Why evaluating the sound of a single speaker is essential

[moleman]

Welcome to ASR!

This is discussed earlier in this thread. The short answer depends on the recording. For the most part, listening to summed mono is fine. For speaker evaluation, working with reference music or pink noise is best. Amir listens to one channel of the stereo mix but has curated his selections (so potential negative mix-down issues are avoided).

Thanks so much!

 

[Holmz]

If you want my off-the-cuff opinion I'd say mono listening would probably make it easier to hear distortion and tonal colorations, but there are certain things that might only show up in stereo. In the case of apparent changes in imaging, I wonder if maybe there was something going on with crosstalk or channel-specific phase distortion happening to throw off the image...

If something like that is happening you will never know based on mono...

And a cabinet resonance could make a focused image jump to a speaker.
If tonal frequency response flatness is the main point, then a mono speaker may be ideal.

 

[xanalog]

Actually, that brings up a question related to mono listening!

I also have a related question, which this thread has brought up.
In the minds of many and based on some replies, it appears that some ambiguity still exists about the 'psycho-' variables of psycho-acoustics.

I wouldn't trust a single person on audibility of such artifacts. This is why we have measurements. Subjective uncontrolled testing is just a small compliment to measurements in a review. Anyone confused about this is firmly in anti science subjective world.

The ASR test/evaluation of electronics are achieved through the use of a "golden" (reference) unit, namely the AudioPrecision system.
Similarly, for the ASR speaker test/measurements, the Klippel becomes that "golden" standard.
Wouldn't it, then, make sense to use a "golden" mono speaker for all A/B testing.
I think @amirm already has that third "golden" unit. (his Revels).
Such a 'mono-y-mono' (A/B) comparison and his conclusions may bring about a balanced soundstage for the those who fancy his subjective and/or objective views.

 

[livinon2wheels]

Here is the result of evaluations by Sean Olive in 1994,(Olive, S.E. (1994). “A Method for Training Listeners and Selecting Program Material for Listening Tests”, Audio Eng. Soc. 97th Convention, preprint 3893.) and as reported in Figure 3.15 in the 3rd edition of my book. This shows the success rate of listeners reporting hearing resonant flaws in loudspeakers. You will see poor correlation with entertainment value.


View attachment 448682

thanks for this list Dr. Toole, I have a copy of your book that i refer to pretty often. I'd like to add one more album to your list for consideration for system evaluation. If I am not mistaken it was Linda Ronstadt's last album called 'Winter Light' I have always thought female vocalists were in general a great way to evaluate speakers and their associated electronics. This particular album has some really deep bass on a few of the tracks, features her voice close miked for the sibilance test, and the breathy airy highs that her voice possessed. This album is a bit unique in that does not have a bunch of overdriven screaming guitars, record levels are not pushing against the limiter either. If you aren't a fan, it might still be worth a listen. I want to thank you personally for all the work you have done in the audio field and lending credence to some things I thought were right but didn't know for sure. Your insight has clarified much of my thinking and I thank you for that.

 

[MarkS]

I have never understood why female vocals would make for a good listening test. Linda Ronstadt has never sung to me with her lips inches from my ear, as they are from the microphone. I have absolutely no clue what that should sound like.

 

[Heinrich]

I also have a related question, which this thread has brought up.
In the minds of many and based on some replies, it appears that some ambiguity still exists about the 'psycho-' variables of psycho-acoustics.

The ASR test/evaluation of electronics are achieved through the use of a "golden" (reference) unit, namely the AudioPrecision system.
Similarly, for the ASR speaker test/measurements, the Klippel becomes that "golden" standard.
Wouldn't it, then, make sense to use a "golden" mono speaker for all A/B testing.
I think @amirm already has that third "golden" unit. (his Revels).
Such a 'mono-y-mono' (A/B) comparison and his conclusions may bring about a balanced soundstage for the those who fancy his subjective and/or objective views.

Exactly, but we have measurement errors. No measurement is ever exact, not even when counting (Poisson statistics). The spinorama.org tells that a difference of 0,6 in Olive preference ranking is insignificant. All specimen ranking within this (again not that exact) interval (while distance is not defined on the Olive scale), have to be considered equivalent. There‘s no ‚golden‘, it‘s all just metal.

But then the standard ranking, even the measurement‘s items (freq/ resp, direct/, …) are derived from statistics, averaging person to person variations. To have a single person evaluating one of a plethora of equivalent speakers reliably, not the least sighted, after measurement, in a particular room is a major achievement, if once accomplished.

In short, the spin is great, as it sets a standard. So the mixer in studio knows roughly what you are going to hear at home. And that was it. Perfection impossible. (Loving it, actually.)

 

[xanalog]

Exactly, but we have measurement errors. No measurement is ever exact, not even when counting (Poisson statistics).

I personally value -more than any other- @amirm's "Golden Ears".
Even if/when he is on the negative slope of the Poisson distribution curve, possibly caused by caffeine or alcohol, or even a bad hair-day!

 

[Travis]

Here is the result of evaluations by Sean Olive in 1994,(Olive, S.E. (1994). “A Method for Training Listeners and Selecting Program Material for Listening Tests”, Audio Eng. Soc. 97th Convention, preprint 3893.) and as reported in Figure 3.15 in the 3rd edition of my book. This shows the success rate of listeners reporting hearing resonant flaws in loudspeakers. You will see poor correlation with entertainment value.


View attachment 448682

You posted this reminder awhile back Dr. @Floyd Toole, and Dr. @Sean Olive made a couple of social media posts about the song, and eventually a more in-depth interview about how the two you selected test music at the NRC.

The interview is here:

SoundStage! Access | SoundStageAccess.com (GoodSound.com) - Dr. Sean Olive on 35 Years of “Fast Car”

Dr. Olive discussed the hx of test tract selection, what to listen for in fast car, and also that it was very revealing of distortion:

From the linked article:

DB: If you were having a conversation with someone who’s just getting into hi-fi, and they were looking to maybe go to a stereo shop or two and audition some gear with “Fast Car” in-hand, what would you tell them to listen for? What would they be hearing that might tell them what makes one speaker better than another?

SO: First, listen to the bass. There’s a fair amount of it in “Fast Car.” The song has an electric bass and kick drum. Listen for which speaker can recreate the depths of bass better. Secondly, the hi-hat is quite prominent in the mix. Is it sparkly? Is it dull?

Also, listen carefully to her voice. This track has been found to be very revealing of distortion. The combination of the bass and her voice is such that if the speaker has trouble reproducing the bass—and we’ve heard this with bookshelf speakers—it’ll basically max out the woofer excursion, and then it will start modulating her voice.

You can actually hear her voice wavering a bit in pitch. This is something that Wolfgang Klippel, who makes test and measurement equipment, became aware of, and when he does seminars that demonstrate how to measure distortion and nonlinearities in speakers, he always uses “Fast Car” because it’s so revealing.

I think one of the reasons is, not only is the bass prominent, but there aren’t a lot of things around it, so if there are any distortion components, they’re not well-masked by other information in the recording. So the distortion tends to be quite audible.


[Emphasis mine]

 

[Travis]

In short, the spin is great, as it sets a standard.

First. I agree with what your are saying in that post, and thank you for pointing out the preference score a threshold (.6) I have been trying to find that.

I did want to clarify though, Spin (CTA-2034) sets a standard for how to measure a loud speaker, right?

It doesn’t specify a +/- for FR, either on axis, listening window, or other mix of measurements does it? Nor does it set any standards for DI/soundpower does it?

It’s proprietary, and costs a lot of money, so I don’t know the answers to those questions.

Nor have I seen any studies from AES or the Acoustical Society that discuss a correlation between recording/mixing with monitors with little or no resonances results in better spectral balance (or anything else). There might be studies that say the exact opposite (short of blown speakers, no statistical significance).

Apparently, from this interview,

SoundStage! Access | SoundStageAccess.com (GoodSound.com) - Dr. Sean Olive on 35 Years of “Fast Car”

the songs selected at NRC for listening tests were used because they knew the recording techniques, monitors used, microphones selected, etc. to try and control for “Circle of Confusion” variables. They also appeared to find some spectral balance preferences.

I have to go in and dig more to see what, if anything, got published about spectral balance, and what role monitors played on it.

 

[Newman]

...The spinorama.org tells that a difference of 0,6 in Olive preference ranking is insignificant. All specimen ranking within this (again not that exact) interval (while distance is not defined on the Olive scale), have to be considered equivalent. ...

First. I agree with what your are saying in that post, and thank you for pointing out the preference score a threshold (.6) I have been trying to find that...

Dr Olive defines a strong preference as a <=2-point spread, a moderate preference as a <= 1 point spread and 0.5 point as a slight preference.

But he also says that savvy speaker buyers like us shouldn't use it. It's only for naive beginners. We should use the full spinorama suite and interpret it with more nuance than the preference score can do.

...I did want to clarify though, Spin (CTA-2034) sets a standard for how to measure a loud speaker, right?

It doesn’t specify a +/- for FR, either on axis, listening window, or other mix of measurements does it? Nor does it set any standards for DI/soundpower does it?
...

It doesn't do any of that. It tells you how to measure a speaker, not what standards need to be passed by the speaker being measured.

cheers

 

[Travis]

Dr Olive defines a strong preference as a <=2-point spread, a moderate preference as a <= 1 point spread and 0.5 point as a slight preference.

But he also says that savvy speaker buyers like us shouldn't use it. It's only for naive beginners. We should use the full spinorama suite and interpret it with more nuance than the preference score can do.


It doesn't do any of that. It tells you how to measure a speaker, not what standards need to be passed by the speaker being measured.

cheers

Thank you on the preference scores. I heard him say that if he got a chance to do it over again he would raise the top number higher so it spread the numbers out a bit more.

Dr. Toole answered a question of mine, regarding purchase decision, way up above, that I think aligns Dr. Olive, just from a different angle.

 

[amirm]

I have never understood why female vocals would make for a good listening test.

Not every random female vocal is good for this use. The right ones though are very good at detecting midrange to treble frequency errors. Exaggerating highs is much more irritable with female vocals than male. And it is easy to realize of female vocals are recessed or forward. EQ effects as a result, are quite audibly noticeable. While I don't use these tracks, the idea has been properly tested using controlled listening tests at Harman: https://www.audiosciencereview.com/...sic-tracks-for-speaker-and-room-eq-testing.6/

AES Paper, The Subjective and Objective Evaluation of Room Correction Products
Sean E. Olive, John Jackson, Allan Devantier, David Hunt, and Sean M. Hess

JW - Jennifer Warnes, “Bird on a Wire”
TC - Tracy Chapman, “Fast Car”
JW - James Taylor, “That’s Why I’m Here”


AES Paper, A New Listener Training Software Application
Sean Olive, AES Fellow
Harman International Industries

· Tracy Chapman, "Fast Car", Tracy Chapman
· Jennifer Warnes, "Bird on a Wire", Famous Blue Rain Coat
· James Taylor "That's Why I'm Here", “That’s Why I’m Here”
· Steely Dan “Cousin Dupree”, “ Two Against Nature”
· Paula Cole, “Tiger”,” This Fire”
· “Toy Soldier March”, Reference Recording
· Pink Noise (uncorrelated)


AES Paper, Differences in Performance and Preference of Trained versus Untrained Listeners in Loudspeaker Tests: A Case Study*
Sean E. Olive, AES Fellow

James Taylor, “That’s Why I’m Here” from “That’s Why I’m Here,” Sony Records.
Little Feat, “Hangin’ on to the Good Times” from “Let It Roll,” Warner Brothers.
Tracy Chapman, “Fast Car” from “Tracy Chapman,” Elektra/Asylum Records.
Jennifer Warnes, “Bird on a Wire” from “Famous Blue Rain Coat,” Attic Records.

I have bolded the female vocals.

I have absolutely no clue what that should sound like.

No one does until you hear the same track across a number of speakers -- the thesis of Dr. Toole's work. Then you realize which speakers do it wrong, and which do it right. As Dr. Toole puts it, every one of us has an internal compass when it comes to what sounds correct. That compass works in multi-way testing. Or in my case, using EQ to detect with or without the filter.

 

[livinon2wheels]

Yes, and a few people here have scoffed at yours and Sean's use of "Fast Car" in tests, without realising that it is actually a science-based choice!

Can you explain why fast car is a science based choice?

 

[livinon2wheels]

I think the idea is that the more speakers used, the harder it is to spot their flaws, so they sound better, not dumbed-down to 'average'.

I think this speaks volumes on why home theater systems in a box sell as well as they do...in what other medium can you use cheap inferior speakers to create a soundfield that still manages to provide an acceptable experience. All the spatial cues tend to draw our attention away from a focus on what a single speaker can do.

 

[sm5]

Are there any studies on bass frequency response vs volume? For example: KH150 + KH750 sounds great - in mono or dual KH150+dual KH750'd - but switching out the sub and turning the volume up about 2-3dB I'd say the KH150 alone sounds the same if not better and turning it up a few more dB - the KH150 alone definitely sounds better? Obviously at some point it would run out of steam without the subwoofer and the sub would always win, but for at least nearfield listening levels/distances: it seems to track that way at least for me.

Otherwise is there a certain low-frequency cutoff where preference changes? For example do people generally prefer more bass down to 35-40hz and below that it's not so much of a change if there's a sub or not - such as diminishing returns below a certain frequency? Playing most frequencies below 35hz I can't really audibly hear much of anything besides rattling in things - definitely not much in most music. (Even listening to pure tones at 35 and below - there's not much audible below 35hz or so it seems)

*Another reason I'm asking this is I've also tested calibrated KH80 and KH120ii with dual KH750 subs and A/Bing them (one above the other and both slightly tilted to ear (and switched so other is on bottom and top) I still usually prefer the KH150's which I find odd. (the KH80 and 120ii still sound like smallish speakers even with the subs right under them, vs the KH150 sounds more like a normal-sized speaker?) Obviously they aren't blind test though - but a curious finding.

I suppose a similar question could be asked for Mono/Stereo vs Surround/Immersive. I find usually when you turn up the volume a bit louder on a stereo track it usually sounds better than surround at a lower volume (either a separate stereo mix vs the downmixed Atmos mix) Same with In-ear-Monitors: They usually sound better to me than an in-room surround system with subs. I'm personally finding loudness + low frequency down to 35hz or so beats the 'spatial aspect' of surround and wondering if that is what others feel too or if it's been studied. Obviously you're losing the unique spatial aspect of it ... but I guess that's why I'm asking in this Mono Single-Speaker thread.

As far as for consumer reasons: I suppose the question can be boiled down to: are two full range speakers with lots of power more preferable to most people than a multi-speaker Atmos setup with smaller speakers that can't go as loud or as low? Obviously my findings with the KH80+subwoofers being less preferred than KH150's alone is perhaps a reason I'm wondering this.

 

[amirm]

Otherwise is there a certain low-frequency cutoff where preference changes?

The research used standard speakers so no subs. And no giant speaker that went down to 20 Hz. Such testing becomes hard also due to room modes interfering. I sometimes find a speaker with more bass extension to sound worse due to this and have to deploy EQ.

 

[Holmz]

I have never understood why female vocals would make for a good listening test.

It is common for men to like a woman’s voice at times.

Linda Ronstadt has never sung to me with her lips inches from my ear, as they are from the microphone. I have absolutely no clue what that should sound like.

It is better when they are whispering in one’s ear…

 

[MattHooper]

Saw a photo on Reddit today that reminded me of some of the discussion that occurred in this thread (and some others that I can’t recall ) . Some surround enthusiasts seemed to have implied that once you are introducing a stereo system into a space, introducing discrete surround sound isn’t much tougher.

Grabbed this photo from a Reddit thread. I don’t mean to dunk on this person’s set up as I’m sure he’s quite happy.

But I think it shows some of the obvious obstacles to the adoption of discrete surround sound for consumers and even enthusiasts:

I would find those surround speakers an obvious and awkward intrusion in to the space, both ergonomically and aesthetically.

Take away those speakers and you see the relative ease of how the two channel speakers adapt to the room without intrusion.
(and I think it also shows how the adoption of three channels instead of two, which some wished happened, would also pose harder issues than two channels)

The problem is if somebody wants to make surround speakers unobtrusive generally you start looking at things like on-wall or in-ceiling speakers which typically involve having to run cables through walls or ceilings.
I did that myself for my surround system. But then I was also fortunate to have a room where that could work. Plenty of people are also working with spaces like the above photo without the type of enclosed parallel walls on which to mount every speaker.

 

[Kal Rubinson]

Saw a photo on Reddit today that reminded me of some of the discussion that occurred in this thread (and some others that I can’t recall ) . Some surround enthusiasts seemed to have implied that once you are introducing a stereo system into a space, introducing discrete surround sound isn’t much tougher.

That is a tough room.

 

[xanalog]

That is a tough room.

Made even tougher by positioning and placement.
Would a 90 degree CW rotation (where the TV+speakers are moved to the wall w/the bookcase) make things less tough?
hypothetically....