may different entirely from what someone else prefers.

 

[ahofer]

Here's the thing about preferences, it really doesn't matter and isn't terribly useful datum for any individual know what other people, or the mean, or the majority prefer, because and individual might have a difference preference. For example, most people like chocolate, but if you don't like the taste of chocolate, it doesn't matter in anyway to you or your experience of chocolate that most people like it. C'est la vie. If you're building products to sell -- and Toole worked for Harman -- it's valuable and useful knowledge. But for any one of us, it's kind of useless info.



I don't think the comparison with musical instruments quite holds up. First, our playback gear are not musical instruments, they are playing back spatial cues and information about the sonic dispersion of the original sound in its original acoustic space. If that information is not on the recording, it doesn't really matter what's happening at playback.

Second, instruments do focus their sound -- think of what a clarinet or a trumpet sounds like if you listen to them right in front of the bell or to the side of the bell. Sound power and harmonic spectrum vary because the instruments don't generate a flat frequency response signal in an omnidirectional pattern.

But, as to the opinion part of this all, I think Floyd Toole may have come up with some useful information for product developers building products that more people will like than not to be use in common area living space type of acoustic situations and doing mostly casual type of listening -- social listening, walking around, the kind of stuff were people are commonly listening off axis.

But I never do that kind of listening. I only listen to music when I can just sit down and do nothing else but focus on the music. I hate headphone and find using them a totally unnatural and cognitively dissonant experience, but I also hate room effects, I find them, together with noise, to be among the most disruptive and distracting -- the opposite of natural and engaging -- things in hifi playback. They seem to always be calling attention to the fact that I'm listening to a canned experience, not a real one, they muck up the frequency response with boundary effects, they much up the ability to hear into the recorded soundstage by layering on their own local acoustic ambient cues that create ambiguity between the recorded "ambience" and the local one.

Like, you might not want to record a sound in a highly damped room. You likely wouldn't get a "natural" sound on the recording, but when you play it back you don't want to add a bunch of additional, cognitively confusing local ambient. Ironically, of course, that's exactly what happens with most popular music. You stick the instrumentalists is a damped, isolated space, you close mic (or you even take the sound from an electronic source direct into the board) so there's little to no ambience in the recording to begin with, maybe you add a bunch of artificial ambience as a special effect during mixing, and then people play it back in their living rooms where they season to taste with local ambience.

I suppose it can be somewhat effective if you're trying to get a "they are here" experience out of playback instead of a "you are there" kind of experience of being transported into the acoustic space of the recording (presuming there is one, and I'm mostly a jazz and classical listener so often there is).

But most home living spaces kindy sound crappy -- they're full of bright flutter echo and hard reflective surfaces in parallel, they have uneven decay times across the frequency spectrum, below 200 Hz response is totally swamped by room modal effects, room dimension are such that you there are all kinds of deep and high boundary interference effects, at high SPLs and low frequencies you have all kinds of things rattling, etc. You can stick audio gear in there that delivers flat frequency response, high spls without distortion, wide dispersion if you want (hell, if you like wide dispersion, go get yourself some omnis, they certainly can produce a seductive "realism" effect) -- but you still wind up to a large degree listening to the room, and the room inevitably is mucking things up.

It may or may not be true that most people have a particular shared preference -- we have 8 billion people in the world, I think Toole's and researchers who have followed haven't quite got to a wide enough, cross cultural enough, big enough set of subjects for use to definitively say that and I think there's enough variety in both use -- such as the music we listen to and the way we listen -- and preference that, per that Evans et al. paper that @Keith_W cited, what "most people" prefer is inconclusive, and what any one of us prefer may different entirely from what someone else prefers.

I think that’s about right, and why you should still listen to speakers before buying/committing. However, the progress we’ve made in measurement now provides an *explanation* of my preference that, once noted, the consumer can use to narrow down their shopping time. Everyone looking at speakers should try a linear-wide and linear-narrow speaker in their own space. I do think that the *evenness* of that dispersion is a more universal preference. I’d rather listen to KEFs (narrow) than Magnepans (wide and uneven), and I’ve owned both. One possible exception is the “Gundry/BBC dip” that is built into, say, Harbeths, where the dispersion dips in the presence region (not the on-axis response). But that’s another example where measurements are illuminating preferences rather than displacing them, as some would aver.

 

[bmc0]

If you're building products to sell -- and Toole worked for Harman
[...]
I think Floyd Toole may have come up with some useful information for product developers building products that more people will like than not to be use in common area living space type of acoustic situations and doing mostly casual type of listening -- social listening, walking around, the kind of stuff were people are commonly listening off axis.

A good portion of Toole's work was done at the National Research Council in Canada, funded by Canadian taxpayers. It is incorrect (and insulting, in my opinion) to imply that his work is biased by corporate interests cynically seeking to sell more product. If this is not what you meant, I apologize, but some people are likely to interpret your statements this way.
"Casual listening" does not describe the conditions of the listening tests conducted by Toole and his collegues.

That said, my own preferences don't seem to align perfectly with Toole's conclusions (clearly yours don't either). My main speakers are fairly narrow directivity (±45°; DI of ~9dB), which helps to suppress very early reflections, but I also don't think that a listening space nearly devoid of reflected enegery is desirable for 2-channel playback. Standard 2-channel is fundamentally flawed in several ways and some reflections help to mitigate these problems. Even better, additional playback channels—either real or sythesized from 2-channel source material—can offer a significant improvent in spatial quality and general realism if done properly. My preference is firmly in the "you are there" camp for virtually all music.

 

[Keith_W]

A good portion of Toole's work was done at the National Research Council in Canada, funded by Canadian taxpayers. It is incorrect (and insulting, in my opinion) to imply that his work is biased by corporate interests cynically seeking to sell more product. If this is not what you meant, I apologize, but some people are likely to interpret your statements this way.

He may not mean that, but I would certainly make that imputation. All medical publications make authors disclose any conflicts of interest, e.g. studies funded by drug companies and so on. Then there are authors whose personal biases make for some pretty interesting publications. e.g. there was an ICU doctor who loved his dialysis machines, and published study after study asking if dialysis machines can improve outcomes in ICU patients even for those who did not need dialysis (for example, "can dialysis machines remove inflammatory prostaglandins and reduce severity of systemic inflammatory response syndrome?"). Then there was another guy who thought that low dose antidepressants should be in our drinking water. Constant exposure to these academic nutjobs in my professional life makes me read every publication with a healthy dose of scepticism and critical thinking.

I treat audio studies the same. I assume there is some kind of conflict of interest and read the study accordingly. It is not my intention to impugn Toole and call him a corporate lackey, but reading studies with this mindset helps you keep alert as to whether author bias influenced the study. The data is the data, and provided the experiment was conducted properly and the results reported honestly, the data does not lie. But interpretation of the data is something else - some authors might overstate the problem, or overstate the significance of the outcome - like make generalisations that they should not. This is very common in the academic world.

For example, can we use Toole's 1985 study to argue that wider directivity loudspeakers are preferred in general? I would argue that we can not. The sample size was too small (3 loudspeakers, 10 listeners), and the experiment was flawed (all the tested loudspeakers had issues). It needs corroboration with other studies. To me, it's a data point and not a definitive experiment. I don't think Toole would argue otherwise.

 

[bmc0]

I would certainly make that imputation.

I am not making the claim that one should uncritically accept Toole's interpretations, nor that he is immune to bias and has no conflicts of interest. The other poster appeared to be asserting that his studies were designed and conducted with the intent to inform the development of mass-market products aimed at uncritical listeners. As far as I can tell, this is not correct.

Absence of reproducibility trials by independent parties seems to be general problem in this particular field.

 

[Yasuo]

Here's my point: chasing absolute linearity in speakers is like obsessing over SINAD in amplifiers—past a certain threshold, it becomes meaningless.

This applies to (probably) anything. It's the point of diminishing returns. My personal rule is to buy the best ratio of cost/performance. Sure, a few times I bought high-end models (flagship?), but fortunately not when it comes to audio equipment.

 

[chervokas]

A good portion of Toole's work was done at the National Research Council in Canada, funded by Canadian taxpayers. It is incorrect (and insulting, in my opinion) to imply that his work is biased by corporate interests cynically seeking to sell more product. If this is not what you meant, I apologize, but some people are likely to interpret your statements this way.
"Casual listening" does not describe the conditions of the listening tests conducted by Toole and his collegues.

That said, my own preferences don't seem to align perfectly with Toole's conclusions (clearly yours don't either). My main speakers are fairly narrow directivity (±45°; DI of ~9dB), which helps to suppress very early reflections, but I also don't think that a listening space nearly devoid of reflected enegery is desirable for 2-channel playback. Standard 2-channel is fundamentally flawed in several ways and some reflections help to mitigate these problems. Even better, additional playback channels—either real or sythesized from 2-channel source material—can offer a significant improvent in spatial quality and general realism if done properly. My preference is firmly in the "you are there" camp for virtually all music.

The point is not that there's anything wrong or bad or biased in Toole's research, whether it was done in the decades before or the decades after he went to work for Harman (though I think his well known book, Sound Reproduction, was published long after he went to work for Harman IIRC) -- the point is that the matter of "what kind of speaker do most people prefer?" is basically a commercially useful question, but not an individually useful one.

If I want to build a product that is going to have the widest potential marketplace acceptance, building something that conforms to the research on what most people prefer will be helpful and useful and a risk mitigating factor to me. If I'm an end-user determining what I prefer, knowing what most other people like is meaningless to me. Knowing that all of my neighbors like chocolate isn't going to be useful at all in determining whether or not I like chocolate. I need not concern me, it's not helpful or useful to me.

Also, when we look around and see what people DO prefer -- including apparently both you and me -- it often doesn't conform the "use the early reflections" approach. I think, in fact, when we look around and see what people do choose to listen to -- from beamy horns, to dipoles or even omnis, in live rooms and in treated rooms -- there's a really wide variety of preferences, reflecting, I suspect, not only a wide variation among individual likes and dislikes but also a wide range of differences in preferred program material and types of listening. And I'm not sure that preferences studies of speakers have cast a wide enough net in terms of test subjects and real world program variety, nor a large enough sample size, to capture the actual variation of preference we see around us in the world day to day.

 

[tmuikku]

Hi,
it's about "how to listen". Can you hear singular early reflection, what does it sound like? If you turn on your stereo right now, do you localize the reflection and "hear" how it sounds like? How does it affect the whole perception of sound?

I mean there is single perception of sound at any given moment in a small room, we don't hear distinct echo as the walls are really too close to hear distinct echoes. So, as yourself what part of the "one perception" you are currently having is due to a reflection? What in the perceived sound is due to reflections?

Go ahead and figure out a listening experiment to subject yourself to reflected sound and you should be become bit more sensitive for reflections for few minutes at least, not too hard. Just hold your mobile on your hand with some makeshift horn play some white or pink noise and point at things, listen the reflection you are making. Experiment with yourself like this, by actively changing sound and actively listening how the sound changes!

Now that you've subjected yourself how reflected sound sounds like, you should be able to hear it with your stereo as well. Try this, using one speaker only, ask your friend move a mattress around at first specular reflection point, or ask them to turn the speaker if it's a directional model and try to listen how the reflection changes and what happens to the sound in general. If you are alone you could go behind a speaker and point at various walls in your room, can you hear the room sound change? Can you change audibility of a reflection with turning your head or moving your position, like getting closer to the wall than speaker, or vice versa? At what distance you can't hear it any more, or can you? What you heard from behind the speaker, does it still hear if you now walk to your listening spot or did it go away? What if you turn your head? what if you move closer to speakers iow increase direct / reflected sound ratio among changing angles toward the specular reflection points, how does the sound change or does it?

There is literally million tests to subject yourself to how your room and stereo sound like and how your auditory system works with it, I mean some reflection might be obvious while testing but if you rotate your head slightly it might go away and so on. Do this as long as needed to become acute to the sound so you can position the speakers as you like, to what sound you prefer there in your room. This is all done by ear like so, one doesn't have to know any details of anything, just experiment and do what sounds nice to you, or interesting, what ever. I bet, that in the process you'll find out thing or two what you wanna hear, assuming your speakers can do that. It could be possible they don't and you never get anything from this experiment. But, if there is even a glimpse, you'd know what kind of speaker (directivity) would be better!

Have fun experimenting everyone

 

[Matias]

Wide dispersion: sounds wider (duh), more "surround sound" from 2 speakers, immersive. Can be problematic if a room is asymmetric as it will be highlighted.

Narrow dispersion: soundstage is less immersive/surround but more focused and defined, more room friendly too.

IMO there is no right or wrong here, just different types of presentations with pros and cons and our personal preferences.

 

[Heinrich]

... according to Toole’s research, wide-dispersion speakers generally outperform narrow or controlled-dispersion designs in terms of perceived sound quality.

That can't be said without mentioning, that the statement refers to the average of some population. It is not said whether there is a greater or lesser preference for one or the other, or whether individual judgments favor either one or the other entirely. Just as an example of how quickly science can be distorted.

What kind of early reflections do people actually prefer?

That may sound a bit strained, but why are you interested? Are you building your stereo for random people you can't ask?

And why should it be the early reflections, but not the diffuse field as a whole, the 'steady state'?

Now you already own two fundamentally different systems at some cost. The change also takes effort, I suppose. How would you make the discussed property adjustable?

Here's the thing about preferences, it really doesn't matter ...

Second, instruments do focus their sound -- ...

Exactly, it came to me likewise many times. Same with the instrument's dispersion pattern. How would it be simulated in a stereo context, havin 'virtual' sound sources. And so forth ...

 

[MattHooper]

My experience….

I’ve never liked head-in-a-vice listening experiences. I don’t like noticing obvious imaging or tonal shifts with even minor shifting of my head or position on my sofa.

So that is one reason for me to favour that speakers that have even off axis performance and some nice dispersion.

And while I really do like imaging precision, I like it more in the sense of hearing really dense, palpable images. (That’s one reason I’ve been attracted to Thiel speakers for so long, which to my ears have those qualities).

However, I don’t like heavy toe-in. I find this often results in Sonic images that seem too crisply defined at the edges, and often a more washed out tonal balance.

So I have almost always had some toe out of my speakers, occasionally facing straight ahead.

Just as many others experience in a typical room, this widens the soundstage, and I find images tend to float more freely in the room.
As well, Sonic images expand a bit more, edges soften a bit, and so I find this to sound more pleasing and natural than the “ squeezed too tight” images with heavy toe-in.

It’s always a trick to get just the right amount of slight toe out which allows the sound to relax and open up, without sacrificing too much imaging precision, or Sonic density.

But I always find my sweet spot for this.

Then there’s the other aspects of the room interaction.

I have a lot of control over reflectivity in my room.

The more I engage room reflections, the more lively and “live” the presentation. Which can be quite gratifying.

On the other hand, the more I control reflectivity, the more subtleties I hear in the recording in terms of timbral nuance and the exact and often subtle nature of the recorded ambiance or added reverbs.

The thing is I really really like that timbral nuance (it’s one reason I bought my speakers) and I love hearing the exact ambience/reverb in different recordings.
I’m sort of a “ reverb fan” in a way.
So I don’t want some totally generic “airy presentation” but rather I want to hear the really specific character that changes from recording to recording.

And yet again, I want that sense of relaxed wall-melting presentation when available from a recording.

So for me, it’s always been a careful dialling in my loudspeaker positioning, listening position, and playing with opening up or cutting down reflectivity, to get a balance in which I hear into the delicate nuances of each recording, which doesn’t become too “ headphone” and doesn’t lose that “ life and energy.”

So my ideal isn’t, as some ultra neutral narrow dispersion systems present the music - which is like a portal that opens up only in between the speakers, and so it feels like a separate portal into the recording from the room I’m in.

Rather, I want to engage just enough reflection so the room beyond my loudspeakers - including out to the sides of my speakers - shape shifts with the nature of each recording.

And that is what I have got, I feel.

I’ve mentioned before that my favourite new hack for acoustics has being the introduction of a curved diffuser that I place in between my loudspeakers near the back wall.

This particular diffuser adds a bit of lively reflectivity. What allows me to do is cut off more of the room reflections so I really get the most subtle nuances of the recording and enter into that recording space. But the diffusor adds some more life to the Sonic images within that space - the imaging becomes more dense and palpable without being squeezed tight, and also seems to add a little bit more upper mid range or lower trouble, energy.

For me, this is made for the best of both worlds. I am really experiencing The ambiance of the track and not too much of my room, and yet sonic images have that dense, lively character I crave.

 

[sigbergaudio]

But after living with the KEFs for a year, I noticed something fundamental. Coaxial speakers like the KEFs have limited spatial imaging, especially in smaller rooms. They sound focused and clean, almost like listening directly to near-field studio monitors. Minimal early reflections, yes, but the room never felt truly filled.

Just came here to say that you can't generalize your experience with this KEF speaker to every speaker with a coaxial driver. The widest soundstage and the most immersive sound I've ever heard has come from coaxial speakers with moderate dispersion width.

 

[ahofer]

How would it be simulated in a stereo context, havin 'virtual' sound sources. And so forth ...

But the recording captures that…I may be misunderstanding but it seems you are mixing up the source and the reproduction.

 

[thewas]

Just came here to say that you can't generalize your experience with this KEF speaker to every speaker with a coaxial driver.

As I had written above someone cannot even generalize that experience with KEF loudspeakers to different rooms and personal preferences.

 

[chervokas]

How would it be simulated in a stereo context, havin 'virtual' sound sources. And so forth ...

Various stereo micing techiniques, like XY and mid-size configurations can be effective in capturing a the sound of instruments in space in a sound field reproduceable in two-channel stereo playback, spaced omnis can work too, just that you can capture the actual mix of focused and disbursed sound form the original instruments in the original space. More complicated micing patterns can be used too, but they can often result in somewhat confusing or ambiguous soundstage cues for the listener. For headphone listening, some of these dummy head type binaural recordings can be really effective.

 

[anphex]

What I can definitely say after some time of active testing is that yes, the KEF Reference 3 present music more precisely, more intimately, and render the reverb/room ambiance set in the mix very clearly—making it feel like you're "live" in the same recording studio.

The NuVero 70 lack this finesse, as they strongly excite my own room, causing my early reflections to mix heavily with the space embedded in the music material.

In the end, it comes down to this:
Both speakers are fun, but in their own way. The KEFs are like a magnifying glass into the music—very detailed and clean, though sometimes a bit dry. The Nuberts make every song enjoyable and bring the music live into my room. Almost like a somewhat reversed effect compared to the References. Of course, they sacrifice some detail and linearity for that.

Interestingly, this effect persists even two rooms away. Even there, these core differences between KEF and Nubert remain noticeable. So it really seems to be a "battle of dispersion patterns", since both speakers are relatively linear at 10–30° off-axis.

I'll provide pictures later today. My living room setup is kind of funny lately. If interested, I can provide measurements of both speakers on and off axis in my living room on my listening position.

 

[ahofer]

“You are there”/“they are here”

 

[Heinrich]

What I can definitely say after some time of active testing is that yes, the KEF Reference 3 present music more precisely, more intimately, and render the reverb/room ambiance set in the mix very clearly—making it feel like you're "live" in the same recording studio.

The NuVero 70 lack this finesse, as they strongly excite my own room, causing my early reflections to mix heavily with the space embedded in the music material.

It would be interesting to know, which characteristic in particular drives the impression. Is it the diffuse field, the steady state of reverberation, or is it mostly the first reflections.
The main difference between these is that first reflections interfere with the direct sound, leading to deep dips and relatively mild peaks, even comb filter effects. The steady state interferes in more chaotic fashion plus minus something.

Presumably first reflecions generate directional cues that may be decoded by the listener. But those for sure won't enhance the stereo effect, but would to the contrary reveal the speaker's position in the room. I wonder how such an effect could be acvided, if it is real. Dr Toole speaks of 'listening through a room's acoustic'?

Dr. Toole also speaks of the virtues of a not too subdued steady state. For instance it fills up the lobing effects around the crossover frequency. And additionally, it may fill up dips, equalize peaks originating in early reflections?

In this context it may be interesting, that sound is identified mostly by the first attack, some 10ms plus. The spectrum is more continuous, not so much an assembly of 'notes' and its overtones. That mitigates the impact of said peaks to some degree, I speculate.