The “Real” Speaker Preference Test

[Duke]

@Duke :
Can you elaborate a bit more on this?

Thanks for asking! I see now that when I wrote “That [i.e. "significant interaction with the music genre"] hasn't been my experience, nor the experience of my beta-testers”, what I had in mind was not clear. Sorry about that! Let me try again:

That statement describes the situation with the more evolved configuration of the multidirectional approach I had described, wherein the secondary array has a much longer reflection path than the earlier-and-discredited “twisted bipole” configuration. And I may have mis-understood what @Floyd Toole meant. Anyway let me explain my statement: I had asked my beta-testers specifically whether they found any music genres or particular recordings where the additional reflection-field energy was detrimental to either sound quality or spatial quality (with the aforementioned more evolved configuration), and they all told me that it was never detrimental. The initial set-up matters, but then it seems to be “set it and forget it”, with no potentially annoying need to optimize settings from one recording to the next.

I inserted an edit into my previous post where I think it was needed, and again THANK YOU for calling my attention to my lack of clarity.

Do you mean that the impression of "more sense of space" versus reduced clarity were independent of genre?

With the more evolved, longer-reflection-path-length configuration, assuming good initial set-up, there seems to be no degradation of clarity. The "sense of space" seems to correspond to the recording rather than to the playback room.

Or does that mean that some people were in favour of same-side-wall reflection and others were not?

Again I should have made it clear that my statement applied to the more evolved (longer-reflection-path-length) configuration and not to the "twisted bipole" which fired its additional reflection-field energy into the short-path-length, same-side-wall first reflection zone.

Imo 60-degrees-off-center to either side is probably the ideal from the standpoint of reflection arrival direction, but the issue is the arrival time of those reflections. Ime the earlier (and stronger) those first same-side-wall reflections, the more they widen the soundstage, but also the more they degrade the spatial quality in other ways. So my approach trades off having the soundstage extend well beyond the speakers (unless that's engineered into the recording), in exchange for what I believe to be a more effective presentation of the sense of space in the recording itself. To be fair, having the soundstage extend laterally beyond the speaker is arguably a more noticeable spatial attribute, so many if not most people might well prefer to have those strong early same-side-wall reflections.

In a sense, I'm trying to eliminate having strong early same-side-wall reflections but still have just as much (imo desirable) later-arriving and spectrally-correct reflection energy, and my approach uses an unorthodox bi-directional radiation pattern instead of relying on room treatments.

And this being more a general preference of the listener, not changing with genre?

You ask good questions. Yes to "not changing with genre", as far as I can tell.

As for "general preference of the listener", my answer is a "yes, But... ":

Imo there is arguably a "correct setting" for the relative loudness of the additional reflection-field energy that may well be essentially the same (or nearly so) from one listener to another. My suggesion is to find the relative loudness setting for the additional reflection-field energy where a degradation of clarity is just barely noticeable, and then set the level just barely below that threshold. I've had multiple beta-testers wind up at the exact same setting under blind test conditions, using a remote volume control with one decibel increments in a test set-up. My sample size is too small for me to be sure this "clarity threshold" level setting would be the same for different listeners, but it might be.

That being said, here's the "But... ": Some listeners may prefer the net result with the additional reflection-field energy louder than the "clarity threshold". I'm not yet aware of any of my beta-testers or customers choosing that, nor is it my preference.

 

[olieb]

Ime the earlier (and stronger) those first same-side-wall reflections, the more they widen the soundstage, but also the more they degrade the spatial quality in other ways. So my approach trades off having the soundstage extend well beyond the speakers (unless that's engineered into the recording), in exchange for what I believe to be a more effective presentation of the sense of space in the recording itself.

Thank you for the very elaborate answer.
This all seems rather plausible to me. Reflections that come later are tending to have less effect, but as this effect has positive and negative aspects the mildness can be a good compromise that has the advantage of being so benign that different kinds of recordings all go well with it.
With more normal speaker arrangement (outside of treated/advanced studios) it is just not so easy to have these late reflections only. (What amount of delay of the reflections did you have in your arrangement?)

I am actually more interested in the “normal“ situation of absorbing or not the early lateral reflections. (In stereo and in 5ch).
I was looking either there were general recommendations from the research/experience. And I would be interested mostly in classical music, so in acoustical recordings that carry significant recording room/concert hall signature.

I was wondering whether in this case/genres absorption or reflection is more preferred (or diffusion?). My idea was to go with absorption, but ….
So, I checked and found a passage in Sound Reproduction 4th ed. where this is discussed.
Chapter 12.6 ”Discussion of Acoustical Treatment Of Listening Rooms“:

In stereo listening, “perfection” does not exist. Listeners preferring pinpoint imaging soundstage can choose to absorb or diffuse side-wall refections. Those preferring a more open, spacious soundstage can leave those areas as fat wall surfaces and choose wide dispersion loudspeakers. The choice is influenced by music preferences. Hard-panned left and right sounds are essentially monophonic and usually benefit from some local refections.

As these hard-panned sounds are rare in classical recordings, my guess would be that the side-wall reflections are more helpful for mixed music and in particular for people who prefer a ”They are here“ impression.
Pinpoint location is not necessarily my ideal but I would prefer the spaciousness of the concert hall over the spaciousness of the small listening room.
So having less specular early reflections seems like a good idea.

And regarding multichannel, where it seems that many acoustical differences tend to level out somewhat:

When multiple channels are operating simultaneously, these reflections are swamped by the recorded spatial sounds and become neutral factors. With well-designed loudspeakers, having relatively uniform directivity, the effects of these side-wall refections range from neutral to slightly beneficial, the choice can be left to the customer, there is no right or wrong.

OK, that kind of closes it by leaving it open ;-).

 

[Andysu]

In the old days, people bought the 4622 and later 4722 for both home stereo and home cinema:
View attachment 521163
and sometimes 4670or 4675

View attachment 521166
screenshots from this catalog:
View attachment 521165
Meyer is in this hall now.
What can JBL (or Meyer)) offer for home use now instead of 4722?..

alien earth encounter speaker technology , take me to your to your alien JBL leader
fully JBL professional at this THX cinema up to 156 speakers

 

[Duke]

Yes, the center channel is a real dilemma for multichannel mixers - the timbre of a real loudspeaker does not match the corrupted timbre of stereo phantom images because of the crosstalk...

Any thoughts?

It looks to me like there is evidence the 2 kHz phantom center dip does not corrupt the perceived timbre nearly as much as the measured comb-filter dip implies when the phantom image is good.

In 2014 Francis Rumsey gave a lecture in which (citing the work Günther Theile) he states that the perception of coloration from stereo's 2 kHz comb filter dip goes away when we have a good phantom image. Rumsey goes on to say that as the ability to detect a spatial image improves, the ability to detect a timbral coloration decreases. He calls this phenomenon "decoloration". Here is the video of his lecture, cued up to that part:

What are your thoughts about what Francis Rumsey refers to as "decoloration"? Is it related to the phenomenon you've mentioned wherein (if I understand correctly) as the number of channels goes up, sensitivity to the sound quality of the individual channels goes down?

Am I overlooking something?

 

[Flaesh]

Any thoughts?

Let's install a good, full-fledged center channel and, with the remaining money, get a couple of LRs. Will that be good? How bad can LRs be? For what we consider "flawless" (considering the price) playback.

ps: Personally, I'm used to sounds coming from space and don't even want to see the center speaker.
By the way, in my system, it visually feels like wall-mounted subwoofers are playing all music, and the speakers are just sitting in the corners like unwanted furniture.

 

[HalSF]

I'm a bit puzzled that the OP made this thread about Harman research as it relates to speaker sales.

My anecdotal impression is that 95%+ of the discourse and discussion I've see that mentions Harman research in relation to audio products is centered on headphones, and on the Harman Curve as a point of reference for what people prefer in headphone sound. Speakers enter the topic mostly as a consideration of the way good speaker sound in a listening space can be reproduced by headphones, My old NAD Visio HP50 headphones, for eample, were marketed and praised as an effort to offer a Harman Curve-compliant closed-back headphone together with a sound that was marketed as Room Feel. My brand-new Dan Clark Audio Aeon Core headphones are being touted as a premium headphone that pretty faithfully follows the Harman Curve using more accurate up-to-date research, and that it basically hits this sweet spot without needing EQ to refine it.

I'm stumped trying to think of times that I've seen speakers being marketed or praised or reviewed by relating them in any way to Harman research. It's overwhelmingly a headphone thing, isn't it?

 

[a4eaudio]

I'm a bit puzzled that the OP made this thread about Harman research as it relates to speaker sales.

Because Sean Olive started with loudspeakers before focusing on headphones. Because Olive and Toole, at Harman, did a lot of research on preferences and measurements and how they correlate for loudspeakers. Because Sean Olive actually wrote a paper using linear regression that calculates a "Preference Score" for loudspeakers, which every review on ASR calculates. Because every review on ASR is based on Spinorama measurements, originating with Toole, who spent 28 years at Harman.

 

[HalSF]

Because Sean Olive started with loudspeakers before focusing on headphones. Because Olive and Toole, at Harman, did a lot of research on preferences and measurements and how they correlate for loudspeakers. Because Sean Olive actually wrote a paper using linear regression that calculates a "Preference Score" for loudspeakers, which every review on ASR calculates. Because every review on ASR is based on Spinorama measurements, originating with Toole, who spent 28 years at Harman.

Thank you for this extremely concise and informative clarification, and for making me feel like a fucking idiot. I guess I deserved it?



I suppose what I might have said is that all that Harman stuff seems to have been involved in promoting and successfully selling way more headphones and IEMs than speakers, and thus might not have sounded quite so stupid. But perhaps that is idiotic too.

 

[a4eaudio]

....and for making me feel like a fucking idiot. I guess I deserved it?

LOL, no need to feel like an idiot. There is no reason everyone on ASR needs to know the history of all this stuff, some of us just happen to know it. (Maybe we need to get out more )

 

[Flaesh]

Harman research

I hope the research was applied to this wonderful installation:

 

[Floyd Toole]

What are your thoughts about what Francis Rumsey refers to as "decoloration"? Is it related to the phenomenon you've mentioned wherein (if I understand correctly) as the number of channels goes up, sensitivity to the sound quality of the individual channels goes down?

Am I overlooking something?

I recall responding to this a while back, but to repeat: See Section 4.3.2 in row 4th. edition and slightly differently in Section 7.1.1 in the 3rd edition. The crosstalk coloration for centre images is clearly audible with pink noise as anyone can demonstrate for themselves using the method described in the text box on page 110 in the 4th edition. With normal recordingsl the audibility varies because of spectral content, as would be expected, but several researchers (referenced in the book) are reported as noting it as a real issue, even affecting intelligibility.

As also would be expected, the effect is measurably and audibly diminished if room reflections are present to complicate the acoustical interference - it is primarily a direct-sound phenomenon.

Theile's experiments were done in an anchoic chamber, there were no reflections, the experiment used a dummy head and. headphones (accuracy of HRTFs?), and the head positional shifts were large. Once away from the sweet spot the comb filters are asymmetrical and more complex. It is probably relevant that interference dips are harder to hear than resonant peaks so the cues that influence perceived timbre are not strongly affected.

Not discussed is old fashioned adaptation, in which experience has shown that listeners can find pleasure listening to seriously corrupted loudspeakers if they live with them. Spending time listening to the relatively minor coloration of stereo acoustical crosstalk that exists only in the sweet spot seems like a simple perceptual adjustment.

As for the suppression of coloration when listeners are aware of the loudpeakers, surely this is easily disproven by substituting an identical centre loudspeaker for a phantom image created by a pair of identical loudspeakers. This is the situation faced by recording engineers creating programs for multiple playback formats. Recording engineers who sit in the sweet spot are challenged by the timbral mismatch between a real center loudspeaker and a phantom centre image. For example: King, R., Theriault, M., Massenburg, G. (2023). “A Practical Approach to the Use of Center Channel in Immersive Music Production”, Audio Eng. Soc., 155th Convention, Paper 168.

On and on the debates go . . .

 

[Floyd Toole]

I hope the research was applied to this wonderful installation:

Clearly no research was required, merely good salesmanship. I cannot imagine what motivated this, but it was probably good for business

 

[Duke]

@Floyd Toole, thank you very much for taking the time to reply in-depth. Please don't take my having divergent opinions and pushing back a bit as me intending any disrespect whatsoever.

See Section 4.3.2 in row 4th. edition and slightly differently in Section 7.1.1 in the 3rd edition. The crosstalk coloration for centre images is clearly audible with pink noise as anyone can demonstrate for themselves using the method described in the text box on page 110 in the 4th edition. With normal recordings the audibility varies because of spectral content, as would be expected, but several researchers (referenced in the book) are reported as noting it as a real issue, even affecting intelligibility.

The outcome of the monophonic pink noise test described in the box on page 110 of your 4th edition is consistent with Theile's finding that the comb filter's effect on sound coloration is clearly audible with a monophonic signal reproduced through two loudspeakers. He used white noise instead of pink noise but I don't think that affected the outcome.

But once we move away from monophonic sound and get into two-channel sound with phantom images, Theile's findings become very interesting! He finds a strong correlation between the presence of a phantom image and the suppression of coloration from the comb filter:

“The suppression of sound colouration in the superimposed, broadband sound field for the case of two coherent sound sources takes place to the extent to which a phantom source arises. The sound colour of the phantom source does not correspond completely to the spectral properties of the ear signals.”

Theile's experiments were done in an anchoic chamber, there were no reflections...

Seems to me the complete absence of in-room reflections to fill in the comb filter dippage arguably makes his finding that comb-filter coloration is suppressed to the extent that a phantom image is present that much more robust.

Not discussed is old fashioned adaptation, in which experience has shown that listeners can find pleasure listening to seriously corrupted loudspeakers if they live with them. Spending time listening to the relatively minor coloration of stereo acoustical crosstalk that exists only in the sweet spot seems like a simple perceptual adjustment.

It's not clear to me whether adaptation was controlled for in Theile's experiments. But if we are to believe him, even the expected relatively minor coloration is suppressed in the presence of a good phantom center image.

As for the suppression of coloration when listeners are aware of the loudspeakers, surely this is easily disproven by substituting an identical centre loudspeaker for a phantom image created by a pair of identical loudspeakers. This is the situation faced by recording engineers creating programs for multiple playback formats. Recording engineers who sit in the sweet spot are challenged by the timbral mismatch between a real center loudspeaker and a phantom centre image. For example: King, R., Theriault, M., Massenburg, G. (2023). “A Practical Approach to the Use of Center Channel in Immersive Music Production”, Audio Eng. Soc., 155th Convention, Paper 168.

I don't dispute that there will be a clearly audible difference between using phantom center mode and using a discrete center channel speaker. What I do question is whether the one is obviously colored in comparison with the other in the presence of a good phantom center image. In other words, does Theile have a valid and relevant point, or not?

Ironically it was a rapid A/B comparison (rather than a deep dive into the literature) that changed my mind from believing a center-channel was required for best sound to believing something more like “not necessarily”. I was of course expecting the comb filter issue to audibly, if not obviously, favor the discrete center channel (which was identical to the left and right channel speakers), and likewise expecting a spatial quality difference in favor of the discrete center channel. The spatial quality difference dominated my perception (and that of others present), and it was in favor of the phantom center mode. If there was a sound quality difference it was sufficiently minor than none of us noticed it. I do not think this would be the universal outcome of such a comparison; I think the specifics matter, and in this instance the specifics favored the phantom center configuration.

So, within the inevitable limits of our informal experimental set-up, my two observations were that 1) the perceived sound quality in phantom center mode was on a par with using a discrete center channel; and 2) the perceived spatial quality in phantom center mode was superior to using a discrete center channel.

Is either observation plausible? Of course I could be wrong, but perhaps there is at least a little bit of supporting evidence:

Regarding the perceived sound quality parity of the phantom center configuration obviously we have Theile's finding of what Francis Rumsey calls “decoloration” in the presence of a good phantom image; and we also have this: Later in his lecture, in the context of wavefield synthesis, Rumsey reports that the least coloration comes from using only two channels! (Go to 47:16 in the video embedded in my previous post.) So, maybe two-channel sound quality doesn't necessarily inherently suck compared to using a dedicated center-channel speaker.

Regarding the perceived spatial quality superiority of the phantom center configuration, that too might be possible; quoting from the King et al paper you cited above: “Research shows that phantom center images are often preferred over a discrete center, because of the added spaciousness, envelopment, etc.” So, maybe two-channel has desirable spatial-quality attributes compared to using a dedicated center channel speaker.

In any event, imo the specifics matter.

On and on the debates go . . .

Do you think I'm being unreasonable? I'm sure I have my blind spots.

 

[TLEDDY]

I subscribe to Consumer Reports¨… here is their approach:

If you want speakers that truly excel in sound quality, a few brands consistently stand out in our lab tests.

BRAND	PROS	CONS
Sonos	Top sound quality, multiroom flexibility, easy use	Some models are WiFi-only, pricier
Edifier	Benchmark sound quality, strong value	Fewer smart features on some models
JBL	Impressive sound, versatile connectivity	Bulkier designs, higher price points
Ultimate Ears	Great portable sound, rugged build	Not as strong for mul

Their circulation is in the millions, including on-line. Just FYI…

 

[Suono]

the ability to detect a spatial image improves, the ability to detect a timbral coloration decreases.

I take it for granted that "focused" reproduction guarantees tonal fidelity. A credible spatial reconstruction is achieved only when the correct balance between the various frequencies regarding energy and time is ensured

 

[BookshelfBob]

As probably the least well informed member taking part in this thread.

Let me assure you that audio is an opaque market to the vast majority of consumers. So their preferences beyond pricing and aesthetics*, don't really mean much. Just like in Healthcare, it's opaque. No one has any idea how to gauge what they're buying.

Obviously sites like this have changed things for people like me. You can make a vaguely rational decision with surface level info. But most consumers google fu is not strong and they will never find ASR while researching.

 

[Floyd Toole]

I subscribe to Consumer Reports¨… here is their approach:

If you want speakers that truly excel in sound quality, a few brands consistently stand out in our lab tests.

BRAND	PROS	CONS
Sonos	Top sound quality, multiroom flexibility, easy use	Some models are WiFi-only, pricier
Edifier	Benchmark sound quality, strong value	Fewer smart features on some models
JBL	Impressive sound, versatile connectivity	Bulkier designs, higher price points
Ultimate Ears	Great portable sound, rugged build	Not as strong for mul

Their circulation is in the millions, including on-line. Just FYI…

They had massive circulation, and considerable market influence in HiFi loudspeakers in 2004 when Sean Olive published his AES papers proving that their methods resulted in essentially random product ratings - and they stopped reviewing HiFi loudspeakers. At least, at the time, they published their method of testing, and serious audio people knew it was wrong. Chapter 11 in the 4th edition.

Does anyone know how they arrive at their current ratings of "small" loudspeakers? It is important to know. Massive circulation is great if the information is reliable. They are in the publishing business.

I happen to have good information that Sonos and JBL do a respectable job of delivering relatively neutral sound quality. No data on the others. I own some Sonos "smart" loudspeakers for background music throughout our living area. They blend in very nicely with a few Revels.

 

[kyuu]

I think RTINGS.com has a much better methodology for rating those sorts of "lifestyle" speakers. And in any case they publish all the data so you can trawl through and make your own conclusions rather than simply relying on the site's recommendations, if you are so inclined.

 

[Floyd Toole]

I think RTINGS.com has a much better methodology for rating those sorts of "lifestyle" speakers. And in any case they publish all the data so you can trawl through and make your own conclusions rather than simply relying on the site's recommendations, if you are so inclined.

Yes, some information is better than none, but they measure "room curves" so some caution and speculation is required to draw meaningful trends.

 

[krabapple]

Does anyone have any actual facts and figures to show if speakers that have been designed to apply the Harman research findings have actually made any significant sales inroads with the buying public?

tl;dr answer for thread: No