Magnepan LRS Speaker Review

[Francis Vaughan]

Well the many worlds interpretation of quantum mechanics retains realism, determinism and locality. You have to give up at least one of those with pretty much every other interpretation, which is not an easy metaphysical choice to make.

If you are somehow forced to make a choice, and choose the least unsatisfactory. I'm happy to sit this out and not make any choice of metaphysical interpretation. It isn't as if making a choice is compulsory.

 

[MattHooper]

Even though I've never owned panel speakers one of my most memorable audio experiences resulted from, years ago, spending the day at an audiophile's home listening to a pair of ELS-63s and Gradient subs.

I´m not sure if these would work with the 2805s or if Gradient makes a dedicated sub for them but it might be worth exploring such an alternative.

I owned the ESL-63s and the Gradient dipole subs made for the 63s. To this day it's still the best integration of a dynamic driver system with a stat that I've heard.

 

[temps]

Interesting measurements for sure. I can't help but feel a full planar speaker this undersized makes for a terrible listening experience... but it would be interesting to pair this with something like Rythmik Audio's FM8 midbass/subwoofer module, but at that point, why wouldn't you just get Martin Logans? Curved panel for a wider sweet spot, a woofer to play what a small panel can't, way way less ugly, and ESLs are only slightly more money.

 

[Joppe Peelen]

Interesting measurements for sure. I can't help but feel a full planar speaker this undersized makes for a terrible listening experience... but it would be interesting to pair this with something like Rythmik Audio's FM8 midbass/subwoofer module, but at that point, why wouldn't you just get Martin Logans? Curved panel for a wider sweet spot, a woofer to play what a small panel can't, way way less ugly, and ESLs are only slightly more money.

mostly because the disperion of a curved panel are not what you expect by the curve size restricts dispersion far more then that little angle also distortion of a ML because of there curvetation is worse then typical flat esls's since mylar cant be bend... and stay that way. only by pulling on the vertical plain. forcing it to somewhat form a curve. but the mylar is never exactly in the middle. over the whole thing.
still very nice, but if i could chose an esl i would go for segmented panel they can be segmented to get any direction plot you like (well almost)

think of a big dome midrange, its a dome so you would be on axis of A spot on that dome at all times. but still there dispersion is not great , you need to go tpo a smalll dome to achieve that.

 

[Juhazi]

I owned the ESL-63s and the Gradient dipole subs made for the 63s. To this day it's still the best integration of a dynamic driver system with a stat that I've heard.

That was the onset of a Legend! And as far as I know a major inspiration for Linkwitz. Double woofers in H-frame.

https://www.gradient.fi/en/loudspeakers/11-gradient-xo-amp.html
https://www.gradient.fi/en/content/9-gradient-story

 

[Wes]

Well the many worlds interpretation of quantum mechanics retains realism, determinism and locality. You have to give up at least one of those with pretty much every other interpretation, which is not an easy metaphysical choice to make.

glad you used "interpretation" and not theory or hypothesis - it is not physics, not science and belongs in the philosophy dept. (to borrow a comment from an anti-string guy)

 

[tomtoo]

glad you used "interpretation" and not theory or hypothesis - it is not physics, not science and belongs in the philosophy dept. (to borrow a comment from an anti-string guy)

Do we realy have to get into the deepest deeps of quantum philosophy while we just measure soundwaves produced by ordinary electro-mechanical speakers??

 

[BYRTT]

Do we realy have to get into the deepest deeps of quantum philosophy while we just measure soundwaves produced by ordinary electro-mechanical speakers??

Yes ....

 

[tomtoo]

Yes ....

View attachment 84776

So you have to add some quantum fluctuations on the sine waves.

Edit says: oh im stupid sure you already have added them. They are just to tiny i can see them.

 

[williamwally]

Im going to agree with whichever quantum philosophy gives dipoles the most in-room bass

 

[BYRTT]

.....oh im stupid sure you already have added them. They are just to tiny i can see them.

Spot on in below they special filtered and humongos magnified for LRS panel ..

 

[Wes]

Do we realy have to get into the deepest deeps of quantum philosophy while we just measure soundwaves produced by ordinary electro-mechanical speakers??

my post is not about quantum anything - I just pointed out that science has certain requirements

 

[ruinevil]

Wonder how the Caintuck Alpha Bass Baffles would work with Magnepan. A pair with amplifier modules would actually come out to be cheaper than another set of LRS, which are cheaper than most good subwoofers.

 

[Newman]

Apparently so, but when asked if listening to a pair of speakers in an anechoic environment would be good, the answer I've seen to that question is "nooooooo".

Toole explains this. Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections. However, we have also the ability to distinguish direct sound from sources when indoors. We don't only hear the 'sum total'. So, if the direct sound is unnatural, we dislike it indoors as well as outdoors. And Klippel is showing us that the LRS direct sound is quite unnatural.

So then some numbers get juggled from the anechoic or pseudo-anechoic measurements to simulate their room response, but I've not seen a description of the simulated room (coulda missed it, my attention span and knowledge of the scientific data and/or proprietary algorithms is not what it could be).

Amir always shows a simulated in room response in speaker tests.

But if the direct sound is bad, then it has fallen at the first hurdle, and it's just wallpapering to look to the total room result for help. The speaker is doomed and can never reach the top shelf.

I'm just glad I'm deaf and dumb enough to like the horrible things I already have in a practical, and often measured, way.

We all are. It's human nature to pre-condition incoming raw sense data with a suite of unconscious personal cognitive biases, which we are wired to mistake for the raw sense data.

That's why bad speakers still get so much praise. Something non-sonic about them (or their back story) is so well aligned with widespread biases or values that many of us hear their sound as good anyway.

cheers

 

[Blumlein 88]

Toole explains this. Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections. However, we have also the ability to distinguish direct sound from sources when indoors. We don't only hear the 'sum total'. So, if the direct sound is unnatural, we dislike it indoors as well as outdoors. And Klippel is showing us that the LRS direct sound is quite unnatural.

Amir always shows a simulated in room response in speaker tests.

But if the direct sound is bad, then it has fallen at the first hurdle, and it's just wallpapering to look to the total room result for help. The speaker is doomed and can never reach the top shelf.

We all are. It's human nature to pre-condition incoming raw sense data with a suite of unconscious personal cognitive biases, which we are wired to mistake for the raw sense data.

That's why bad speakers still get so much praise. Something non-sonic about them (or their back story) is so well aligned with widespread biases or values that many of us hear their sound as good anyway.

cheers

Okay, let us forget about the LRS specifically and think of larger panels that aren't failed at the first hurdle of direct sound. In the section prior to where Olive's formula is presented are Klippel research results about pleasantness. Now unlike the Harman mono situation, where a panel fires its rear wave into the rear wall and straight back, in stereo, the panel is angled to the rear wall, it fires an identical form back, which will bounce over to the side, and in most rooms will then bounce toward the listener as an identical copy from the side wall reflection. Close enough in time to be integrated together with the direct sound. Klippel research shows such a result in very high in pleasantness as judged by trained listeners. In most rooms most surfaces will have an attenuation with rising frequency. Meaning those sidewall reflections aren't so different than off axis downward sloping response of conventional speakers. Plus a dipole will engage fewer room modes making for a smoother response. I do think in the mono centered presentation of dipoles used at Harman there is an artificial handicap in the results versus stereo results. Or more specifically, perhaps the Harman protocol could be improved if it were a mono presentation of either a left or right channel position. Not a centered one. Conjecture on my part, but one not without some rational support.

Box monopole speakers as tested good in the Harman protocol are undeniably very good in stereo. However, something is wrong with the dipole results. They are more satisfactory than the formula indicates.

In addition I wonder about a hybrid Frankenstein situation where I put wide narrow dispersion panels up, and augment them with a direct radiator box speaker just outside the panels, to provide the proper sidewall reflections. I see no reason in principal such a thing could not be made to work. For that matter perhaps such a strange 4 speaker for stereo arrangement could result in better Harman formula results than any single speaker is capable of achieving.

 

[tuga]

Toole explains this. Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections. However, we have also the ability to distinguish direct sound from sources when indoors. We don't only hear the 'sum total'. So, if the direct sound is unnatural, we dislike it indoors as well as outdoors. And Klippel is showing us that the LRS direct sound is quite unnatural.

What do you mean by "unnatural"?

Amir always shows a simulated in room response in speaker tests.

This is Amir's Predicted In-doom Response.

This is Stereophile's actual listening window plot:

Magnepan LRS, anechoic response on mid-panel tweeter axis at 50",
averaged across 30° horizontal window and corrected for microphone response,
with the nearfield panel response plotted below 300Hz. (source)

If you chop 6dB off the 75Hz region (an artifact caused by the nearfiled measurement) the response becomes a lot flatter below 350Hz and down to perhaps 60Hz.
Having seen hundreds of JA's measurements I am inclined to believe that his is a much more accurate illustration of reality than the Klippel's.
In other words, Klippel don't do dipoles well (enough).

Amir always shows a simulated in room response in speaker tests.

But if the direct sound is bad, then it has fallen at the first hurdle, and it's just wallpapering to look to the total room result for help. The speaker is doomed and can never reach the top shelf.

Bad (direct) sound is not just tonal balance issues; it's also distortions.

We all are. It's human nature to pre-condition incoming raw sense data with a suite of unconscious personal cognitive biases, which we are wired to mistake for the raw sense data.

That's why bad speakers still get so much praise. Something non-sonic about them (or their back story) is so well aligned with widespread biases or values that many of us hear their sound as good anyway.

Indeed.
As Toole says "Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections".
In other words, biased. So much so that many people actually enjoy room-generated distortion resulting from boundary reflection above the transiton range even though some of those are happy with treating the room or using digital correction below that...

"Preference" is a lot more complicated than we're made to believe. Taste - everyone's got one.

 

[MZKM]

What do you mean by "unnatural"?



This is Amir's Predicted In-doom Response.

This is Stereophile's actual listening window plot:

Magnepan LRS, anechoic response on mid-panel tweeter axis at 50",
averaged across 30° horizontal window and corrected for microphone response,
with the nearfield panel response plotted below 300Hz. (source)

If you chop 6dB off the 75Hz region (an artifact caused by the nearfiled measurement) the response becomes a lot flatter below 350Hz and down to perhaps 60Hz.
Having seen hundreds of JA's measurements I am inclined to believe that his is a much more accurate illustration of reality than the Klippel's.
In other words, Klippel don't do dipoles well (enough).



Bad (direct) sound is not just tonal balance issues; it's also distortions.



Indeed.
As Toole says "Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections".
In other words, biased. So much so that many people actually enjoy room-generated distortion resulting from boundary reflection above the transiton range even though some of those are happy with treating the room or using digital correction below that...

"Preference" is a lot more complicated than we're made to believe. Taste - everyone's got one.

The PIR always is a bit lower in the bass. It is more representative of those people who have their speakers like 5ft off the front wall.

The distortion graphs are done in-room, but near-field (Amir did the original tweeter axis, so ignore the treble region):

It can be called “flat” down to 50Hz. Keep in mind though we don’t like bass that is flat in room, 20Hz should be at least ~5dB higher than 1kHz, which in turn should be somewhere around 5dB higher than 20kHz.

 

[BYRTT]

What do you mean by "unnatural"?



This is Amir's Predicted In-doom Response.

This is Stereophile's actual listening window plot:

Magnepan LRS, anechoic response on mid-panel tweeter axis at 50",
averaged across 30° horizontal window and corrected for microphone response,
with the nearfield panel response plotted below 300Hz. (source)

If you chop 6dB off the 75Hz region the response becomes a lot flatter 350Hz.
Having seen hundreds of JA's measurements I am inclined to believe that his is a much more accurate illustration of reality than the Klippel's.
In other words, Klippel don't do dipoles well (enough).



Bad (direct) sound is not just tonal balance issues; it's also distortions.



Indeed.
As Toole says "Humans have spent enough time indoors/incave to feel it's unnatural if indoors and hearing sound without wall reflections".
In other words, biased. So much so that many people actually enjoy room-generated distortion resulting from boundary reflection above the transiton range even though some of those are happy with treating the room or using digital correction below that...

"Preference" is a lot more complicated than we're made to believe. Taste - everyone's got one.

Think Amir's PIR is really not so long from stereophile's one please see my post 647 few pages back, i dont say this speaker is bad also its not expensive but rather cheap and looks beautifull, but objective its bass light have non linear response or tonality and its own special signature for directivity that looks like a kind of effect box and sum of objective flaws or the directivity flaw alone could be a reason why users subjective report its superior for acoustic material.

Directivity flaw is visual in the normal spinorama if you look how directivity index goes wiggle nuts at 600Hz and up and for on axis you can see it same 600Hz and up the curve start look like sharp saw tooths, probably its because of comb filtering effect from the long panel stack having relative different distance from its bottom up to the top relative to listener area or microphone, also the low end band stack verse high end band stack are side by side..

Direct sound what is that on objective graphs, will mean its somewhere within the directivities belonging to CTA2034 standard listening window and for LRS there is not much coherence among them curves, lower graph is normalized or lets say EQeq flat on axis..

For comparison below is directivities belonging to CTA2034 standard listening window for Genelec 8341A EQed to same tonality on axis as Magnespan LRS, lower graph is normalized or lets say EQeq flat on axis..

 

[tuga]

The PIR always is a bit lower in the bass. It is more representative of those people who have their speakers like 5ft off the front wall.

The distortion graphs are done in-room, but near-field (Amir did the original tweeter axis, so ignore the treble region):

It can be called “flat” down to 50Hz. Keep in mind though we don’t like bass that is flat in room, 20Hz should be at least ~5dB higher than 1kHz, which in turn should be somewhere around 5dB higher than 20kHz.

If the PIR represents speaker at 5ft/1.5m from the front wall then it's useless for those of us in the UK and in many European homes.

The actual in-room response does indeed look closer to JA's LW plot yet different from the PIR.
It is not wrong to conclude that the PIR plot does not reflect the IR of dipole speakers with the necessary accuracy.

 

[tuga]

Think Amir's PIR is really not so long from stereophile's one please see my post 647 few pages back, i dont say this speaker is bad also its not expensive but rather cheap and looks beautifull, but objective its bass light have non linear response or tonality and its own special signature for directivity that looks like a kind of effect box and sum of objective flaws or the directivity flaw alone could be a reason why users subjective report its superior for acoustic material.

Directivity flaw is visual in the normal spinorama if you look how directivity index goes wiggle nuts at 600Hz and up and for on axis you can see it same 600Hz and up the curve start look like sharp saw tooths, probably its because of comb filtering effect from the long panel stack having relative different distance from its bottom up to the top relative to listener area or microphone, also the low end band stack verse high end band stack are side by side..
View attachment 85022

Direct sound what is that on objective graphs, will mean its somewhere within the directivities belonging to CTA2034 standard listening window and for LRS there is not much coherence among them curves, lower graph is normalized or lets say EQeq flat on axis..
View attachment 85025

For comparison below is directivities belonging to CTA2034 standard listening window for Genelec 8341A EQed to same tonality on axis as Magnespan LRS, lower graph is normalized or lets say EQeq flat on axis..
View attachment 85026

I agree that the off-axis response is far from smooth but this is somewhat mitigated by the narrow dispersion.

I also wrote earlier that the LRSs would work well if complemented by dipole bass units and preferably listened midfield to avoid the high distortion at high SPLs.