30.7 for Condos tour

[josh358]

Wendell Diller from Magnean asked me to pass this on:

No, Magnepan's tour is not scientific market research

Dear Press and Future Attendee,

Last night, I had the privilege to talk to Robert Deutsch of Stereophile at the "30.7 for Condos" event at Audio Excellence in Toronto. Robert was there to report on the concept of a public focus group (not on the speaker we were showing). As Robert noted, there has been a lot of research on the value of focus groups. Is a public focus group more effective? (good question)

Toronto was only the second stop on this tour, but it already has the same shortcoming as the 30.7 tour in terms of reliable feedback for Magnepan--- low participation. Attendees on all the 30.7 tours were very vocal about their opinions of the 30.7, but when it came to making their opinion public, approximately 1 out of 20 attendees (or less) went online to give Magnepan feedback. This tour is much the same.

There is one obvious question on the minds of attendees--- "Has Magnepan 'sold out"? They were too polite to say it in such stark terms. Most attendees are familiar with Maggies and the problem of integrating a dynamic woofer with a panel speaker. This issue is the focal point of these tours and there was much discussion.

If you are able to attend one of the upcoming events, you can see the response of the group for yourself. But, in terms of an online consensus, the sampling rate will rate will probably be too small for a reliable indicator.

https://www.canuckaudiomart.com/forum/viewtopic.php?t=55342

- Wendell Diller, Magnepan

 

[TitaniumTroy]

Watching the video really gave you a better sense of being there than my weak pics, thanks Josh. Everything that young man said I agreed with, midrange integration with woofer was truly seamless. It will really be interesting to see the final product if it makes it to production. Which it should as it really helps take the Maggie bass to another level. IMHO in all areas regarding bass, impact, dynamics, and lower in frequency it beats the full on 30.7

 

[josh358]

Can't wait to hear it!

 

[Deleted member 2944]

I have a hard time taking seriously the opinion of anybody who has an Eric Alexander product sitting behind him.

Dave.

 

[q3cpma]

Can anyone explain the interest to me? The measurements I could find (on stereophile) are abysmal. From an outsider perspective, this looks like a brand that stuck to a concept only because their entire identity is built around it; the only thing I can give them is that most models are gorgeous.

 

[g29]

Can anyone explain the interest to me? The measurements I could find (on stereophile) are abysmal. From an outsider perspective, this looks like a brand that stuck to a concept only because their entire identity is built around it; the only thing I can give them is that most models are gorgeous.

Have you heard them in person before ?

They are a dipole planar/ribbon line source competing in a sea of cone point sources. Line sources behave differently than point sources. People tend to either love them or hate them based on their personal preferences and priorities.

The 2" thickness is also pretty cool. They were the original "flat screens" before flat screens were cool.

 

[q3cpma]

Have you heard them in person before ?

They are a dipole planar/ribbon line source competing in a sea of cone point sources. Line sources behave differently than point sources. People tend to either love them or hate them based on their personal preferences and priorities.

Do that invalidate the measurements?

 

[g29]

Do that invalidate the measurements?

Personal preferences can override measurements. A former acquaintance liked the sound of my Maggies, but his wife told him he prefered 2" Bose cubes better.

Here is a set of 20.1's with OB/dipole subs in room response.

 

[mitchco]

@josh358 any chance coming to Vancouver, Canada?

 

[josh358]

@josh358 any chance coming to Vancouver, Canada?

Last I asked, Wendell was talking about extending the tour by swinging west. He doesn't plan to visit almost every dealer the way he did on the 30.7 tour, but hopefully he'll do an event near you!

 

[josh358]

Can anyone explain the interest to me? The measurements I could find (on stereophile) are abysmal. From an outsider perspective, this looks like a brand that stuck to a concept only because their entire identity is built around it; the only thing I can give them is that most models are gorgeous.

Planars measure differently and the way Stereophile measures the on-axis frequency response of Maggies is just plain wrong.

If you look back at their review of the MG3.6, there are a series of interesting letters appended that discuss why one meter measurements don't work for a dipole. The problem is that in a dipole, he back wave is out of phase, and so back wave and front wave cancel -- the cause of the figure 8 pattern. Unfortunately, this means that below a frequency that depends on baffle size (Fequal), they roll off at 6 dB per octave.

To restore the bass response, dipole woofers use equalization (accomplished with resonant sections in planar dipole woofers, electronically with dynamic ones). However, this means that in the near field they have a response that rises as frequency decreases, because at that distance the wave hasn't diffracted around the baffle. And that in turn means that you can't measure the frequency response of a dipole woofer at 1 meter, you have to do it at the listening position. I asked John Atkinson once about why he didn't do some in-room measurements, and he said it wasn't logistically possible.

OK, problem number two -- reflections. When you're measuring the waterfall plot of a point source, you can gate the measurement so you're measuring only the output of the speaker and not the reflections from the floor and other surfaces. But what about a line source that goes almost all the way to the floor and ceiling? It's difficult or impossible to gate that out -- you'd have to put the speaker in an anechoic chamber or on a tower. So the waterfall plot ends up looking hashy. This is something that John always mentions in his writeups, but I suspect that a lot of people just look at the graphs and skip the fine print. One way to get around this is to measure the waterfall of the dipole up close -- when you do that, the hashiness goes away.

Then there are some positive that just don't show up in the measurements that are usually made. For example, for a given on-axis SPL, dipoles have a total radiated power that is 4.8 dB less than monopoles. This means that they sound good in a room with an Rt typical of that in the average listening room, while the usual omni/cardioids do better in a somewhat drier (less reverberant) acoustic. This is something that can be measured, of course, but it isn't measured in the typical review.

Another advantage of dipoles -- they tend not to excite the lateral and vertical room modes, because they're in the dipole null. This gives them smoother and more realistic bass in an untreated, unequalized setup. (They do have a suckout, usually a bit above 100 Hz, where the rear wave reflects off a front wall that's half a wavelength away, and cancels the front wave. This is a weakness that doesn't show up in the measurements either. Omnis have the same problem at 1/4 wavelength, but since they're omnis, you can split the difference between the x, y, and z axes, avoiding complete cancellation.)

In general, dipoles require less room treatment than omnis for optimal results (it's often been observed that in an anechoic chamber, speakers tend to sound the same). The converse of that is that they're notoriously fussy about placement!

There are also differences in spatial presentation. Linkwitz asserted that to properly experience a sense of space, the reflections had to be spectrally similar to the direct sound. They typical box is a cardioid at higher frequencies, omni at lower ones. That I think is why some manufacturers put rearward-facing high frequency drivers on their speakers. It also means that conventional speakers have to use baffle step compensation, something that's unnecessary in a dipole.

There's another consideration as well. Omni/cardoids tend to radiate to the sides, dipoles to the front and rear. This I think is one of the reasons that dipoles are so uncanny in their ability to convey a sense of space in a large venue like a concert hall. As you move them further out from the wall, you hear more and more depth. The ideal is a delay of over 20 ms, because the best concert halls have an ITD (time to first reflection) of 20-25 ms, and reflections earlier than that will mask that reflection and make the venue seem smaller and the sound less clear. (This is the "pre-delay" setting on a reverb processor if you have access to one and want to get a feel for what it does.) Omni cardioids tend to spread the sound laterally more than front to back, and there often isn't enough delay in the reflection to get the desired effect.

There are other differences too. The cylinder wave from a line source falls of at 1/R, while the wave from a point source falls off at 1/R^2. Floor and ceiling issues tend to disappear, because the floor and ceiling reflections create a virtual line of infinite extent (in practice, more like 2.5x the height of the room). And the speakers tend to "disappear" in part because there's no floor bounce.

The nonlinear distortion of planar drivers also tends to be low at moderate listening levels.

Here's another thing that Stereophile doesn't measure -- maximum output. It's an area in which planars are typically at a disadvantage compared to a large box speaker like the Wilsons. The larger planars will outpunch a smaller box. Above the bass range, the big ones are pretty comfortable at 110 dB SPL or even more. But a big box will do more than 120, even 130 in a few cases. So there's a disadvantage vs. some large boxes that isn't measured.

Finally, with a line source, you can get both excellent on- and off-axis response simultaneously. The Maggie true ribbon is only 1/4" across, which gives you almost perfect dispersion past 20 kHz. So you have uniform on-axis and room response.

So -- all of this stuff can be measured, but it isn't practical to measure it all in a Stereophile review, and even when the measurements are present and valid, it isn't always easy to connect the measurements to their effect on sound quality.

 

[Sancus]

I own a pair of 1.7i, and I like the dipole/line source effect that has been described as "making music sound wide/tall". I don't own them because they're accurate -- they're definitely, audibly not, no matter how much placement fiddling and room eq you throw at them. But they don't sound like (many) standard "monkey coffins" either, so in terms of making a unique impression, they definitely do that.

I've never seen anechoic measurements of Magnepans. Not too sure in general if you can use the same procedure as regular speakers to measure a dipole. Reflections seem like they would be much more important to capture. The measurements on Stereophile etc are also really old, so hopefully performance has been improved on newer versions, but who knows.

I bought my Magnepans some time ago(they were upgraded to the 'i' version from the base 1.7, because they needed servicing), and while I still like them, I've become increasingly interested in trying other options due to their poor accuracy. I live in an apartment, so I'm pretty much the target market for these speakers, I suppose.

However, I'm not convinced...

Here is a set of 20.1's with OB/dipole subs in room response.

I mean, this looks...okay, but at-listening-position frequency response is a small part of the puzzle. And that's with subs? How much work was that integration?

Here's my problem with the high-end Magnepans. They are really, really expensive for the performance level. Like, the 20.7s are ~$15K and the 30.7s are $29K? These are passive speakers! The 30.7 are way more expensive than the Revel Salon 2! That's unjustifiable to me. E: I should also note they're even more expensive than pre-built LX521s, to compare directly to another extremely well regarded dipole.

I don't know what the pricing plans for "30.7 for condos" are, but anything >$10K to me is DOA. Because the D&D 8Cs exist, provide better performance, and are even more "small space friendly" than any panel speakers that need to be placed ~2-3 feet away from walls are ever going to be no matter how much you shrink the panel width.

I will say that I've always had an absolutely fantastic experience with Magnepan in terms of customer service. And I respect them as a company. I just don't think their high end speakers are "worth it". Maybe if you have unlimited funds and multiple sets of speakers at that price. But for someone for which $10-15K on speakers would be a large investment, they just aren't keeping up with SOTA competitors in value.

 

[q3cpma]

Planars measure differently and the way Stereophile measures the on-axis frequency response of Maggies is just plain wrong.

If you look back at their review of the MG3.6, there are a series of interesting letters appended that discuss why one meter measurements don't work for a dipole. The problem is that in a dipole, he back wave is out of phase, and so back wave and front wave cancel -- the cause of the figure 8 pattern. Unfortunately, this means that below a frequency that depends on baffle size (Fequal), they roll off at 6 dB per octave.

To restore the bass response, dipole woofers use equalization (accomplished with resonant sections in planar dipole woofers, electronically with dynamic ones). However, this means that in the near field they have a response that rises as frequency decreases, because at that distance the wave hasn't diffracted around the baffle. And that in turn means that you can't measure the frequency response of a dipole woofer at 1 meter, you have to do it at the listening position. I asked John Atkinson once about why he didn't do some in-room measurements, and he said it wasn't logistically possible.

OK, problem number two -- reflections. When you're measuring the waterfall plot of a point source, you can gate the measurement so you're measuring only the output of the speaker and not the reflections from the floor and other surfaces. But what about a line source that goes almost all the way to the floor and ceiling? It's difficult or impossible to gate that out -- you'd have to put the speaker in an anechoic chamber or on a tower. So the waterfall plot ends up looking hashy. This is something that John always mentions in his writeups, but I suspect that a lot of people just look at the graphs and skip the fine print. One way to get around this is to measure the waterfall of the dipole up close -- when you do that, the hashiness goes away.

Then there are some positive that just don't show up in the measurements that are usually made. For example, for a given on-axis SPL, dipoles have a total radiated power that is 4.8 dB less than monopoles. This means that they sound good in a room with an Rt typical of that in the average listening room, while the usual omni/cardioids do better in a somewhat drier (less reverberant) acoustic. This is something that can be measured, of course, but it isn't measured in the typical review.

Another advantage of dipoles -- they tend not to excite the lateral and vertical room modes, because they're in the dipole null. This gives them smoother and more realistic bass in an untreated, unequalized setup. (They do have a suckout, usually a bit above 100 Hz, where the rear wave reflects off a front wall that's half a wavelength away, and cancels the front wave. This is a weakness that doesn't show up in the measurements either. Omnis have the same problem at 1/4 wavelength, but since they're omnis, you can split the difference between the x, y, and z axes, avoiding complete cancellation.)

In general, dipoles require less room treatment than omnis for optimal results (it's often been observed that in an anechoic chamber, speakers tend to sound the same). The converse of that is that they're notoriously fussy about placement!

There are also differences in spatial presentation. Linkwitz asserted that to properly experience a sense of space, the reflections had to be spectrally similar to the direct sound. They typical box is a cardioid at higher frequencies, omni at lower ones. That I think is why some manufacturers put rearward-facing high frequency drivers on their speakers. It also means that conventional speakers have to use baffle step compensation, something that's unnecessary in a dipole.

There's another consideration as well. Omni/cardoids tend to radiate to the sides, dipoles to the front and rear. This I think is one of the reasons that dipoles are so uncanny in their ability to convey a sense of space in a large venue like a concert hall. As you move them further out from the wall, you hear more and more depth. The ideal is a delay of over 20 ms, because the best concert halls have an ITD (time to first reflection) of 20-25 ms, and reflections earlier than that will mask that reflection and make the venue seem smaller and the sound less clear. (This is the "pre-delay" setting on a reverb processor if you have access to one and want to get a feel for what it does.) Omni cardioids tend to spread the sound laterally more than front to back, and there often isn't enough delay in the reflection to get the desired effect.

There are other differences too. The cylinder wave from a line source falls of at 1/R, while the wave from a point source falls off at 1/R^2. Floor and ceiling issues tend to disappear, because the floor and ceiling reflections create a virtual line of infinite extent (in practice, more like 2.5x the height of the room). And the speakers tend to "disappear" in part because there's no floor bounce.

The nonlinear distortion of planar drivers also tends to be low at moderate listening levels.

Here's another thing that Stereophile doesn't measure -- maximum output. It's an area in which planars are typically at a disadvantage compared to a large box speaker like the Wilsons. The larger planars will outpunch a smaller box. Above the bass range, the big ones are pretty comfortable at 110 dB SPL or even more. But a big box will do more than 120, even 130 in a few cases. So there's a disadvantage vs. some large boxes that isn't measured.

Finally, with a line source, you can get both excellent on- and off-axis response simultaneously. The Maggie true ribbon is only 1/4" across, which gives you almost perfect dispersion past 20 kHz. So you have uniform on-axis and room response.

So -- all of this stuff can be measured, but it isn't practical to measure it all in a Stereophile review, and even when the measurements are present and valid, it isn't always easy to connect the measurements to their effect on sound quality.

Thanks for the detailed answer, I didn't know much about dipoles except the fact they existed (and the cancellation). I quoted Stereophile because these are the only ones I found; Magnepan themselves giving nothing. Do you have more measurements? Especially off-axis, since this the one that seemed the least good in the Stereophile articles.
By the way, are there dealers in Europe? It seems strange to me that I didn't find any, even if I heard about Magnepans a LOT more than a good number of brands (Rythmik and Ascend are other brands I didn't find in here).

 

[josh358]

Thanks for the detailed answer, I didn't know much about dipoles except the fact they existed (and the cancellation). I quoted Stereophile because these are the only ones I found; Magnepan themselves giving nothing. Do you have more measurements? Especially off-axis, since this the one that seemed the least good in the Stereophile articles.

I have only measurements I've made on my own system, which aren't complete enough to be of use. And the only measurements I've seen from Magnepan have been on axis. But I looked at the Stereophile LRS measurements

https://www.stereophile.com/content/magnepan-lrs-loudspeaker-measurements

and a number of things came to mind.

First of all, the off-axis measurements look terrible in large part because they're dipoles! They're *designed* to roll off as you approach 90 degrees off axis. So you have to take that into account. The ideal would be a straight line, its height decreasing with angle on both sides.

The second is that they don't measure off-axis response *towards the rear.* Try doing that with a box. The LRS's behavior here will be superior to most boxes.

Number three: the quasi-ribbon models like the LRS don't have the amazing high frequency dispersion of the true ribbon models because the quasi ribbon is wider, so you get lobing as you move off axis. This is apparent in the top octave and is a flaw, though at those frequencies, you're mostly hearing what's ahead of you.

The fourth is that crossover lobing, which is inevitable, occurs side to side in a line source and top to bottom in a point source, because of the way the drivers are arranged (horizontally vs. vertically). So you'll see the crossover lobes in the lateral off-axis response of Maggies, but you won't see it in the vertical off-axis response. Whereas in boxes, the same lobing occurs, but vertically. And here's the thing: JA doesn't measure very far vertically, just +5, -15 degrees. So the lobing is present in both the Maggies and the boxes, but you can only see it on the Maggies.

Number five: vertical dispersion is essentially perfect with a line source -- not so much with a point source. Again, this is something that JA doesn't measure.

Finally, I don't know what averaging JA used but narrow peaks and dips at high frequencies are pretty much inaudible and I saw a lot of those.

I've never owned the LRS, just heard it at Axpona (where it was jaw-droppingly good for a speaker at that price), but I had a pair of MMG's and to the extent I heard issues in the polar pattern it was when I moved off axis, the tonal balance would change and they beamed some (though my old ears are much less sensitive to beaming than they used to be). My Tympani IVA's are much better behaved. But I listen on axis, and the LRS, heard on axis or reasonably close (one seat over), sounds amazing. To get a real idea of how the on- and off-axis response affect sound quality you'd presumably have at the least to apply the Harman formula, and when we look at a plot in Stereophile we can't do that.

When I did the search for the review this came up:

"As Herb Reichert noted, and Measurer-in-Chief John Atkinson agreed, the LRS requires greater-than-average care in setting up, and it rewards that care with, in HR's words, 'microscopic detail, accurate timbre, and pure-water transparency'."
https://www.stereophile.com/content...nt-budget-components-2019#ymFdRZDBpzRfIaRb.99

Just a reminder of how hard it is to interpret measurements, particularly with something like planar loudspeakers that are difficult to measure.

 

[g29]

... I mean, this looks...okay, but at-listening-position frequency response is a small part of the puzzle. And that's with subs? How much work was that integration? ....

I have monitored more than just the frequency response, but that is what most people initially look at. What other parts of the puzzle are you referring to ?

Integration effort varies with what hardware/software you have at your disposal and your level of OCD.

One thing cool about line sources is the soundstage doesn't collapse when transitioning between sitting and standing.

 

[Sancus]

I have monitored more than just the frequency response, but that is what most people initially look at. What other parts of the puzzle are you referring to ?

I mean, it tells you a limited amount about how the speakers themselves actually perform. One person's in-room measurement(assuming it's correct, not overly smoothed, etc, yours seems fine but I'm speaking in generalities here), especially in a system including other speakers, doesn't tell anybody that much about general case performance, or performance compared to other speakers.

Those measurements are okay-ish. They seem honestly 5db too bright, since it should be sloping down, but maybe you did that intentionally and it could be changed without too much trouble. 60-300hz seems pretty messy also.

But as I said in my previous post, the main issue for me is the cost involved for that performance. I have heard the 20.7s, and they sound impressive, but they damn well better for a $15,000 set of speakers...and then when you start adding subs, electronics, and all the rest, it's a LOT of money when it's outperformed by many, many cheaper speakers(in terms of accuracy) with less work.

Also, they are freakin' huge, so I would hope that the soundstage is stable when you stand up

 

[g29]

Those measurements are okay-ish. They seem honestly 5db too bright, since it should be sloping down, but maybe you did that intentionally and it could be changed without too much trouble. 60-300hz seems pretty messy also.

I prefer a flat response, others prefer a tilted response. Either are doable with today's software/house curves, so that is more of a preference issue. As for the region you mentioned, I have a lot going on there due to asymmetric room issues (but +/- 2.5dB [15Hz-18kHz] isn't bad). You are correct, it is hard to compare inroom measurements because you are also listening to the room.

But as I said in my previous post, the main issue for me is the cost involved for that performance. I have heard the 20.7s, and they sound impressive, but they damn well better for a $15,000 set of speakers...and then when you start adding subs, electronics, and all the rest, it's a LOT of money when it's outperformed by many, many cheaper speakers(in terms of accuracy) with less work.
...

I have had 1.X, 3.X and 20.X and would be happy with any of them, especially now with the ability to add OB/dipole subs to augment the lower end. As for price comparison, could you recommend some cheaper speakers that sound better (especially in the LRS to 1.7i price range) ?

... Also, they are freakin' huge, ...

That's what my wife said.

 

[Sancus]

As for price comparison, could you recommend some cheaper speakers that sound better (especially in the LRS to 1.7i price range) ?

I'm specifically referring to the high end Magnepan stuff when I say it's too expensive vs. performance. That said,I don't have a REW graph of my 1.7s, but they have audible issues in the treble and mid-bass especially. And frankly, you can feel and hear the panels vibrate substantially more at certain frequencies, which is a serious flaw for any speaker. Could some of that be resolved with room EQ? Maybe, but you can't really resolve resonances that way. I use Audyssey XT 32 w/1 sub and it does make a huge improvement in bass, but I'm just not super interested in diving into REW. I'd like to get Dirac at some point though.

The Vanatoo T0 that I have on my desktop sound more accurate than my 1.7s, and they were only ~$350 for the pair. Being tiny, they don't play particularly loud, so perhaps not a fair comparison. I am planning to try out a pair of the new Kali IN-8s(only $400/ea!) and I suspect they will resoundingly trounce the 1.7s in accuracy. They should also play quite loud. Whether or not the hiss will bother me or I will miss the dipole effect too much, I don't know! We'll find out.

The value for money, in general, definitely seems to be in the world of studio monitors compared to older passive speakers, across the board. That said, while I haven't heard them, I also would expect the Revel F35 to outperform the 1.7s, based on what I've experienced from the Revels I have heard(F206).

I don't really know too much about the LRS, other than the Stereophile review which seems to have the same flawed measurements as their other reviews of Magnepans. Is there some reason to believe they're more accurate than the 1.7i? Magnepan's page states they're supposed to "give you an idea what the 20.7 or 30.7 sounds like" but they're a 2-way quasi-ribbon design without a ribbon tweeter, so I'm skeptical.

 

[josh358]

The LRS is amazing. But if you have the 1.7i, you already know what it sounds like, with a slightly different spectral balance -- it's hearing that kind of sound out of a speaker that costs $650 that's mind blowing. But of course with 55 Hz bass extension and a quasi ribbon tweeter it's probably truer to say that it gives you a taste of what the quasi ribbon midrange on the larger models sounds like at moderate levels.

I'm still a bit mystified when you say that other speakers are more accurate. How so? The inexpensive boxes I hear don't sound realistic at all. They sound like someone is humming along with the music, rather than reproducing it, which is to say that they don't have the "microscopic detail, accurate timbre, and pure-water transparency" that Herb Reichert described. This, along with their great spatial rendition, is why I went from dynamics to planars years ago and never looked back. Dynamics have their own virtues, but realistic reproduction of acoustical music isn't one of them -- unless you can spend an inordinate amount of money, stats or planars are uparalleled in that.

I beg you to listen with your ears rather than worrying about incomplete and often poorly conducted measurements that don't tell you very much -- I can tell more about how a speaker sounds merely by knowing whether it's a ribbon, horn, electrostatic, or dynamic than I can by looking at the typical measurements in a magazine from which you'd be very hard pressed to tell why these technologies have the distinctive sounds that they do!

 

[Sancus]

I beg you to listen with your ears rather than worrying about incomplete and often poorly conducted measurements that don't tell you very much -- I can tell more about how a speaker sounds merely by knowing whether it's a ribbon, horn, electrostatic, or dynamic than I can by looking at the typical measurements in a magazine from which you'd be very hard pressed to tell why these technologies have the distinctive sounds that they do!

Um, as I've said, I own 1.7is and have listened to them for many years. They are less accurate than $350 studio monitors. I don't know what to tell you. That's the reality of the matter.

I think I'm going to bow out of this thread though -- I'm not really interested in having this argument with multiple people who are convinced of the superiority of Magnepans. It's not productive.