Blind Listening Test 2: Neumann KH 80 vs JBL 305p MkII vs Edifier R1280T vs RCF Arya Pro5

[Floyd Toole]

Hi Floyd, as you say, the evidence so far suggests that wider dispersion speakers, preferred in mono, seem to also be preferred in stereo. Presumably for a on center listener. But wouldn't wide dispersion work against a speaker if the listener moves slightly off axis? Wouldn't the wider dispersion speaker image pull more towards the closer speaker, than a narrower dispersion speaker? I believe Earl Geddes did such demo's with his larger horn speakers "cross fired", to show a wider sweet spot. No blind tests/AES etc paper though. For that, I think only Davis may have published, which I'm sure you're aware: https://www.linkwitzlab.com/Links/Optimized-listening-area-Davies.pdf

I discuss "stereo image stabilization" in Section 15.3.1 in the 3rd edition. I discuss the Davis "modified omnidirectional" loudspeaker - it can hardly be called "directional", but it is somewhat direction controlled. The time-intensity trade-off on which such schemes rely is a very imperfect concept, but the human imagination fills a lot of gaps, I can only imagine.

 

[ROOSKIE]

Hi Floyd, as you say, the evidence so far suggests that wider dispersion speakers, preferred in mono, seem to also be preferred in stereo. Presumably for a on center listener. But wouldn't wide dispersion work against a speaker if the listener moves slightly off axis? Wouldn't the wider dispersion speaker image pull more towards the closer speaker, than a narrower dispersion speaker? I believe Earl Geddes did such demo's with his larger horn speakers "cross fired", to show a wider sweet spot. No blind tests/AES etc paper though. For that, I think only Davis may have published, which I'm sure you're aware: https://www.linkwitzlab.com/Links/Optimized-listening-area-Davies.pdf

I have a lot of speakers I am testing as I am a nerd planning some designs of my own(plus in the winter here it is just so easy to buy one more thing to test vs actually do anything outside or anything really).
I sit next to My GF about 60% of the time and we listen in stereo almost always because we are ultimately in for the musical enjoyment and we both like 'magical' stereo effects taking place in a room and the effect of a performance in front of us. We love it.
So, I have my reasonably extensive subjective experience to draw from here.

When listening as duo

• I usually aim the L speaker at the R listener and the R speaker at the L listener, but not always exactly that. (Crossfire)
• We sit pretty close together.
• The speakers are about 8 feet apart C-C
• The spot between us is about 11feet from each speaker's face
• The back wall is about 8 feet behind us.
• 9 foot ceilings, approx 12x23" room

As a side by side duo, both wide dispersion, medium and narrow generally seem to image really when comparing a speaker to itself in the center position if the treble/beam width are fairly constant.
So, imaging is perceived as diminished in 2 person side by side stereo sessions in my space when the treble response falls rapidly off axis. It is also diminished when a speaker has certain issues around the crossover point to the tweeter. Imaging is further reduced when there is excessive response and/or room mode issues from 200-400hrz

Speakers like the JBL 4309 still image bananas side by side, supremely well - stunningly well. Even the PEQ'd Klipsch Rp-160&600mVer 1 speakers I have do it well for a duo with a really fairly narrow dispersion. (Note, the Klipsch speakers I have sound horrible off axis. Such as when used for background music and moving all around.)
Speakers with uneven off axis and rapidly falling off axis just do not do as well.

This seems even further improved when the crossover point to the tweeter is lower(based on my active DIY experimenting and subjective listening to several known commercial designs) and/or the blending involves less acoustical issues. Especially within the context of holding a center image. Some speakers draw the center closer to the near speaker much more than others. I do not have a theory yet. For example the JBL 530 in a crossfire keeps the center very near the middle of the speakers(60/40)but the LS50meta often pulls toward the near speaker fairly strongly(75/25) despite having all the properties that I would think would not cause this.
ALL of this varies a bit with particular tracks and sometimes if I switch speakers with very different designs and I have not acclimated.
Going from the BMR curved cab minitor to the Klipsch can be weird for a few tracks. (and other times I adapt immediately ----> so shrug???)

Using PEQ between 200-400hrz seems to shore up vocal 'centering' and speakers with a thick responce there benfit the most.

I actually have no notable preference between single and duo listening anymore in terms of enjoyment level. Well actually I prefer listening with my GF as it is more fun, sonically the benefits of single listening position are not that great to me anymore as I have adapted well to the duo.

This is all anecdotal if that was not clear. I understand the reasons this would be better tested blind. It simply not possible for me.

 

[dasdoing]

is there any way to have an mdat of those inroom meassurements?

 

[I0000dayz]

Interesting finds. Btw what EQ settings did you use for the Edifier RT1280?

 

[thecheapseats]

...It is a wonder it all sounds as good as it does...

that^^^ is a fact... will it get better with successive generations?... future jurys will remain deadlocked, as they are today, imo...

 

[fineMen]

Stereo is both mono and double-mono.

I don't know what 'double mono' could be. Having two identical signals, one from the left, one from the right is still 'stereo' but describes a situation in which stereo doesn't work so well, to say the least. It should be a center event in front of the listener, but that image is especially fragile.

Of course two channel stereo is just a mess and the vast majority of consumers dismiss it completely whilst formally listening to a stereo setup. To increment the count of channels doesn't help, but that may depend on the target.

I said that testing speakers using "music" which was mixed on other speakers could become self referrential. That's fair, science is always self-referrential--in hindsight. Here it lead to a standard (spinorama) which as a standard usually is also self referrential.

One other way more important thing. I dare to speculate that the participants in the test were pre-occupied to select a certain type of speakers. The question was not which speaker they prefer personally, but what they think others would prefer, given that typical, partly comical hifi-sound as the target.

In this "test" here the music selection is especially hifi-ish. All this computer rendered reverberation for instance. And what really bothers me is that any spacial separation of the instruments is missing, each having its own stereo from left to right, one sitting on the other. What is that for?!

In short, I think it is very naive to expect that people think and do what they say ;-) They themselves don't know. Please get the "humanities" in next time. From their perspective a test like this is a joke, but that's their perspective, right?

 

[Svensson]

We did. I didn't describe the make up, but we had a mix of professional musicians, "audiophiles" / amateur musicians, and general public. The general public made up about 20%.



With our very limited data, we saw a similar trend. When we run another test, we plan to survey the listeners and collect data about their music and hifi background. We even toyed with administering an audiogram from an iOS app. Ultimately, we just were too limited in time on this round to add all these bits in.

@MatthewS thanks for the clarification. I do think this adds weight to your findings. We each have our preferred 'sound' but musicians and probably audiophiles also, will be more attuned to the way instruments should sound.

 

[Svensson]

I appreciate the box / whisker presenttion of the data too. Intuitively, the Neumann still seems superior, but realistically few of us have perfect rooms, so JBL (in this group) should be more than acceptable.

 

[fineMen]

I appreciate the box / whisker presenttion of the data too. Intuitively, the Neumann still seems superior, but realistically few of us have perfect rooms, so JBL (in this group) should be more than acceptable

I own the latter. For what it is it can be recommended, stronly. But what is it?

First, what is the Neuman intended to be? It is not targeted at the consumer, it is not designed to deliver the full experience. Very much like the BBC's LS3/5(A), which for some comical reasons became an icon among 'audiophiles', it is targeted at the pro/ audience for, you name it, monitoring tasks. Restrictions in bass and available volume are well acknowledged by the knowing audience. Crackling microphon lines, loss of lines, residual hum, overdrive, echoes, resonant peaks, feedback, ambient noise are of interest, not the music, speech as such. For work, not for fun.

Slightly different with the JBL. Cheap mixing in a bedroom studio, supporting especially stereo (!!) panorama, depth and an overall sketch of the balance bass vs/ treble. Quite close listening distance, taking the early reflection from the mixing console / desk into account (see frequency response). In short, by its attitude the JBL is not for fun either.

Where's the preference coming from? From liking the music, or from thinking that this speaker does "it" right? That is a crucial question, despite the quite subtle difference in formulation.

I know, we as engineering focused folks don't feel that difference that easily. Hence you can ignore it, right?

My personal experience with the fun factor of the JBL is mixed. There are lots of deficiencies that won't be of any significance when used as intended. But on the long run, say within 20minues listening to a record side, passively *g*, they become all too obvious. On the other hand, we had a house warming party using it ... (elder people only).

 

[Postlan]

It has been once again proven that flat speakers are preferred.

It is likely that JBL scored higher than Neumann, which appears flatter, because JBL's low end extend about 10Hz lower. On the other hand, it may suggest that at this level of difference, it may not be necessary to obsess over being perfectly flat.

The RCF has a typical BBC dip. so even if this speaker scored the highest in the test, I would not be surprised. However, in this case, this may be due to the fact that the mid-low frequencies are too pronounced, resulting in an overall imbalance.

 

[Thomas_A]

I own the latter. For what it is it can be recommended, stronly. But what is it?

First, what is the Neuman intended to be? It is not targeted at the consumer, it is not designed to deliver the full experience. Very much like the BBC's LS3/5(A), which for some comical reasons became an icon among 'audiophiles', it is targeted at the pro/ audience for, you name it, monitoring tasks. Restrictions in bass and available volume are well acknowledged by the knowing audience. Crackling microphon lines, loss of lines, residual hum, overdrive, echoes, resonant peaks, feedback, ambient noise are of interest, not the music, speech as such. For work, not for fun.

Slightly different with the JBL. Cheap mixing in a bedroom studio, supporting especially stereo (!!) panorama, depth and an overall sketch of the balance bass vs/ treble. Quite close listening distance, taking the early reflection from the mixing console / desk into account (see frequency response). In short, by its attitude the JBL is not for fun either.

Where's the preference coming from? From liking the music, or from thinking that this speaker does "it" right? That is a crucial question, despite the quite subtle difference in formulation.

I know, we as engineering focused folks don't feel that difference that easily. Hence you can ignore it, right?

My personal experience with the fun factor of the JBL is mixed. There are lots of deficiencies that won't be of any significance when used as intended. But on the long run, say within 20minues listening to a record side, passively *g*, they become all too obvious. On the other hand, we had a house warming party using it ... (elder people only).

When it comes to making "it right" you can always have a personal view as seen in previous studies (variation from the mean). Logically, reproduction should not add or subtract to what is on the record meaning a linear frequency response and no/low distortion. Then there are room effects which can and should be dealt with (partly at least) and some stereo flaws that may be reduced to some extent. And last but not least, variations in the recordings makes the ideally linear frequency response from the reproduction chain just a starting point for adjusting weak bass or too much treble that is present in the recording from start.

Personally, using various recordings of human voices, especially females/tenors etc is a rather good tool to evaluate speakers that make "it" right. This also equals liking the music (natural sounding, no irritating harshness, or booming mids etc), in my experience.

 

[fineMen]

When it comes to making "it right" you can always have a personal view as seen in previous studies (variation from the mean). Logically, reproduction should not add or subtract to what is on the record meaning a linear frequency response and no/low distortion.

I think such subleties are not easily detected by a person asked to evaluate a speaker. Maybe the mixer did a bad job and so the balance is wrong, which makes the speaker win which unvoluntarily corrects it?

People don't know how it is supposed to sound. You could ask the sound engineer who authored the recording, but not the consumer. There's an information gap.

I may reformulate my ideas from before, hope you don't get it wrong. People don't pick the speaker that renders the musical performance the best. They prefer the speaker that renders those markers the best that are believed to be typical for a hifi-sound. That could be clear and smooth synthetic reverberation. It is omnipresent in contemporary recordings because it is cheap and effective. A wide dispersion would engage the room more, which resembles the out-of-phase effect that people are used to hear when listening to stereo out of axis in daily life. Big bass is a clear marker for expensive speakers simply because of size, and so on and on.

As to avoid such effects, and I don't claim that these are actually at work, would be to present recordings that are far off the mainstream. Recordings that people are not used to automatically identify as 'good record, good speaker happy maker'. The "Fast Car" was such a thing when it came out--I recall this anecdotally as a personal experience ;-) It made every speaker sound 'better'! Take a flute solo, but without the artificial wind sounds; hard to find ...

Anyway, I don't want to spoil the thread with too much of my personal musings. Let me reiterate my suggestion to get the 'humanities' in, because here we address peoples' illogical feelings, sheer preference namely. I'm done ;-)

 

[Newman]

I think such subleties are not easily detected by a person asked to evaluate a speaker. Maybe the mixer did a bad job and so the balance is wrong, which makes the speaker win which unvoluntarily corrects it?

People don't know how it is supposed to sound. You could ask the sound engineer who authored the recording, but not the consumer. There's an information gap.

I understand your logic, but the research results are in. People can, and routinely do, cross that gap. It has been described as a "remarkable ability" to detect colourations in the speakers despite that information gap.

In effect, your hypothetical problem has been tested and found to not be a barrier.

 

[fineMen]

I understand your logic, but the research results are in. People can, and routinely do, cross that gap. It has been described as a "remarkable ability" to detect colourations in the speakers despite that information gap.

In effect, your hypothetical problem has been tested and found to not be a barrier.

Don't get me wrong, but 'remarkable' translates to 'magical', right? Could you point me to some research, best not behind a paywall?

 

[Newman]

Buy and read Toole's book. If you are not willing to do that, then it's not easy to credit your demands for evidence.

 

[MediumRare]

As to avoid such effects, and I don't claim that these are actually at work, would be to present recordings that are far off the mainstream. Recordings that people are not used to automatically identify as 'good record, good speaker happy maker'. The "Fast Car" was such a thing when it came out--I recall this anecdotally as a personal experience ;-) It made every speaker sound 'better'! Take a flute solo, but without the artificial wind sounds; hard to find ...

Anyway, I don't want to spoil the thread with too much of my personal musings. Let me reiterate my suggestion to get the 'humanities' in, because here we address peoples' illogical feelings, sheer preference namely. I'm done ;-)

Unfortunately, at least some of your arguments do not align with provable facts. For example, the statement above is the opposite of true: that track was demonstrated to elicit one of the greatest differences in ratings of speakers; it did NOT make all speakers sound good. The suggestion to read Toole’s book is a good one. Or simply read what he has written earlier in this thread.

 

[Blumlein 88]

I know I'm just adding to the last two posters. Buy, read and understand Toole's book. It is some of the best money and time you can spend if you are an audiophile. It does not answer all questions or give all the answers you will ever need. It is highly informative and many ideas people have about speakers have been tested beyond what many know.

The circle of confusion where in a sense nothing could be a reference, and the efforts to break that circle to have a chance to learn the truth.
A blog about that circle.

Audio's Circle of Confusion

Audio’s “Circle of Confusion” is a term coined by Floyd Toole [1] that describes the confusion that exists within the audio recording and r...

seanolive.blogspot.com

 

[thecheapseats]

...Buy, read and understand Toole's book...

and as Mr. Toole said just a few days ago up here regarding recorded music, "It is a wonder it all sounds as good as it does."... and I agree with him...

as to the 'Circle'... I understand the underlying logic of this 'circle' - but at some point-> you have to trust someone in the chain when you hit 'play' on your home system...

having seen this 'Audio Circle of Confusion' as well as discussing the phenomenon for years with others who make and record music for a living - a few issues might be considered regarding skilled engineer's methods as they apply to mics, eq, room-ambience and effects during a recording - see #1 in the "Circle" illustration above...

many people here may know this - it may be interesting for those who don't... these methods have evolved over decades as engineers are often called to work in unfamiliar studios with no previous familiarity with its monitor system - therefore, it's always wise to 'not' fully trust what you hear when you're tracking/recording... watch your meters on each piece of gear in the signal chain, pay attention, use your skills and don't screw up... it's a 'do no harm' gig...

the same methods apply whether it's a world-class studio or an artist's home studio - or even if you know the room and it's monitor system intimately...

(1) microphone best practices have always been selecting the best sounding mic for a vocalist's voice or a player's instrument - if it doesn't sound great, or at a minimum, acceptable (this is not subjective) - then try another mic, or more if possible, hopefully without interrupting the workflow of the session...

(2) no drastic amount of eq is going to fix a poor mic selection - and living with a radically eq'd performance may well be, or rather is almost guaranteed to be, seriously problematic at mix time...

(3) very minimal eq might be applied to a recorded signal - saving final eq decisions for mixing as it is preferable to have a signal that's not been radically altered before mix time... no fun having to live with a poor eq choice made when recording...

(4) ambient room quality where a performance is recorded can be treated in a pinch (with limits) using various treatments - gobos, baffles, pads - but if a poor ambient environment is insurmountable, you're recording in the wrong room/place...

(5) the only effects typically recorded with a mic signal are peak limiting and/or very mild compression in the mic's signal chain to the recording device - effects like reverb, delays, modulation effects are not recorded on the same track with the mic signal - and if recorded (which is rare) they're put on separate tracks from the dry, no-efx mic signal...

at the end of this beginning (of the recording process), there are other skilled engineers doing the mixing and mastering - and hopefully there's a quality end product after all of that... like I said at the beginning of this too long word salad - "at some point you have to trust someone in the chain when you hit 'play' on your home system"...

anyway, that's my two cents from the cheap seats...

 

[Thomas_A]

I think such subleties are not easily detected by a person asked to evaluate a speaker. Maybe the mixer did a bad job and so the balance is wrong, which makes the speaker win which unvoluntarily corrects it?

People don't know how it is supposed to sound. You could ask the sound engineer who authored the recording, but not the consumer. There's an information gap.

I may reformulate my ideas from before, hope you don't get it wrong. People don't pick the speaker that renders the musical performance the best. They prefer the speaker that renders those markers the best that are believed to be typical for a hifi-sound. That could be clear and smooth synthetic reverberation. It is omnipresent in contemporary recordings because it is cheap and effective. A wide dispersion would engage the room more, which resembles the out-of-phase effect that people are used to hear when listening to stereo out of axis in daily life. Big bass is a clear marker for expensive speakers simply because of size, and so on and on.

As to avoid such effects, and I don't claim that these are actually at work, would be to present recordings that are far off the mainstream. Recordings that people are not used to automatically identify as 'good record, good speaker happy maker'. The "Fast Car" was such a thing when it came out--I recall this anecdotally as a personal experience ;-) It made every speaker sound 'better'! Take a flute solo, but without the artificial wind sounds; hard to find ...

Anyway, I don't want to spoil the thread with too much of my personal musings. Let me reiterate my suggestion to get the 'humanities' in, because here we address peoples' illogical feelings, sheer preference namely. I'm done ;-)

As I mentioned, the individual variation is already there in the research and this includes both the recorded music and listeners. There is some research left to do, read peer review published work, that includes stereo errors and dispersion. And, as I said, human voices are quite revealing for errors, IMO, and "strangely" I still end up with an almost linear frequency response when flaws have been corrected by ear. The 1-5 kHz region stereo flaws is however still open for debate both on and off-axis, but the corrections needed are still small from a linear response.

 

[IAtaman]

Audio's Circle of Confusion

Audio’s “Circle of Confusion” is a term coined by Floyd Toole [1] that describes the confusion that exists within the audio recording and r...

seanolive.blogspot.com

Very nice article, confusing illustration..