Reviewing Speakers - Measurements and Listening Tests (Video)

[Kvalsvoll]

What are your thoughts on "what to look for", if you don't mind?

To much to cover here, but first to look at are faults in the speaker that shows up as irregularities in the frequency response - mainly off-axis, then we have the radiation pattern. But - even though a more strict and narrow pattern tends to present better images, a speaker with wider pattern can be just as good, in a properly treated room, it just is different - size, location of objects, depth, height.

Example - the C1 center (old design now, not commercially available). It was difficult to make the voice appear fixed to the screen, when the center was placed below or above, the speaker itself was ever so slightly detectable as the source of the sound. Solution was even steeper crossover slope, to get very high attenuation of cone breakup -> speaker disappear.

A speaker-room combination is easier to evaluate - all rooms that have insufficient attenuation of early decay will suffer from reduced presence and realism of sound objects.

 

[PeteL]

The beneficial qualities of spatial audio generated by stereo imaging also is HIGHLY DEPENDENT on the room: size, acoustic treatment, speaker placement near a wall, seating position near the back wall, etc. Now you end up conflating these environmental qualities with the speaker's innate qualities - this is why some speakers sound like they have muddy bass in one room while in another room the bass is anemic - does the speaker suck or does the room suck? Nearfield mono listening does help eliminate some of these issues.

That make sense for reviewing, sure, but in real life, it don't matter if it's the that speaker suck or the room, You want them to sound good in your room no? What's the point in buying something only for it's innate quality if in real life it doesn't work for you. Let's just say I agree that you can better evaluate a speaker with listening to just one, but if you are trying to evaluate if it works for you, if it's a keeper or you should return, then evaluate them in the way you will be listening the music, if that makes sense.

 

[amirm]

Do spatial effects also matter?

Not in the context of a reviewer telling me about it. Exactly what would validate them and make them translatable to the reader/viewer's situation? And how did they disentangle it from content's natural spatial effects vs speaker's?

 

[amirm]

Because most people listen to two speakers?

People deploy two speakers. What they hear as far as imaging is by far dependent on the content. Have an instrument hard panned to left and then right. What controls that imaging? Speaker or content? Answer of course is content.

And two speakers reinforce sound in a room differently than one speaker. There is good reason to measure using two speakers. Just not what Amir is attempting to accomplish. That's the answer you are looking for.

I will do whatever makes the best sense and stereo testing does not. I have two speakers at times but even then I don't use both. Why would I want to heavily dilute my impression of the fidelity and engineering of the sound by testing a combo? Who says anything I hear as far as imaging applies to you? Even if you have identical room, you would have to listen to identical music for it to connect to your experience.

 

[Duke]

Not in the context of a reviewer telling me about.

Okay, that makes sense.

How about in the context of listening to music, rather than reading reviews? I assume spatial quality matters to listeners since most of us listen in stereo if not multichannel.

How would you suggest someone go about evaluating the spatial quality (or "credibility of the stereo image" to borrow @Kal Rubinson's imo well-thought-out words) of a speaker?

 

[Newman]

Why would I want to heavily dilute my impression of the fidelity and engineering of the sound by testing a combo?

You wouldn’t be diluting anything if you use 1 speaker for the tonality part of your assessment and two for the ‘other’ part.

Although I am still not sure what the ‘other’ part is: in terms of traditional reviews, it is just another excuse for reviewers to largely imagine things that aren’t even happening in the sound waves, plus massively confuse cause and effect (just like they always do), thinking ‘imaging’ is due to things like baffle construction or driver material, instead of being the simple sum of how their speaker/listener/room layout and reflections is interacting with one of the speaker’s off axis measurements, twice.

 

[amirm]

How would you suggest someone go about evaluating the spatial quality of a speaker?

Thinking out loud, I think you need a specific control room with certain properties on the reflective surfaces.
Second, you need content created in that room with certain set of speakers showing best case imaging (pinpoint accuracy in blind tests).
Third you roll in different speakers and see how they impact that imaging.

This relationship with the room and content makes it a hard problem to solve.

 

[amirm]

You wouldn’t be diluting anything if you use 1 speaker for the tonality part of your assessment and two for the ‘other’ part.

Although I am still not sure what the ‘other’ part is: in terms of traditional reviews, it is just another excuse for reviewers to largely imagine things that aren’t even happening in the sound waves, plus massively confuse cause and effect, thinking ‘imaging’ is due to things like baffle construction or driver material, instead of being the simple sum of how their speaker/listener/room layout and reflections is interacting with one of the speaker’s off axis measurements, twice.

That's exactly the problem. I can play music in stereo and say this and that happened. But what does that mean? How does it translate to a different room and content? How do we deal with the height and size differences of speakers?

I don't even know how good our memory is of spatial effects. Or our acuity in this regard.

It is a hairy, hairy problem.

 

[Newman]

How would you suggest someone go about evaluating the spatial quality (or "credibility of the stereo image" to borrow @Kal Rubinson's imo well-thought-out words) of a speaker?

Well, if you want to isolate the review to the speaker’s qualities, you had better put them in an anechoic chamber along with yourself, and listen to them “blinded”.

Otherwise, sighted, there will be the usual tendency to experience large speakers as producing a large soundstage, dipoles as producing a deep soundstage, etc, even in a non-reverberant chamber.

It’s just a gigantic mess, and, almost no matter what method is used, the result will have nothing to do with what you would experience at your home with those speakers. So what is the point of the reviewer going through it all?

 

[Duke]

This relationship with the room and content makes it a hard problem to solve.

It is a hairy, hairy problem.

Agreed. We now have measurement methods and analytic tools which correlate well with controlled blind listening results, to the point of predicting loudspeaker sound quality preferences with a remarkable degree of confidence based on sufficiently comprehensive measurements alone. But it seems that we still have a ways to go when it comes to spatial quality (or spaciousness, or "credibility of the stereo image"), though there does seem to be a correlation between sound quality preference and spatial quality preference, perhaps because both place a premium on freedom from resonances.

According to a study done by Wolfgang Klippel (perhaps you've heard of him?) and cited by @Floyd Toole starting on Page 185 of the 3rd edition of his book, sound quality and spaciousness contribute equally to the perception of "naturalness", while spaciousness dominates the perception of "pleasantness". Toole concludes:

"Therefore, whether one is a picky purist or a relaxed recreational listener, the impression of space is a significant factor."

Imo it's a judgment call whether or not to include subjective impressions of spatial quality in a review, as I can see arguments for both sides.

 

[Newman]

@Duke you are mixing things up. Toole and co never said the spatial attributes of a speaker are at issue. As long as the speaker has the right directivity characteristics, easily measured, then its responsibility for spaciousness and so on is done with.

He is talking more about the limitations of what you can do with stereo versus multichannel. And/or recording techniques.

 

[Spocko]

Well, if you want to isolate the review to the speaker’s qualities, you had better put them in an anechoic chamber along with yourself, and listen to them “blinded”.

Otherwise, sighted, there will be the usual tendency to experience large speakers as producing a large soundstage, dipoles as producing a deep soundstage, etc, even in a non-reverberant chamber.

It’s just a gigantic mess, and, almost no matter what method is used, the result will have nothing to do with what you would experience at your home with those speakers. So what is the point of the reviewer going through it all?

A well designed speaker with neutral frequency curve and well controlled directivity requires less room treatment to sound good in all rooms than an imperfect speaker. So reviews grounded in measurements and double checked with mono listening creates a hierarchy of speakers where the higher ranked speakers are more tolerant of environmental imperfections than lower ranked ones that measure far worse. And like all guidelines there are always exceptions (unique room shape, listening preferences, etc) but the vast majority would benefit from this baseline list to begin their research while manufacturers can use this list to improve their offerings or explain their design compromises.

 

[Spocko]

That make sense for reviewing, sure, but in real life, it don't matter if it's the that speaker suck or the room, You want them to sound good in your room no? What's the point in buying something only for it's innate quality if in real life it doesn't work for you. Let's just say I agree that you can better evaluate a speaker with listening to just one, but if you are trying to evaluate if it works for you, if it's a keeper or you should return, then evaluate them in the way you will be listening the music, if that makes sense.

Absolutely always bring the speakers in your room. These reviews simply help cull the massive number of available speakers to the few well designed ones within your budget otherwise you’d be buying way too many speakers to try at home.

 

[Kal Rubinson]

Not in the context of a reviewer telling me about it.

Agreed and that points to another thread speculating about determining how we can measure it.

 

[Duke]

As long as the speaker has the right directivity characteristics, easily measured, then its responsibility for spaciousness and so on is done with.

What are "the right directivity characteristics" for spaciousness?

Have they been established with the same experimental rigor as "the right directivity characteristics" for tonality?

 

[Newman]

You seem to be missing the point.

There is only one set of directivity characteristics that leads to the most preference for speakers — taking all attributes into account. Wide, and either constant or smoothly varying to a limited extent.

The rest of the ‘imaging impression’ is not something that can be ‘reviewed into’ the speaker, and instead, is down to:-

• Speaker/room/listener layout and room characteristics.
• Number of channels.
• Recording techniques and mixing/mastering techniques and multichannel technologies.
• The usual sighted bias effects.

Reviewing it as a property of the speaker is typical subjective sighted nonsense and ignorance.

cheers

 

[Duke]

You seem to be missing the point.

In that case, surely you can explain the point to me?

Ah, I see your edit now.

There is only one set of directivity characteristics that leads to the most preference for speakers — taking all attributes into account. Wide, and either constant or smoothly varying to a limited extent.

But sound quality preference is not the same thing as spatial quality preference. They do tend to track one another, but are not the same thing:

"Sound and spatial qualities both contribute to our musical pleasure, but to what extent? Here [Section 7.4.2 of his book, 3rd edition] they are shown to be of comparable importance. In fact, sound quality and spatial quality ratings have a strong tendency to track each other..." - Floyd Toole; see also the listener questionnaire in Figure 7.13, with clearly separate sections for "Sound Quality" and "Spatial Quality".

What are the spatial quality goals? It's pretty easy to articulate the tonality/sound quality goals in terms of either measurements or perceptions, but are the spatial quality goals so clearly known? And if so, what are they?

 

[Newman]

A well designed speaker with neutral frequency curve and well controlled directivity requires less room treatment to sound good in all rooms than an imperfect speaker. So reviews grounded in measurements and double checked with mono listening creates a hierarchy of speakers where the higher ranked speakers are more tolerant of environmental imperfections than lower ranked ones that measure far worse. And like all guidelines there are always exceptions (unique room shape, listening preferences, etc) but the vast majority would benefit from this baseline list to begin their research while manufacturers can use this list to improve their offerings or explain their design compromises.

Exactly. Nothing to do with sighted subjective reviewer mumbo-jumbo, and already easily available from current spinorama measurements. So no need for some here to claim new measurements are needed, or worse, it’s impossible to measure so we need subjective reviewers to educate us with lyrical prose.

 

[MattHooper]

As to a speaker's "spatial abilities" (about which we are presuming "imaging/soundstaging" etc)...

I disagree that a listener's report about listening to a stereo pair are without any value or transferable to what others may hear in other rooms.

For instance, off the top of my head, some speakers from Audio Physic (e.g. Virgo 2, Scorpio that I've had), Waveform Mach speakers (egg shaped mid/tweeter modules) or MBL Omnis...these speakers do have certain imaging/soundstaging characteristics that are pretty consistant. That is, they really "disappear" as apparent sound sources, in an effortless way. In other words, images sound "out of the box" to the degree that even hard panned sounds, voices, instruments don't seem to be "stuck in the speaker" but rather just float in the same area of the speaker, in a detached way, as if the speaker isn't producing the sound.

Yes, plenty of speakers image and soundstage and do a magic trick. But it can take more effort to approach what those speakers just seem to do effortlessly, with very little fiddling with set up. That was the case in every room I ever heard those speakers, big, medium, small, they stuck out as more easily "disappearing" and imaging more completely than most other speakers. That's also what people would remark about when they heard them in my place.

Whereas something like the Devore O (wide body 2-way) have, I find, a rightly earned reputation for being more difficult to get depth and precise imaging. They are known for casting more of a "wall of sound" with big thick sonic images, more diffuse and less pinpoint, and a bit "stuck in the speaker" compared to the types of designs I mentioned earlier. Or compared to Devores more traditional slender floor standing speakers, which just seem to effortlessly "disappear" and image with great depth and specificity....which is exactly what many people cite for why they went with the slender Devores after comparing to the wider versions...and it's exactly what I have heard in those speakers in several different set ups and rooms.

Another off the top of my head: I found Thiel speakers to have a really consistent quality about their imaging - a particularly precise focus and density to the imaging. Whenever I'd directly compare them to other brands in the same room, the other speakers imaged and soundstaged too, but compared to the Thiel they had a vague, swimmy, see-through diffuse quality. As if the sonic information for each instrument had been spread out a bit and blurry by a lack of focus. The same tracks on the Thiels was like dialing in optics to a sharper focus, like they took all the information associated with an instrument in the soundstage and focused it to a tighter, "denser" more palpable sound source in the soundstage. I had for instance Harbeth speakers and no matter how I set them up in my room, they could never achieve the depth and imaging precision of the Thiels.

That's one reason I've kept coming back to Thiel speakers over the years because I've found that quality very fleeting in other designs, and...again...it was something I heard pretty consistently, over time, in different show rooms, audiophile homes, and having owned several in my own room.

Once again...is all this to a scientific level of reliability? Nah. We've been here before. But I can't agree that the imaging/soundstaging quality of a pair of speakers can't be usefully reported on, at least in some cases. Or that a speaker's particular tendencies in stereo imaging can't be discerned in different set ups or rooms. That's just not my experience.

 

[Newman]

But sound quality preference is not the same thing as spatial quality preference. They do tend to track one another, but they are not the same thing.

Fortunately, I never said ‘sound quality preference’.

Re your Toole reference, like you say, his conclusion was that sound quality and spatial quality ratings have a strong tendency to track each other — so IMO if you have got the sound quality rating sorted, spatial won’t be much different (so why bother?). He also concluded that for stereo, the nature of the recordings is the overriding factor in spatial ratings, and “the essential ingredients of ‘imaging’ are in the recordings.” Not the speakers.