CD² (Constant Directivity and Critical Distance)

[Y~BNA]

HI EVERYONE, MAJOR EMERGENCY EDITING ALERT!

I posted an opening statement on this thread that was hastily and sloppily phrased.
This caused it to contain inaccuracies and even false information and to be prone to misinterpretation.
I am like that. Jumping in. Would usually do better when i wait a bit and let it settle.
Not too smart to begin with (in certain respects anyway), jumping in with sloppy posts makes me look even dumber than i already am!

SO HERE'S MY REVISED AND EDITED OPENING STATEMENT FOR THE THREAD ('FORUM'). HOPE I DID A LITTLE BETTER THIS TIME.

2CD is key in the search for reproducing a credible illusion of live recordings (from concert hall or music studio) in our home listening environment.

CD simultaneously happens to be the acronym of two parameters highly relevant to attaining this goal:
-Constant Directivity (ideally closely approached)
-Critical Distance (ideally far, to create a wide listening area)

Both very important and difficult to manage well in small(ish) reproduction listening environments, both critical and, on top of that, the second depends on good implementation of the first.

Both Open Baffle configurations and specific "Living Room Line Arrays" (among other speaker concepts) have shown to be a step, or several, in the right direction when implemented well. Since a while actually, as a very famous and specific type of line array (Don Keele's CBTs of the 1970's) and very famous dipoles (Quad ESL63s and Siggy's {RIP} Orions/LX521s of the 1960's and 1980/1990's) remind us.

Quoted years may not be entirely correct, but the concepts have been around for at least half a century. Yet (almost) no manufacturer has made serious attempts at continued and sustained mass production for home audio, for obvious marketing and other practical reasons.

That means little funding has gone into the research also, leaving us in the hands of some great contributors in the area between Pro and DIY audio, among which the aforementioned greats. There's some on ASR, some on DIYaudio, many AES or BAS members, some have their own websites, some are professionally involved but share their knowledge unreservedly and their white papers are freely accessible. .

I'd like to participate in this research, accompanied by folks some of whom are better educated and more experienced than myself, hopefully.
My own research focuses on other's research. Meta- or comparative research so to speak, so mostly theoretical so far. I invite anyone to shoot me down at any time. If you have valid arguments, knowledge or experience: put me in my place. But i found comparative research can be illuminating and very satisfying.

I lack a formal electronics or acoustics education, but i am an engineer, tend to grasp logics (and a tiny bit of logic) and am generally able to adequately absorb basic concepts discussed in white papers slightly outside my field (by academics, engineers or otherwise). Of which there is still an abundant treasure to be found about Open Baffles and Line Arrays on www and in print.

Standing on the shoulders of giants i hope to take some small steps i could never accomplish on my own, and then do a test build and give back measured results.

If any of the above resonates with you, take your shot.


For those not familiar with the concepts:

1.
Constant Directivity.

The most used term for Equal Power Response, Uniform Radiation Pattern and some similar terms, all meaning the same:

For our auditory senses (two ears + psycho-acoustic brain centre + what else involved?) to recognise an auditory event to be genuine, as in "i am hearing musicians, not a loudspeaker-cone in a coffin", the intensity of the sound should radiate more or less evenly in all directions, even maintaining its spectral balance when reflected. This incidentally means off-axis listening will get better: although the SPL will lower, the spectral content will remain more or less unchanged (best case scenario).

Given the physical limitations of acoustic transducers and their applications, as well as the acoustics of most rooms, only very few loudspeaker-listening room combinations attain this goal.

ALL well designed audio devices sound decidedly identical, given the enormous differences between ANY two makes of loudspeaker in any two rooms, OF ANY QUALITY LEVEL, AT ANY PRICE.

That should tell us something.


2.
Critical Distance.

The distance away from the sound source (speaker) past which reflected SPL is higher than direct SPL.
Also known as the "Near Field/Far Field" transition. They signify the same:

When reflected sound overwhelmes direct sound, articulation becomes blurred (specifically on transients/impulses), intelligibility of speech is compromised and eventually the whole sound stage drowns in a sea of undefined acoustic rubble.

In a very reflective room this distance and the near field are small, in a 100% anechoic room this distance and the near field could (theoretically) cover the entire room.

Whether that makes an anechoic room the ideal listening venue, however, is under debate IMO and will be addressed later (in this thread or elsewhere).
Cheers!

 

[kemmler3D]

Not an expert that's going to help advance the cause here, but I have a couple questions for my own edification -

1)

both are also critical above all else and, on top of that, one can't be good without the other.

Why is this?

2) Why not constant directivity horns, vs. OB / line arrays?

I'm not suggesting your approach is wrong but I don't know the arguments in favor of your assertions off the top of my head either.

 

[Y~BNA]

For those not familiar with the concepts:

1.
Constant Directivity.

The most used term for Equal Power Response, Uniform Radiation Pattern and some similar terms, all meaning the same:

For our auditory senses (two ears + psycho-acoustic brain centre) to recognise an auditory event as being genuine, as in "i am hearing musicians, not a loudspeaker-cone in a coffin", the intensity of the sound should be more or less at an even level for all frequencies in all directions at all distances.

Given the physical limitations of acoustic transducers and their applications, as well as the acoustics of most rooms, extremely few loudspeaker-listening room combinations attain this goal.

ALL well designed audio devices sound decidedly identical, given the enormous differences between ANY two makes of loudspeaker in any two rooms, OF ANY QUALITY LEVEL, AT ANY PRICE.

That should tell you something.

2.
Critical Distance.

The distance away from the sound source (speaker) where reflected soundlevels start matching direct soundlevels.
Also known as the division "Near Field/Far Field". They signify the same:

When reflected sound overwhelmes direct sound, articulation becomes blurred (specifically on transients/impulses), intelligibility of speech is compromised and eventually the whole sound stage drowns in a sea of undefined acoustic rubble.

In a very reflective room this distance and the near field are small, in a 100% anechoic room this distance and the near field (theoretically) cover the entire room.

Whether that makes an anechoic room the ideal listening venue, however, is under debate IMO, and will be addressed later (in this thread if it is approved, or elsewhere).

 

[Y~BNA]

So just in case (we can always copy/paste posts to another thread i hope?), i will elaborate a lil bit about what i'm trying to do here.

There's indeed some Open Baffle Line Arrays on this site, Siggy's LX521s are of course mentioned, built, emulated, etc. Don Keele's curved CBTs also (they're a separate kind of animal, but still undeniably Line Arrays, complete with temporal shading by the carpentry and amplitude shading by passive PCBs), but the combination of both principles is scarcely executed and rarely described, as well as poorly supported by research and testing.

Perry Marshal has some nice examples on this site (https://www.diyaudio.com/community/search/2380719/?q=dipoles&c[users]=perrymarshall&o=date) and there's more to be found, but by the time i arrived at this concept myself, it was already clear that this is a search!

It is, however, a viable way to expand the near field for HF (something point source OB has trouble with) and expand Constant Directivity into the upper three octaves, something OB on its own has trouble with as well.

I have run into planar tweeters that have been found testing outstandingly by Mike Chua (ampslab.com). They are usable down to 400Hz, meaning they could potentially be crossed at 800Hz, and with some luck and very low load (as in LA) even a bit lower. Incidentally, this relieves us of the need to build a four-way, as this 800Hz-18/20kHz range actually covers two traditional ways.

They are shaped like true line sources, with a grid featuring small openings (6mm wide) stacked vertically. These are 1/3 of the wavelength at 20kHz, loaded by the much larger diaphragm behind it, meaning they will tend to behave somewhat like independent 180°/half omni directional 6mm sources (horizontally!).

Stacked in an array of 4-6 units they will form a true line source that will be narrower than the wavelength way past 20kHz. At 80-120cm height, the gaps between units are the size of 7,150Hz waves (48mm), so they will start to slightly misbehave only in the upper octave (where actual fundamentals are virtually absent).

All this makes for a line array that is (in theory) ’half-omni' without any significant lobing, from its XO to 20k. It is a dipole by design, enabling figure-of-eight radiation pattern into frequencies point source OB can't reach, even if it's not on account of the actual dipole null (fortunately, because that would have turned into comb filtering at these frequencies).

Plus its LA nature compensates for the early HF fall-off through air absorption. The level, as well as the cylindrical wave front, if all parameters are strictly met, may in this particular line array hold up longer (to around 14m/47ft) according to first calculations. If it's half that in practice, this will already be a huge step in acoustics.

So combining said line array with a 12" dipole midwoofer (effective piston diameter 265mm = 1/2 wavelength of 650Hz) and a pair of 18" dipole subs, we'll possibly attain a system where the bottom four octaves are spherical figure-of-eight, the next five up to 18kHz are mostly upholding cylindric figure-of-eight. Only vertically, the non fundamental upper 10-20kHz octave will probably slightly misbehave.

All with reduced room interaction, especially in the lower and upper voice (between 300Hz and 4kHz) where the dipole and line array characteristics will reinforce each other.

I am hoping for further reduced room interaction than based on dipole alone, which will improve both direct to reflected ratio and late to early reflection ratio. Resulting in a room that will consist of 80% near field.

So far this build plan is completely based on acoustic design, no DSP filter to be found anywhere. The woofers will be crossed way below their dipole peak, the mids also, and all drivers operate below their beaming range, based solely on their physical dimensions. Also acoustic centers will be time aligned physically, since delay in a symmetric OB doubles at the rear the exact problem you are fixing at the front.

Only NOW will we use a little DSP for flat tuning FR and driver anomalies per way, and then FIR filtering some nice XOs.

We no longer have to sit still to enjoy a sweet spot, we'll all fit inside it while walking around. If only...

OK, those were the basics, all evolved from wishful thinking, now shoot me down all o' yous!

 

[amirm]

For our auditory senses (two ears + psycho-acoustic brain centre) to recognise an auditory event as being genuine, as in "i am hearing musicians, not a loudspeaker-cone in a coffin", the intensity of the sound should be more or less at an even level for all frequencies in all directions at all distances.

Where did this assumption come from? And how do you know it is even achievable?

 

[Y~BNA]

Not an expert that's going to help advance the cause here, but I have a couple questions for my own edification -

1)

Why is this?

2) Why not constant directivity horns, vs. OB / line arrays?

I'm not suggesting your approach is wrong but I don't know the arguments in favor of your assertions off the top of my head either.

1).
Constant Directivity is what defines the soundstage in a room, if your speakers and your room together don't produce it, the illusion of a musical event can not be accomplished.
If it is accomplished a little bit, like at a small sweet spot 7ft in front of your speaker pair, your Critical Distance is small. That means the area where direct sound dominates is small. Most of the room will not have a credible soundstage, because reflections dominate.
So without Constant Directivity your Critical Distance will be shit. And without good nearfield behavior your Constant Directivity will be shit.

2)
I don't know. What are you asking. What do you mean "versus"? There is no versus. Any means to reach both CD's is valid. As long as you reach 'em.
Although i would think horns are inherently problematic, as they are very directional, and their directionality is not constant between ways. I might be wrong, but horns main attraction is sensitivity AFAIK.
Are they known for an Equal Power Response? A Uniform Radiation Pattern regardless of frequency, direction and distance? If so, they suit the purpose of CD².

 

[sqrl]

For our auditory senses (two ears + psycho-acoustic brain centre1) to recognise an auditory event as being genuine, as in "i am hearing musicians, not a loudspeaker-cone in a coffin"2, the intensity of the sound should be more or less at an even level for all frequencies in all directions at all distances3.

1. Auditory processing is distributed and integrated with other sensory modalities.
2. I see no musicians.
3. Inverse square law.

 

[Y~BNA]

Where did this assumption come from? And how do you know it is even achievable?

Isn't Constant Directivity the holy grail of speaker building? Read Toole, Keele, Vanderkooy, Beranek, Angus, Linkwitz, etc. etc. I could go on for quite a while really. Anyone who has published on loudspeakers in rooms basically.

Some speakers achieve it to quite a satisfying extent, but the ideal speaker-room combination is still largely theoretical, as demonstrated by the fact that ALL well designed audio gear is interchangeable (can not or hardly be distinguished between in double blind listening tests) except loudspeakers and rooms.

Even among the best of those, no two sound the same.
But there ARE those that immediately captivate the listener by creating a credible illusion of live music. So that comes pretty close.

I have never heard a boxed speaker do that, but i have heard dipoles do that. And they often don't really care if the room is perfect, as long as it is not too out of the ordinary.

Other than that i truly don't know if it is achievable, didn't say it is, but i'm sure gonna have fun doing the research and trying to approach it.

Tx & cheers

 

[Y~BNA]

For our auditory senses (two ears + psycho-acoustic brain centre1) to recognise an auditory event as being genuine, as in "i am hearing musicians, not a loudspeaker-cone in a coffin"2, the intensity of the sound should be more or less at an even level for all frequencies in all directions at all distances3.

1. Auditory processing is distributed and integrated with other sensory modalities.
2. I see no musicians.
3. Inverse square law.

Yes, and? Our auditory senses cooperate with other mental and physical functions, so? They work, generally. Do we need to see musicians to hear them? Yes the sound is reduced by the increase of distance squared. But with a good speaker design this effect can still be as directionally balanced as it is at a live event? What's your point exactly? Hamburgers? Sunset?

 

[Curvature]

Critical distance is when direct sound is at the same level as reverberant sound. It is not relevant, as a concept, to small rooms other than in the high frequencies, where something like a diffuse field forms. And even then any point within the main polar pattern will be inside the critical distance. This is assuming a standard listening situation. Human sitting in front of speakers.

The idea is only useful for acoustically large rooms, like concert halls, as defined by wavelength. It's the point beyond which localization and intelligibility suffer.

If you want to hear what happens when you exceed the critical distance at home, in an unconventional way, listen to your system from another room.

 

[sqrl]

Yes, and? Our auditory senses cooperate with other mental and physical functions, so? They work, generally. Do we need to see musicians to hear them? Yes the sound is reduced by the increase of distance squared. But with a good speaker design this effect can still be as directionally balanced as it is at a live event? What's your point exactly? Hamburgers? Sunset?

Word salad.

 

[Arindal]

Critical distance is when direct sound is at the same level as reverberant sound. It is not relevant, as a concept, to small rooms other than in the high frequencies, where something like a diffuse field forms.

It is correct that this definition of critical distance is mainly relevant for microphone placement and recording. For reproduction, there is a different form of indirect soundfield at play, as reverb is both contained in the recording, ad (accidentally) created in the listening room, both with contradicting reflectograms, i.e. time pattern of reflections. So I would agree that the mathematical definition of critical distance is not applicable to stereo reproduction, not even at high frequencies. If the reverb field is mathematically diffuse or not, is rather irrelevant, compared to the question if it is sufficiently subtle and ´hiding behind the original reverb´.

The interaction between directivity, listening distance, reflections and reverb in the listening room is of critical importance, though, and I would as a theory support the concept of following both constant directivity and balanced tonality of the room, plus defining critical distance for reproduction.

My personal experience says that the latter can only be defined with practical experiments. I suggest to define practical critical listening distance as the listening distance between speakers and listener that if not critically altering/deteriorating

- localization precision
- localization stability
- reverb tonality
- ambience perception in terms of room size and propiertes
- depth of field

compared to corresponding ideal conditions, which can with most speakers be easily achieved by moving to nearfield conditions.

The idea is only useful for acoustically large rooms, like concert halls, as defined by wavelength. It's the point beyond which localization and intelligibility suffer.

That is a definition you might find in older books. In practice, localization and intelligilibty are vastly independent from the critical listening distance defined by same level of direct and indirect sound, if only the reflection and reverb pattern does serve intelligibility and localization. In the latter case, reverb can even be dominant in level but both properties are retained.

For those who do not believe me, I recommend to visit the following opera/concert venues and enjoy an acoustic performance in there (anyways highly recommended):

Both have surprisingly high RT60 (2.4s for the concert hall, or the opera house approx. 2s), but they do offer a level of transparency, detail resolution and intelligibility which is unrivaled in the world to my knowledge. This is the result of very clever concepts of diffusing reverb.

If you want to hear what happens when you exceed the critical distance at home, in an unconventional way, listen to your system from another room.

The problem with this experiment is that not only the ratio between direct sound and reverb is affected, but even more the tonal balance of the reverb and the direct sound alike. So it will sound very different from an example where the direct sound stays the same, but the reverb gets louder with increasing distance between source and listener.

 

[Arindal]

Why is this?

If the general aim is an indirect soundfield in the listening room which is diffuse, tonally balanced and sufficiently attenuated at the listening position at the same time, it is pretty obvious this goal is jeopardized by either non-constant directivity or overly low directivity, so both parameters must be ideally met at the same time.

Imagine the two extremes of a loudspeaker concept:

omnidirectional - ideally constant directivity, but way too low, leading to dominant indirect sound in almost every room
steeply increasing directivity as with some large horns, planar or fullrange speakers - sufficient suppression of indirect sound at higher frequencies, but dominant midrange reverb which will cause it to stand out from the reverb on the recording easily

Why not constant directivity horns, vs. OB / line arrays?

An ideal constant directivity horn should be doing the job here as well, probably better than an open baffle/dipole concept. Yet, they are more difficult to design, if you really want a constant directivity, and would not accept a widening midrange directivity and overly narrowing one in the upper treble region. Most horns labelled as ´constant directivity´, do not actually meet these conditions, and those which do are pretty big.

There are other concepts imaginable to achieve these goals, like cardioids, arrays, curved planar transducers. And of course we can think of any combination of the aforementioned, as some concepts are easier to design for midrange or treble, some for the bass.

To my knowledge, GGNTKT M3 is the existing concept which comes closest to constant directivity, and it is actually a combination of several.

have run into planar tweeters that have been found testing outstandingly by Mike Chua (ampslab.com). They are usable down to 400Hz, meaning they could potentially be crossed at 800Hz, and with some luck and very low load (as in LA) even a bit lower. Incidentally, this relieves us of the need to build a four-way, as this 800Hz-18/20kHz range actually covers two traditional ways.

I am not aware of a planar tweeter delivering anything close to a constant directivity from 800Hz to 18K. Which one are you referring to?

Any truncated line source known to me shows a significant increase in directivity index either this or that way. To use such for a constant directivity concept, you have to severely limit the frequency band used, and control lobing effects in the transitional bands. Particularly the case for tweeters as you do not want to have complicated (shaded) line arrays for shorter wavelengths.

So combining said line array with a 12" dipole midwoofer (effective piston diameter 265mm = 1/2 wavelength of 650Hz) and a pair of 18" dipole subs, we'll possibly attain a system where the bottom four octaves are spherical figure-of-eight, the next five up to 18kHz are mostly upholding cylindric figure-of-eight.

Dieter Fricke has designed such concept slightly smaller (8" midrange, if I recall it correctly), marketed under the brand Écouton:

The lower subwoofers are dipoles as well, a concept known as RiPol. Have heard them ages ago, and they were excellent. Do you imagine something like that as an ideal dipole concept?

 

[Mort]

Takers

No thank you.

 

[Y~BNA]

Word salad.

Enjoy your meal!
Seriously, i did phrase that poorly: "at any distance", not meaning the intensity should be the same at any distance, but the proportional intensity (also when decreased over distance) should remain as even as possible over the entire frequency range. I know there's people who have phrased it better, but didn't have that quote at hand. If folks wanna know what CD is they can Google it. But it remains paramount for credible (and intelligible) soundstage at any rate. Am i wrong?

 

[Arindal]

But it remains paramount for credible (and intelligible) soundstage at any rate. Am i wrong?

While I did find the general concept to be a surprisingly good indicator for how a loudspeaker would perform in a given room and what would be a good starting point of positioning, I would not see it as a dogma. There are numerous situations in which directivity is not as important, one could think of typical nearfield listening, listening in an overdamped room (like some older studio control rooms) or under free-field conditions.

On the other hand, constant directivity and proper listening distance alone are not a guarantee that everything will work perfectly well. I have encountered several setups, particularly broad-baffle speakers in smaller rooms leading to dominant early reflections from the sides, for which the theory was not applicable and localization was compromised. Same happened with some dipol/bipol style concepts like Linkwitz 521. While tonality was excellent, the rear-firing tweeter was giving the impression that listening distance had to be reduced significantly in order to retain localization and desired proximity (otherwise tendency was a dominance of treble reverb and overly distant imaging).

 

[Y~BNA]

If the general aim is an indirect soundfield in the listening room which is diffuse, tonally balanced and sufficiently attenuated at the listening position at the same time, it is pretty obvious this goal is jeopardized by either non-constant directivity or overly low directivity, so both parameters must be ideally met at the same time.

Imagine the two extremes of a loudspeaker concept:

omnidirectional - ideally constant directivity, but way too low, leading to dominant indirect sound in almost every room
steeply increasing directivity as with some large horns, planar or fullrange speakers - sufficient suppression of indirect sound at higher frequencies, but dominant midrange reverb which will cause it to stand out from the reverb on the recording easily



An ideal constant directivity horn should be doing the job here as well, probably better than an open baffle/dipole concept. Yet, they are more difficult to design, if you really want a constant directivity, and would not accept a widening midrange directivity and overly narrowing one in the upper treble region. Most horns labelled as ´constant directivity´, do not actually meet these conditions, and those which do are pretty big.

There are other concepts imaginable to achieve these goals, like cardioids, arrays, curved planar transducers. And of course we can think of any combination of the aforementioned, as some concepts are easier to design for midrange or treble, some for the bass.

To my knowledge, GGNTKT M3 is the existing concept which comes closest to constant directivity, and it is actually a combination of several.



I am not aware of a planar tweeter delivering anything close to a constant directivity from 800Hz to 18K. Which one are you referring to?

Any truncated line source known to me shows a significant increase in directivity index either this or that way. To use such for a constant directivity concept, you have to severely limit the frequency band used, and control lobing effects in the transitional bands. Particularly the case for tweeters as you do not want to have complicated (shaded) line arrays for shorter wavelengths.


Dieter Fricke has designed such concept slightly smaller (8" midrange, if I recall it correctly), marketed under the brand Écouton:

View attachment 504902

The lower subwoofers are dipoles as well, a concept known as RiPol. Have heard them ages ago, and they were excellent. Do you imagine something like that as an ideal dipole concept?

Tx Arindal for your extensive intel on the subject!

The concept of Ripol i have to research, but the MTM with Sub in Open Dipole config looks very promising, would love to hear those.

The topic of recording venue reflections and reproduction venue reflections superimposed, affecting each other in many ways, including unpredictable ones, is something that has been on my mind for a long time. I have definitely not figured it out, but it seems reasonable we would want to minimize room reflections to keep the acoustics of the recording dominant in near field. If the room is anechoic however, i'm not sure it would be a good extreme of that proposition. As i would defiitely expect our auditory senses (whatever those may include) to register something is "off".

Anyway, the way i foresee my project is by minimizing room reflections by a combination of several physically implemented parameters controlling directivity, including driver- and baffle- -sizes and -shapes, carefully choosing frequency ranges, OB, LA, etc. I may share some sketches and more detail later, after i see where the somewhat hostile athmosphere on this forum is going.

The tweeters i'm referring to are the PT6816-8, which are dipole, emitting through two rows of vertically arranged ovular holes in the rear and four rows in the front. The holes are 6mm wide for a 15mm wide pair of rows in rear and a 35mm foursome in front. The membrane is quite a bit larger in surface than the holes combined, so it will be kind of loading energy, which would theoretically make for a very wide dispersion horizontally, as i expect the 6mm wide holes to physically behave like an array of miniature point sources, smaller than even the 17mm lambda @20k. I may use them facing backward, measurements pending. Application as a short line array would approximate a true line source, up to around 8kHz anyway, and higher with respect to horizontal coverage. They are not expensive, but measure outstandingly according to Michael Chua (ampslab.com).

I was indeed planning some amplitude shading of the tweeters, having the one next to the mid loudest, and then stepping down 1 or 1,5 dB each for the 2nd, 3rd, 4th, etc. above that. To fight temporal smearing by exploiting precedence effect. Not good?

Let's see, what else did you toch on? Yes, your opening statements. I must clarify my project is not based on a listening position, but rather on illuminating as good as possible a rather large open plan room where people live, eat, cook, sit in the couch and listen to music during much of that. People who love music so much they don't have enough time sitting in a sweet spot in their listening room..

Cheers!

 

[Y~BNA]

Critical distance is when direct sound is at the same level as reverberant sound. It is not relevant, as a concept, to small rooms other than in the high frequencies, where something like a diffuse field forms. And even then any point within the main polar pattern will be inside the critical distance. This is assuming a standard listening situation. Human sitting in front of speakers.

The idea is only useful for acoustically large rooms, like concert halls, as defined by wavelength. It's the point beyond which localization and intelligibility suffer.

If you want to hear what happens when you exceed the critical distance at home, in an unconventional way, listen to your system from another room.

Hi curvature, tx for clarifying that. Don Keele turned 87 last month, still kickin'! Your stage name anything to do with his brilliant CBT research?

I am actually aiming for illuminating large open plan living rooms without any fixed listening pos.
More or less like the venues i visit when listening to live music: a big cafe, where any seat in the hose gives you a satisfactory soundspectrum. Most instruments will be acoustic, with an electric guitar or bass here and there, and a mic-ed piano or singer. In the middle between PA and dedicated listening room/position, so to speak.

When you leave that room the spectral balance will suffer, but you still won't mistake it for anything but a live gig.

Tx for your reply!

 

[Y~BNA]

While I did find the general concept to be a surprisingly good indicator for how a loudspeaker would perform in a given room and what would be a good starting point of positioning, I would not see it as a dogma. There are numerous situations in which directivity is not as important, one could think of typical nearfield listening, listening in an overdamped room (like some older studio control rooms) or under free-field conditions.

On the other hand, constant directivity and proper listening distance alone are not a guarantee that everything will work perfectly well. I have encountered several setups, particularly broad-baffle speakers in smaller rooms leading to dominant early reflections from the sides, for which the theory was not applicable and localization was compromised. Same happened with some dipol/bipol style concepts like Linkwitz 521. While tonality was excellent, the rear-firing tweeter was giving the impression that listening distance had to be reduced significantly in order to retain localization and desired proximity (otherwise tendency was a dominance of treble reverb and overly distant imaging).

I learned a lot from spelling out Siggy's website, but he obviously isn't perfect for every application, as he himself generously admitted. So not a dogma, agreed. But anywhere i find myself, i hope the direct sound outlevels the reflected sound, the late reflections WAY outnumber the early ones, and reflections tend to uphold the spectral content of the direct sound. Any means to get there will be good, and it depends on each situation of course. Paradigms remain allowed to shift as insight progresses. Tx for reminding me. Honored you took the time.

 

[Curvature]

That is a definition you might find in older books. In practice, localization and intelligilibty are vastly independent from the critical listening distance defined by same level of direct and indirect sound, if only the reflection and reverb pattern does serve intelligibility and localization. In the latter case, reverb can even be dominant in level but both properties are retained.

You will find no evidence that reverberation can be higher in level than direct sound without detriment to localizability and intelligibility.

For the Elbphilharmonie, here are the acoustical parameters: https://www.akutek.info/Mitt Bibliotek/IOA Auditorium Acoustics Hamburg 2018/Additional/papers/p11.pdf

None of the acoustics parameters show that direct sound is lower than reverberant sound: C80 is positive across all frequencies, center time is low, as is early decay time, particularly in high frequencies where it is important.

This is a modern hall designed with electronic reinforcement and seats positioned to receive direct sound first, with deliberately low early reflections and highly-scattered late reflections. As such, clarity and consistency across frequencies were prioritized for as many seats as possible. This is in contrast to traditional rectangular halls where the above acoustical parameters as well as the strength of early reflections significantly changed depending on the position of the seat.

Again, this has little to do with small rooms.