Evidence-based Speaker Designs

[tuga]

This is the main thing I don't like about the spinorama: it makes it pretty hard to eyeball how 'wide' directivity is and makes little distinction between vertical and horizontal effects. You can sometimes see the impact by looking at the difference between the early reflections and sound power curve(since the early reflections curve includes only a handful of vertical curves but sound-power is the full shebang), but the Salon2 doesn't show much of a difference. Contour plots often use different color grades and scaling.

I agree that scaling is a major problem. I've complained about this in regard to Amir's reviews.

Measurement plots are a lot more useful if the scaling is consistent.

This curve looks quite steep and the dips are more pronounce because it's been shown in 30dB-wide vertical scale when compared to the next plot:

This curve on the other hand looks very less steep and flatter comparitively but now we're on a 50dB-wide vertical scale (so we can't really compare the pitch, can we?):

 

[tuga]

Yes, this seems to reflects Dr Toole's general findings if I'm interpreting things right. In general, spaciousness is associated with better sound quality, and wider dispersion speakers tend to sound more spacious.

But stereo significantly muddies the picture (and more channels moreso). Rankings are overall the same - and this has been shown over several tests -but grouping is significantly closer in stereo. It's harder to tell apart sound quality in general in stereo. Here's one example from the Toole study I referenced earlier.

It's an older study, but as far as I know, there isn't much to invalidate it. In 1985 Toole compared three speakers in a blind test conducted in both mono and stereo. All three have similarly flat on-axis, but rather different off-axis patterns. Two of them, the the Rega Model 3 and Kef 105.2, were in the wider directivity camp but had uneven off axis. The Quad ESL 63 was narrower directivity but had more timbrally balanced off-axis (Toole said it was expected the Quad would win).

Here was their frequency response:

View attachment 52284
And these were the results in mono and stereo, for both sound quality (tonality, presumably), and spatial qualities.
View attachment 52285

This test should showed two things: sound quality is surprisingly correlated with spatial quality, and people had strong feelings about spatial quality even in mono. On the other hand stereo seemed to significantly mask the perceived flaws in mono, even if rankings were overall similar. The Quad was wayyyy behind in mono, but lost by a comparitovely small margin in stereo.

(All this from the latest edition of Toole's book Chapter 7.4.2)

Toole's research found what it found and I can't obviously contest those results.
But I am unsure whether the preference curve can be extrapolated to the general audiophile population, though it may perhaps express the taste of the majority.

Some people like horn or narrow-dispersion panel speakers too, and not everyone is obsessed with "spaciousness".

I have listened to both of those speakers before and I prefer the M2 over the Salon2. I prefer the focused imaging over the spaciousness.

In my (anecdotal) experience as a listener I find that the room-generated "soundstage effect" is actually quite detrimental to the reproduction of classical music (the genre that I most listen to) because boundary reflections in the midrange and treble affect imaging and the overlapping of the sound of your room with the sound produced by the speakers masks the ambience cues that have been registered in the recording and the illusion of "being there"is somewhat shattered...

This view can also be supported by the fact that many people prefer heaphones because they remove boundary interference and also by the number of people who choose a long-wall setup and/or opt to treat first-reflection zones.

Thanks for inviting me to your new forum Amir. For my first post, allow me to refute the notion that early reflections are ever beneficial in a home-sized room:

Early Reflections



--Ethan

But we are indeed talking about personal taste or preference here because you will also find people who enjoy semi-omni or omni speakers and others who can't do without stereo multi-channel surround.

 

[tuga]

This test should showed two things: sound quality is surprisingly correlated with spatial quality, and people had strong feelings about spatial quality even in mono. On the other hand stereo seemed to significantly mask the perceived flaws in mono, even if rankings were overall similar. The Quad was wayyyy behind in mono, but lost by a comparitovely small margin in stereo.

(All this from the latest edition of Toole's book Chapter 7.4.2)

From the article:

Ten listeners, audiophiles, audio journalists and recording professionals, participated in the evaluations.
These people were selected because they were knowledgeable, experienced and opinionated about audio—prerequisites for such a demanding test.

 

[tuga]

That said, vertical reflections can still affect tonality. Though it's not something I think designers should aim for, it's possible that an early reflections dip in the presence region is less offensive than if it existed elsewhere.

I completely agree with that but if a speaker has very wide directivity and is used in a short-wall setup with no first-reflection sone treatment then according to Toole this will have a significant impact in the tonal balance which one would conclude is why it shows in the "1st Reflection" and "Sound Power" curves.

 

[napilopez]

Toole's research found what it found and I can't obviously contest those results.
But I am unsure whether the preference curve can be extrapolated to the general audiophile population, though it may perhaps express the taste of the majority.

Some people like horn or narrow-dispersion panel speakers too, and not everyone is obsessed with "spaciousness".



In my (anecdotal) experience as a listener I find that the room-generated "soundstage effect" is actually quite detrimental to the reproduction of classical music (the genre that I most listen to) because boundary reflections in the midrange and treble affect imaging and the overlapping of the sound of your room with the sound produced by the speakers masks the ambience cues that have been registered in the recording and the illusion of "being there"is somewhat shattered...

This view can also be supported by the fact that many people prefer heaphones because they remove boundary interference and also by the number of people who choose a long-wall setup and/or opt to treat first-reflection zones.

If you saw myother post, you'll see we are in agreement . My point was that it seems the average listener, even including enthusiasts, seems to prefer wider sound. But as suggested by the Salon2 vs M2 test, as well as some other studies, some people seem to have clear preferences for different types of direcitivity. Some people have a clearly defined taste for narrower directivity and/or fewer sidewall reflections. Three people consistently chose the Salon2, three consistently chose the M2. It's just the average of the 15 leaned towards the salons. Similar results have been seen in research towards the amount of sidewall reflections people like. And there's @mitchco's personal anecdote too.

This is in contrast to frequency response, where other than some broad tilt and tonality changes, people seem consistently prefer a flattish sound. So once again, directivity is where I think loudspeaker manufacturers can differentiate themselves significantly while still making good-souding speakers on the whole.

 

[thewas]

And these were the results in mono and stereo, for both sound quality (tonality, presumably), and spatial qualities.

It should be noted that in the stereo test the higher directional Quad did quite better and almost as good at the other ones, which would fit also to the recommendation of Toole that in multichannel systems rather less reflections are preferred (so also higher directivity of speakers) as the enveloping sounds are much better delivered by the various surrounding channels compared to Stereo were "random" reflections are needed for such.

 

[napilopez]

It should be noted that in the stereo test the higher directional Quad did quite better and almost as good at the other ones, which would fit also to the recommendation of Toole that in multichannel systems rather less reflections are preferred (so also higher directivity of speakers) as the enveloping sounds are much better delivered by the various surrounding channels compared to Stereo were "random" reflections are needed for such.

Yep, also explains why there is still at least some hope for panel speakers

To your point about enveloping sounds in multichannel listening, my Yamaha receiver has several tweakable DSP modes emulating different venues. It's blasphemous to the traditional audiophile, but I love trying different virtual venues with different music when not doing 'critical' listening.' No two speakers simply cannot come close to the envelopment of my full atmos setup handling reflections - especially the envelopment from the height speakers. Some of the symphony hall settings are fantastic with certain classical recordings, and it's pretty easy to tweak the amount of atmos reflections on the fly with Yamaha's app.

 

[thewas]

I believe you, Toole is also a big fan of good multichannel upmixing via his Lexicon processor and I also prefer the Harman Logic 7 upmixing in my car to just stereo.

 

[RickS]

As Amir’s speaker testing just got off the ground, this discussion is really key. There will only be solid value in speaker testing if the measurements have more value than just identifying “bad” monopole speakers. I think that identifying really good speakers is where the challenge lies.

This said, I have heard some great monopole speakers, but within the last few years, started to evaluate other options. I have heard great things about dipoles, but have not experienced any personally. At this time I have both omnipolar and CBT speakers and hoping my observations might further the discussion on directivity. I have 2 models of Mirage OMDs and Keele’s CBT24s. I also note that every major smartspeaker is omnipolar, but not saying they necessarily sound better .

I realize that am only a single data point, but hoping my experience helps a bit. Here goes...

1. My Mirage omnipoles would be considered high directivity. I lived with them for years and heard some interesting sound, but always found I was going back to monopoles. This may be implementation related but note Dr. Toole appears to have had a comparable experience (but with Mirage dipoles).
2. I have only had my CBT24s for a few months, but am not going back. Their soundstage is beyond anything I have heard in 40 years. I have heard the bigger models are even better.
3. No matter what I have owned, have never heard anything sound better in mono. That said, I only critically listened to a few and almost always with stereo material (I have broad ranging music tastes), cannot say I have ever listened to a speaker in mono with a source produced in mono. Anybody suggest a good mono recording?

Until Amir tests a number of non-monopoles, we are unlikely to be able to use the testing to state which speakers are truly better. Maybe will have to settle for monopoles initially? I suggest that we get something like an LXmini tested earlier to see if the current tests are really comprehensive enough.

 

[Sancus]

This test should showed two things: sound quality is surprisingly correlated with spatial quality, and people had strong feelings about spatial quality even in mono. On the other hand stereo seemed to significantly mask the perceived flaws in mono, even if rankings were overall similar. The Quad was wayyyy behind in mono, but lost by a comparitovely small margin in stereo.

One of the things that bothers me about the available research is that, due to simple practical limitations, it seems to conflate wide dispersion with quality-of-response-deviation and it's hard to differentiate the two.

If you think about it, it's not really surprising that people would be bothered by narrow dispersion mono speakers in the far field, especially in a larger room. They will have much lower SPL reflections. But this isn't really the case in stereo and certainly not the case in multi-channel.

In general I think the whole idea of relying solely on room reflections to generate spaciousness is inherently flawed and Toole's opinions on multi-channel seem to support this viewpoint.

That said, to prove this with research it seems like you would need speakers with tightly controlled directivity, both wide and narrow, without major frequency response problems. Then you could compare them in mono, stereo, and multi-channel, to see if you can isolate the "directivity width" variable from everything else.

Tangent from Sierra 2 thread:

Your point about "better" is fair. But, explain to me why having wider dispersion would be a negative. I'm genuinely curious about what the rationale behind that idea is.

I think directivity width may depend on your goals/room. For example, if you're setting up a multi-channel system in a problematic asymmetrical room, perhaps you would want narrow directivity so that MOST things the listener hears are generated by direct sound. In this case you are replacing the 'generated reflections' via stereo with 'reflections' generated by your additional channels. In fact, Toole's and other opinions indicate that 'generated reflections' via upmixing in this way may be superior to nearly ANY stereo setup that relies on room reflections.

On the other hand, if you're purely using stereo in a large, symmetrical room, indeed it may be necessary to have wide directivity because otherwise the speakers won't fill the room well and you will be left without that subjective impression of 'spaciousness'.

There is an additional issue: Most of the existing research is to do with "a single seated listener in a midfield or farfield room". However, for many cases, that's not reflective of the real environment. Studio monitors, for example, often have to deal with a desk/console reflection, so it makes sense that narrower directivity and vertical directivity would be more prized. Similarly in the situation where you have multiple people listening to a mix, where some of those people may not be seated and in the exact sweet spot.

To me, that's the case that speakers like the Genelec 83xx series are designed to excel at: Handling situations where listener positioning is somewhat variable. In fact their marketing directly refers to this: “And the 8341 is possibly the best center-channel speaker ever — the image is wide but remains focused, which means others in the control room don’t have to crowd into the sweet spot to hear it correctly. The Genelecs are the tools I need for a more complex sound industry.”

 

[napilopez]

One of the things that bothers me about the available research is that, due to simple practical limitations, it seems to conflate wide dispersion with quality-of-response-deviation and it's hard to differentiate the two.

If you think about it, it's not really surprising that people would be bothered by narrow dispersion mono speakers in the far field, especially in a larger room. They will have much lower SPL reflections. But this isn't really the case in stereo and certainly not the case in multi-channel.

In general I think the whole idea of relying solely on room reflections to generate spaciousness is inherently flawed and Toole's opinions on multi-channel seem to support this viewpoint.

That said, to prove this with research it seems like you would need speakers with tightly controlled directivity, both wide and narrow, without major frequency response problems. Then you could compare them in mono, stereo, and multi-channel, to see if you can isolate the "directivity width" variable from everything else.

While you bring up some good points, I want to highlight that wide dispersion being preferred on average doesn't just come from speaker tests; part of it is from testing preferences of room treatment. Ultimately all wide vs narrow directivity fundamentally tells you is the amplitude of room reflections, which you can also test to a similar extent via treatment. And these tests seem to show that, in general, for recreational listening and mastering, reflections are preferred. This again points to wider directivity being preferred for stereo listening.

I do agree that relying on the room to create reflections is inherently flawed, but the only way to really fix that is with more channels audio or binaural audio. But we just love holding onto stereo =]

 

[Sancus]

While you bring up some good points, I want to highlight that wide dispersion being preferred on average doesn't just come from speaker tests; part of it is from testing preferences of room treatment.

I do agree that relying on the room to create reflections is inherently flawed, but the only way to really fix that is with more channels audio or binaural audio. But we just love holding onto stereo =]

That's a good point, I did forget about the room treatment testing parts.

I do agree that the wide-directivity-preference is almost certainly correct for stereo-in-symmetrical-rooms, which is definitely the most common experience. Since my main room is both asymmetrical AND small/medium at best, probably one of the worst cases for that, I'm personally all-in on the controlled directivity multi-channel approach. It helps that my experiences with real multi-channel music have totally blown stereo away.

 

[napilopez]

That's a good point, I did forget about the room treatment testing parts.

I do agree that the wide-directivity-preference is almost certainly correct for stereo-in-symmetrical-rooms, which is definitely the most common experience. Since my main room is both asymmetrical AND small/medium at best, probably one of the worst cases for that, I'm personally all-in on the controlled directivity multi-channel approach. It helps that my experiences with real multi-channel music have totally blown stereo away.

Totally make sense. And as I noted above, I really enjoy the upsampling on my yamaha reciever when it matches the program material right. Choral music played with simulated cathedral reverberation on my atmos setup is, umm heavenly.

 

[Rollomoto]

Well, here we are with a couple of new aspects that needs to be sorted:

1. Constant directivity (CD) is prefered by listeners (Toole) and harder to achieve (than just controlled dispersion)
2. Wide horizontal dispersion is prefered by listeners, with regards to domestical rooms and typical listening distances (home cinema is a different use case)
3. (1)+(2) do not exclude each other, in fact I postet a diagram that showed horizontal CD with apx. 140° (typically 90°).

It's true that there is no strict definition on constant directivity or controlled directivity, in particular, the range to which they refer. Nevertheless a would call the Revel speaker as nearly constant in the range of 3-8 kHz, where the ear is still very sensitive. It's not certain that the Reval was preferred over the M2 because (a) of the non-CD behavior, especially the dip around 2 kHz or (b) the wider horizontal dispersion.

So an evidence-based approach might look like:

• Do control dispersion as much as you can, ideally to CD over a wide range
• Do expand horizontal dispersion >90° as much as you can
• (Also: if used in very undamped "contemporary" rooms and high listening distences, do narrow vertical dispersion <90° as much as you can, to reduce ceiling and floor reflections)
• Make a solid cabinet, right bracing and damping, that is easily implemented
• Of course use modern (klippeled) drivers, go for a digital and active solution

 

[Sancus]

3. (1)+(2) do not exclude each other, in fact I postet a diagram that showed horizontal CD with apx. 140° (typically 90°).

Hm? That speaker is down -9 to -12db at 90° already. Not sure where you're getting 140° from?

Do expand horizontal dispersion >90° as much as you can

The problem with this is the number of use cases where it's not true. There is no real use case for uneven dispersion or poor frequency response. Those things are just bad.

However if you are building a speaker for use in apartments, or in studios, perhaps you wouldn't want such wide directivity. If you are building a speaker for medium-to-large symmetrical listening rooms, then sure. Is that even a smart business decision, at this point? High-end manufacturers like D&D and Kii seem to think otherwise, both making systems designed to tame rooms that are less than perfect.

E: A speaker that could change its dispersion width on command would be incredibly interesting.

 

[Rollomoto]

Hm? That speaker is down -9 to -12db at 90° already. Not sure where you're getting 140° from?

You need to know that -6dB is the common directivity measurement, so you have to look at the the red-orange curve.

 

[vavan]

A speaker that could change its dispersion width on command would be incredibly interesting

perhaps beolab can do that

 

[thewas]

perhaps beolab can do that

Beolab 50 and 90 actually can change their dispersion according to the needs of user ("party" vs. single stereo triangle listening mode)

 

[QMuse]

This test should showed two things: sound quality is surprisingly correlated with spatial quality, and people had strong feelings about spatial quality even in mono. On the other hand stereo seemed to significantly mask the perceived flaws in mono, even if rankings were overall similar. The Quad was wayyyy behind in mono, but lost by a comparitovely small margin in stereo.

And yet even in stereo Quad didn't manage to score a win but only to reduce the lead of other two. From that perspective, isn't it strange that spatial quality isn't evaluated in the Olive's scoring system?

 

[Kal Rubinson]

Beolab 50 and 90 actually can change their dispersion according to the needs of user ("party" vs. single stereo triangle listening mode)

And at least one setting in between.