Magnepan LRS Speaker Review

[CDMC]

Nope, the losses at car drivetrains etc are huge and not neglible (up to 20%), thats why also a no-load run is performed as I wrote above to determine them.

On our Toyota Land Cruisers we are seeing upwards of 40% losses. 212 at the crank spits out about 130 at the wheels.

 

[David Harper]

That's easy!

They don't.

You just think they do.

have you ever taken an I.Q. test?

 

[RayDunzl]

have you ever taken an I.Q. test?

Hell no.

I might be disappointed with the measurement.

 

[MRC01]

On our Toyota Land Cruisers we are seeing upwards of 40% losses. 212 at the crank spits out about 130 at the wheels.

Perhaps because it's all wheel drive with automatic tranny? Cars I've dynoed over the years were all stick shift performance cars and typically had about 15% losses.
This was an entertaining diversion, but it's time to get back on topic...

 

[DonH56]

I believe that amirm's excellent measurements and expert criticisms of the LRS are all correct. So why do they sound so good to me in my room?
I've owned lots of dynamic box speakers in my 50 years as an audiophile. Some of them very good. But I've never heard anything like these Maggies.
When I reconnected my previous B&W box speakers and listened all I could hear was a box. Surprising to me since I never heard the box before. Clearly, for me,in my room, there's something going on with the LRS that the measurements aren't ,or can't, measure. I do have the lows going to a powered sub. But my subjective experience of these maggies is that their absolute sound quality, within their limitations, is superior to any dynamic speaker I've ever heard. I even gave away a pair of much more expensive floorstander box speakers to my daughter and son in law. They love them.

What Ray said.

Their very large radiating area and directional pattern makes them sound much "bigger" than many conventional speakers. Some of that is probably purely visual, of course. With typical rooms and listening distances the sound field is more enveloping (if beamy, especially for the smaller models). A little off-axis and performance falls off quick, but in the sweet spot they sound much "bigger" than many speakers due to that large radiating surface. Some people sit fairly close as the effect is enhanced. I like that (very close seating) for a little bit, but find it a bit tiring and unrealistic when "everything" comes from "everywhere". That is also why people seem to prefer them for certain types of music (or maybe why people who like certain types of music favor Maggies).

I prefer larger towers in conventional designs, at least partly because they give me that larger "presence". And of course much of it could simply be in my mind. I.e. what Ray said.

 

[KaiserSoze]

For classical physics absolutely.
What you have described it not too far off one of the core philosophical problems in quantum mechanics- that of the measurement problem. Don’t anyone think that this is relevant to this topic. Ever.

The uncertainty principle is a fundamentally different, much more esoteric concern. Of course this is completely off the thread, but if I wanted to point this out, I would probably have replied to the first post that started the tangent. Not that matters. I always enjoy reading your posts by the way. They are almost always interesting, and offer insights and points of view that other people hadn't brought up.

 

[KaiserSoze]

I thought the difference being talked about was in direct (actual) measured vs computed (from other measurements). The latter is subject to both margin of error in measurement as in the former but also the correctness and precision of the computed formula itself. Sometimes, the formulas may be just approximations which can differ from direct measurements if possible to do so.

In audio measurements, you can measure in an anechoic chamber, have a model of room acoustics and compute the "room measurement" without actually measuring in a room. It is only as good or as bad as the model used. This can be done even if it is possible to do the "actual" measurement in a room.

Unless I am misunderstanding the context.

The thing is, when the measured value, of the quantity that has been measured, has been derived partly through some arithmetic formula, or even a formula involving transcendental functions or logarithms or whatever, the measurement is still an "actual" measurement. You are clearly implying that if any math is used in deriving the measured value, the measurement isn't "actual". You are implicitly defining the meaning of the word "actual", and this is actually all you are actually doing. You might be inclined to pull out a dictionary and argue that this interpretation of the word "actual" is consistent with the definition in the dictionary. But it actually isn't. All you are actually doing is implicitly defining the meaning of the word "actual".

Dynamometers do actually measure power. In some cases, but not in all cases, the process of taking the measurement involves measurement of torque and rpm. But even in cases where this is true, power has been measured just as actually as it is in cases where the process does not involve measurements of torque and rpm. It is possible to make it seem otherwise, by using insinuation to define the meaning of the word "actual".

 

[Joppe Peelen]

Hmm i thought i overlay a picture i posted earlier about not having a huge line source and what happens if you measure them slightly of axis vertically. they do match the distortion measurement rather nicely i think. so part of the ragged top end might be because of this?

Here is the picture i posted

and here i traced the distortion measurement from post #1 and the 10cm of axis vertical and 20 cm of axis vertical for a 60 cm line source from the picture above. i know the LRS are slightly bigger. but i dont own this spreadsheet (i think) and i did not make it either thats way above my paygrade and that honer goes to Bolsersts from diyaudio

there is only happening stuff in the upper region of course but thought it was interesting. what do you think ? it does not explain the huge drop though. in the sim of a 60 cm one is also a black line that reseembled a measurement from a panel of mine that was 60-65 cm. also matches that curve really well.

 

[Francis Vaughan]

The uncertainty principle is a fundamentally different, much more esoteric concern.

The measurement problem is not the same as the Uncertaintly Principle. The Uncertainty Principle is a core aspect of QM but really is a much wider thing than just that. You can derive it from first principles (basically a bit of Fourier) without much other QM, and it has fundamental links to the foundational symmetries of physics. The measurement problem is at the heart of the various philosopical questions about the nature of QM, and has no actual theoretical underpinning or indeed any use. It basically askes about the nature of reality - if you make a measurement of a quantum process, the standard interpretation (Copenhagen) says the wave function collapses. But the actual thing that did the measurement is a physical quantum thing, itself a mess of quantum superpositions. So what the heck does it mean when we measure a thing? Is it turtles all the way down? Is QM governed by some hidden variables that actually make this all deterministic? How about a pilot wave guiding the process? Consciouness? All skirt about the core question of the measurement problem.

 

[Burning Sounds]

That will teach me to keep posting from my phone. Autocorrect is very annoying. I actually corrected it back at least once, and it still kept doing it. Humph.

The boxy thing is something I have thought about a lot over the years. I know a couple of dyed in the wool dipole enthusiasts, and the whole boxy sound comes up a lot. So, my personal take on it is that indeed, some boxes have enough internal resonant modes that are not well managed that they colour the sound. (I have read some clearly wrong attempts to explain that boxy modes can't affect the sound, even from some eminent sources.) Vented speakers have a worse time, partly because sound can exit the vent, but also because their nature requires a minimum of damping inside for the vented alignment to work at its best. Two way vented is the worst. However it need not be the case. Three way or more systems don't have frequencies in the main box that may be so affected, sealed alignments can be much more damped, and line systems (aka transmission lines - or quarter wave systems to be more correct) can have a long line of damping to kill off the resonances. They are also inherently highly braced. In my own speaker designs I have spent a lot of time experimenting with damping regimes and measuring the results, and I am convinced that for one - boxiness is very real, and that two - it need not be so.
Whether panel resonances can add to a boxy sound, or are just another nuisance, I don't know. It may depend upon what sonic colour you identify as "boxy", and given that that isn't well defined, it may be part of the problem in communicating the question in the first place. Again, my own speaker designs are silly dead. Constrained layer walls with constrained layer damping braces. Stupidly over engineered, heavy and not suitable for any sort of commercial design. Easy to do however. (And laminating the constrained layers a very messy business.)

In terms of justifying dipoles or panels, IMHO avoiding the "boxy" sound of boxes isn't a particularly good reason. Not all boxes need (or should) sound like a box. That some do isn't a reason to go without a box. The traditional speaker design has quite a few shortconings, and there are lots of attempts to address these flaws with other configurations or technologies, all with differing success, new flaws, and usually quite dubious price performance. There are quite a few reasons the vast majority of speakers are boxes.

The question of how a dipole operates in the room is however IMHO an interesting and open question. I keep citing Linkwitz, and I thing with good reason. He spent a lifetime researching the question, and his work was well respected by his peers. Including Olive. He held some specific ideas about the reasons why his speaker designs worked so well for him, ideas that there is probably some merit in forming the basis of proper research.
It is a pity the first dipole to be measured here is an entry level panel. I think the poor performance of this speaker is clouding the wider question of the value of people like Linkwitz's ideas. If someone could get an Orion or LX to Amir for measuring I think we could have a vastly more productive discussion.

No problem, autocorrect is great when it gets it right, and after spending half my working life in journalism I'm perhaps more sensitive than most. Having said that, just so I don't get hoisted by my own petard, I've become increasingly sloppy since I retired so apostrophes appear where they shouldn't and don't exist where they should. I have no doubt there are examples of my handiwork on this very site!

Yes, I was probably being too generalist about resonances and boxy sound. I was trying to explain why I hear things differently after listening to dipoles for so many years. And it's not like dipoles don't have resonances, either, but in theory they are perhaps easier to control or eliminate. Certainly the LX521 bass unit can shake the enclosure which is why it has the bridge over it to isolate the upper baffle. And it doesn't work well on a springy floor either as it has to rest on the floor for ground plane reinforcement. I've taken mine to a few audio get togethers and they have always performed well except for one occasion where I was in an upstairs room with a very springy floor. Resonating floorboards don't sound good! That's living with Linkwitz - which is probably why there is a thread on here called just that.

I would like Amir to measure an LXMini just so I can see what is happening. I've heard omnis that sound quite good in many parts of the room, but the sound has always been quite diffuse. The LXMini provides this envelopment, but with individual images precisely placed. It's very clever, and in kit form it's about the same price as an LRS.

 

[Vasr]

The thing is, when the measured value, of the quantity that has been measured, has been derived partly through some arithmetic formula, or even a formula involving transcendental functions or logarithms or whatever, the measurement is still an "actual" measurement. You are clearly implying that if any math is used in deriving the measured value, the measurement isn't "actual". You are implicitly defining the meaning of the word "actual", and this is actually all you are actually doing. You might be inclined to pull out a dictionary and argue that this interpretation of the word "actual" is consistent with the definition in the dictionary. But it actually isn't. All you are actually doing is implicitly defining the meaning of the word "actual".

Oh, come on. You are playing semantics by conflating "actual" to mean both direct and indirect measurements. The latter differentiation is common and people colloquially use actual synonymous with direct as opposed to indirect.

When you take indirect measurements, whether they correspond to the same measurements if it were to be measured directly depends (not on math as you erroneously say) but on the computation/formula used to arrive at the direct measurement. In some cases, the formula is exact (V = IR, for example, where you can measure any two directly to get the third indirectly) which is where you seem to be stuck at or it is an approximation from a model because the direct measurement is often impractical to do but the indirect measurement may be good enough for the purpose. But the latter requires separate validation that it is indeed a good representation of the direct measurements if it were possible. This is not an issue of math but the accuracy and precision of modeling that enables indirect measurement to imply a value for direct measurement. Climate change, astronomy. acoustics, quantum mechanics, all involve inexact models to infer a value that cannot be directly measured.

There is no problem in doing so as long as people don't assume the model is necessarily exact and are open to the idea of models being wrong or requiring refinement as one gathers evidence. That is the way most of science works. Not by bringing in metaphysical concepts of what is "actual" to obfuscate the issue.

 

[ctrl]

Since the measurements of the @amirm are doubted again and again, in the following the behavior of a dipole loudspeaker in free field and when placed on the ground will be examined by means of a simple simulation.

The cabinet dimensions correspond to the LRS. To keep the simulation as simple as possible, only the woofer panel was simulated and this was arranged symmetrically. This is not important for the consideration of the low frequency reproduction. On the backside the identical panel was attached for simplification, but with inverse phase - to save computing time, the simulations were only calculated up to about 1kHz.

On the picture you can see the simulated floor in light grey.
The red area was set as ideally radiating with TSP determined by me.
The blue axis represents the "listening position" of the simulation.

First we look at the free field simulation with the forward and backward radiation and the resulting sum:

With an ideal sound source, the "acoustic short circuit" / acoustic cancellation naturally works optimally. This looks worse in the simulation than Amir's measurement of reality.

Now we want to see how much sound pressure we gain in the low bass when we place the speakers on the floor (expect 6dB of course, but up to what frequency?):


To make it clearer, we consider both frequency responses in one diagram.

We can even perceive a slight "floor bounce" depending on the vertical listening position.

You could now continue the game and add back and side boundary surfaces accordingly - really deep and dynamic low frequency reproduction is not to be expected.
But I think it shows that Amir's measurement is realistic (which was to be expected) and it makes clear where the bad low frequency efficiency comes from.

The other measurements of Amir are also confirmed by simulation.
Horizontal spectrogram free-field

Vertical spectrogram free-field. Normalized to the axis frequency response to show how fast the vertical radiation contracts.

 

[Deleted member 2944]

What did you assume for a "native" response of the transducer element, by itself??

Dave.

 

[ctrl]

What did you assume for a "native" response of the transducer element, by itself??

The red area was set as ideally radiating with TSP determined by me.

 

[Joppe Peelen]

Since the measurements of the @amirm are doubted again and again, in the following the behavior of a dipole loudspeaker in free field and when placed on the ground will be examined by means of a simple simulation.

The cabinet dimensions correspond to the LRS. To keep the simulation as simple as possible, only the woofer panel was simulated and this was arranged symmetrically. This is not important for the consideration of the low frequency reproduction. On the backside the identical panel was attached for simplification, but with inverse phase - to save computing time, the simulations were only calculated up to about 1kHz.

View attachment 84285
On the picture you can see the simulated floor in light grey.
The red area was set as ideally radiating with TSP determined by me.
The blue axis represents the "listening position" of the simulation.

First we look at the free field simulation with the forward and backward radiation and the resulting sum:
View attachment 84286
With an ideal sound source, the "acoustic short circuit" / acoustic cancellation naturally works optimally. This looks worse in the simulation than Amir's measurement of reality.

Now we want to see how much sound pressure we gain in the low bass when we place the speakers on the floor (expect 6dB of course, but up to what frequency?):
View attachment 84287

To make it clearer, we consider both frequency responses in one diagram.
View attachment 84288
We can even perceive a slight "floor bounce" depending on the vertical listening position.

You could now continue the game and add back and side boundary surfaces accordingly - really deep and dynamic low frequency reproduction is not to be expected.
But I think it shows that Amir's measurement is realistic (which was to be expected) and it makes clear where the bad low frequency efficiency comes from.

The other measurements of Amir are also confirmed by simulation.

ill make a measurement real quick of a 130cm panel without any filters , not much different then a bass panel from a magnepan(single ended) (mine are a bit bigger though including baffle since there is a ribbon hanging on the side at the moment 18 inches wide and 59 inch high.
planar panel itself is 51 inch high and 9 inches wide ) so yes larger then the lrs.. i did not know they where that small.

The sim assumes a perfect light membrane and no damping of magnets and or metal i guess ? since my measurment looks rather different
no gating used since not possible of course, room interaction is clearly visible in the 1 meter distance measurement. between 90 and 300hz

oh i do have an old SMGA panel lying around that i could fling in and measure see how that coresponds.

Green: 1 meter
Purple: 50 cm
Blue: 25cm
Red 12.5 cm

the insane rise does not occur , all that dramaticaly. but might be the frame and panel being bigger. and i guess the simm does not account for weight of the foil conductors and damping by the small holes, and the magnets being in the way etc ?

BUT : it does look like they crossed the LRS really high from bass to tweeter, what you normally see is a a 12db filter but the cap and the coil are not based on the the same frequency. so the coil might filter already lower (getting rid of the peak around 300-500) and then add the cap at a higher frequency to make sure that crap at the top end (the peak near 5-10 khz (depending on width of magnets) does not show up as dramatically , thats at least what ive seen a few times in the filters.
That or having the filters from tweeter and woofer sit apart rather much so the acoustical crossover would be at 1100 hz for instance but calculated on 400 hz for instance to flatten the rising response for the bass panel

but maybe they chose for less bass since the thing is so small. so they wont suffer on there 500hz efficiency statement ?

 

[GXAlan]

You won't, it is artificial and caused by adding a far more than normal amount of listening room acoustic overlaid on the recording.
It is typical of omni speakers and well positioned panels but is a euphonic colouration as far as I am concerned.

Agree 100% about the euphonic coloration.

I don’t think I would be happy with the Magnepan as my only speakers if I also had Genelec/Revel/JBLs with their accuracy.

That said, with music like Rodrigo y Gabriela (Spanish guitar), and even Disney soundtracks, the MG-III euphonic enhancement is well worth the price of admission.

I don’t know about LRS. It doesn’t seem to measure well and I have never “loved” the modern non-ribbon Magnepans.

 

[Joppe Peelen]

Agree 100% about the euphonic coloration.

I don’t think I would be happy with the Magnepan as my only speakers if I also had Genelec/Revel/JBLs with their accuracy.

That said, with music like Rodrigo y Gabriela (Spanish guitar), and even Disney soundtracks, the MG-III euphonic enhancement is well worth the price of admission.

I don’t know about LRS. It doesn’t seem to measure well and I have never “loved” the modern non-ribbon Magnepans.

haha i like my Aladdin or lion king to on some maggies (nice score from Aladdin , i think )

 

[BYRTT]

ill make a measurement real quick of a 130cm panel without any filters.....

Yours normalized @1kHz and overlaid to sum of ctrl model, now yours 100cm curve look have same tendency (plus other indoor room noice) as sum of ctrl model ..

 

[ctrl]

The sim assumes a perfect light membrane and no damping of magnets and or metal i guess ?

Yep!

and i guess the simm does not account for weight of the foil conductors and damping by the small holes, and the magnets being in the way etc ?

Nope! I can't spit that out in a few minutes
But of course you can quickly try out what happens if for example the backside sound radiation is damped a little bit and additionally a first order lowpass is realized from 500Hz - so the dipole is not ideal anymore.

Then the free field result (4m distance) is much better (red curve is SUM):

 

[bobbooo]

The measurement problem is not the same as the Uncertaintly Principle. The Uncertainty Principle is a core aspect of QM but really is a much wider thing than just that. You can derive it from first principles (basically a bit of Fourier) without much other QM, and it has fundamental links to the foundational symmetries of physics. The measurement problem is at the heart of the various philosopical questions about the nature of QM, and has no actual theoretical underpinning or indeed any use. It basically askes about the nature of reality - if you make a measurement of a quantum process, the standard interpretation (Copenhagen) says the wave function collapses. But the actual thing that did the measurement is a physical quantum thing, itself a mess of quantum superpositions. So what the heck does it mean when we measure a thing? Is it turtles all the way down? Is QM governed by some hidden variables that actually make this all deterministic? How about a pilot wave guiding the process? Consciouness? All skirt about the core question of the measurement problem.

The measurement problem is pretty much resolved by quantum decoherence, in conjunction with a compatible interpretation of quantum mechanics such as the many-worlds interpretation.