How would you design the Ultimate Magnepan?

[DonH56]

How do you prevent beaming from a trumpet?

Or destructive interferences from the soundboard of a piano or the top of a violin or guitar (time delays)?

Does a vibrating string interfere with itself along its length when heard from some perpendicular location (time delay again) ?

If you play instruments in a room, don't they "suffer" similar FR issues as would a speaker playing their anechoically recorded tones?

Aren't we used to all these chaotic effects and largely ignore/process them accordingly on the fly?

---

As for the panel sending resonances back toward the speaker cables, there's a 100:1 (or so) step-up transformer between the amp and the stators in an ESL, so, resonant perturbations (if there) would be reduced by 1/100th going back toward the amp.

Beaming in a trumpet is a function of the bell taper (something I am pretty intimately familiar with).

At the source, e.g. piano or guitar, the interference pattern is part of "the sound" of the instrument. Generation and playback are two different things.

A vibrating string is used as an example to show wave modes/nulls in most introductory physics classes. Wave propagation along the string is a different than the cohesive wavefront most speakers (try to) produce. I wasn't going to dig too deep, at least for now (need to get back to work, test finished).

We ignore them to a certain extent, but comb filter effects and so forth are what muck up the image of a lot of systems, especially if the back wave is not damped. The effect of killing those reflections can provide a dramatic improvement in the sound stage and image. I can ignore a lot of things but prefer to fix what I can.

 

[amirm]

Right. we have to be clear of what sort of "resonances" we are referring to here. Enclosure resonances would best be measured with an accelerometer based system, transducer resonance - FR/impedance.
When Toole/Olive reference "resonances", as in their paper, I assume they are referencing the latter.
I'm never really sure what Amir is referencing.

Happy to provide references. Here is the graph:

Here is the text:

"Loudspeaker C has a similar mismatch in level between the bass and midrange/treble, in addition to a series of resonances above 300 Hz that appear in all of the spatially averaged curves."

Here is the AES paper: http://www.aes.org/e-lib/browse.cfm?elib=16108

Did you read Don's post?

 

[AJ Soundfield]

We ignore them to a certain extent, but comb filter effects and so forth are what muck up the image of a lot of systems

I suspect that to be the main culprit when switching between a piston source and a large panel, in mono, blind. There may be timbral differences due to the sound power at the LP, but the spatial "distortion" might be the most notable, especially with quick switch.

especially if the back wave is not damped. The effect of killing those reflections can provide a dramatic improvement in the sound stage and image.

That would of course depend on their delay/intensity relative to the onset response...and how you define "improvement", unless you really meant "prefer".

cheers,

AJ

 

[AJ Soundfield]

Did you read Don's post?

Which one? Where he said he's not a speaker designer but plays one on TV? I saw something about sharks with laser beams on their freaking foreheads, didn't see anything relevant to impedance.

"Loudspeaker C has a similar mismatch in level between the bass and midrange/treble, in addition to a series of resonances above 300 Hz that appear in all of the spatially averaged curves."

Right and the corresponding impedance showing those resonances is where?
Have you forwarded my post to the good Dr yet?

 

[amirm]

Which one? Where he said he's not a speaker designer but plays one on TV? I saw something about sharks with laser beams on their freaking foreheads, didn't see anything relevant to impedance.

This one: http://www.audiosciencereview.com/f...u-design-the-ultimate-magnepan.263/#post-8127

And don't fall victim to Don's modesty. What he posts, he knows.

Right and the corresponding impedance showing those resonances is where?
Have you forwarded my post to the good Dr yet?

Do you understand how impedance is measured with AP? Do you know the frequency resolution of it? Do you know what panel modes are that Don talks about? Or are you living only in the domain of woofers having low Q resonances at low frequencies that easily manifest themselves in impedance graphs?

 

[AJ Soundfield]

And don't fall victim to Don's modesty. What he posts, he knows.

I read his post.
You're quoting Toole/Olive with the paper, not Don.
Once more, have you forwarded the post to Toole and ask for an explanation as to why those aren't large diaphragm acoustic interference, but resonances?

 

[amirm]

I read his post.
You're quoting Toole/Olive with the paper, not Don.
Once more, have you forwarded the post to Toole and ask for an explanation as to why those aren't large diaphragm acoustic interference, but resonances?

You didn't answer anything I asked you. I am asking if you understood what Don explained about panel modes. And whether you understand the limitations of an impedance measurement as you put forward? Sounds like you don't in which case it is a big mistake in going by what you see in stereophile graphs.

As to going to Dr. Toole, I am not here to make your case. He wouldn't remain my friend long if I kept asking him what some random online poster is saying to challenge him.

Besides, I quoted you exactly what Dr. Olive said. Here is Dr. Toole saying the same thing: http://www.sonicdesign.se/tooleinw.htm

"The boxy sound of an enclosure is caused by acoustical and mechanical resonances. Good engineering can reduce these to inaudible levels. I remember, in a blind test of several years ago, listeners complaining of a "boxy" coloration from a certain product. It was audible to me as well, and moderately annoying. When the curtain was opened, the listeners saw that the boxy sound was coming from a large full-range panel loudspeaker. The resonance was in the panel - there was no box. It turns out to be difficult to design panel loudspeakers that do not have resonances. "

Very clear, right?

 

[AJ Soundfield]

Very clear, right?

No, not at all clear it isn't acoustic interference, instead of resonances/modes. Hence me prodding you towards hearing it from the horses mouth, instead or read/repeat.

 

[amirm]

Oh the "it is not clear" argument. Let me quote you something very simple and see if this is clear. From peer-reviewed, Journal of Audio Engineering Society paper: http://www.aes.org/e-lib/browse.cfm?elib=4445

Do you see how the term resonance is used synonymously with breakup modes that Don explained?

Still not clear?

 

[Thomas savage]

 

[DonH56]

Thanks Amir.

I was thinking of panel resonances, yes, not impedance variations in the sense of electrical impedance. And yes they are akin to cone breakup, the words depend upon what part of the world you were weaned... My grad acoustics textbook has pictures of modes in conventional and planar speakers. It was cool to see them as the B&W testing was a year or two before I took that class, and the book had been around much longer than that B&W experiment.

The correlation between impedance variations, frequency response, and diaphragm modes is complex and not 1:1. An impedance variation may reflect the crossover, port tuning, or speaker (driver) variations that can be voice-coil resonances, spyder/surround (mechanical) resonances, or modes in the diaphragms (and probably ten other things I haven't thought about or don't know). A lot of speaker designs assume a perfect voltage source driving them and thus (appear to) ignore impedance variations; the design target is flat (or whatever their target) frequency response at the acoustical output side no matter the electrical input impedance. Of course, that has led to some pretty nasty impedance plots...

I have rarely measured enclosure resonances and don't have any real comment on them. Accelerometers were rare and expensive back then and it was hard to rig a system that would capture enclosure resonance without being overwhelmed by everything else, like the speaker's output itself. I tried a pillow in front of the speaker, natch...

As for "improvement" in the image with room treatment (note I was not talking about panel resonances at that point), it was a combination of subjective and objective stuff. All IMO since I don't have the data to back it up anymore, just one of those things I know and it is up to you to believe or not. I claim no expertise as a speaker designer; the only successful one I ever did was a servo sub design decades ago. The objective improvement was a better impulse response and much less frequency variation measured by a mic at the listening position with a swept sine source from an ancient (now) HP (later Agilent, now Keysight) audio analyzer. The subjective improvement (or preference if you prefer) was that instruments centered in the stereo field no longer changed position as they played up and down the scale. Mono sources are great for testing things like this and we used some taped recordings of some of us playing our instruments in addition to commercial recordings. While I'd love to say trumpet was the clear favorite, in fact where we most noticed image "wander" was on a sax playing multiple-octave scales scale and some drum tracks. Similar things in "depth" were noticed but I didn't think of a good way to measure that back then (phase info would have made it clear but I plotted only magnitude, alas).

We took a pair of Maggies apart and played with their tuning "dots" and affixing the membrane at various points just to see the impact. I have little memory of the results and so have not discussed them. It changed things, of course... Measuring some of the big full-range panels (like Soundlabs) you could tell the struggles they had with modes and directivity; it is vexing when the image changes with frequency, or when panel modes create standing waves that emphasize frequencies unlike the source material. I did get to talk with Harold Beveridge ("Bev", always a shock to me that he would waste time with a snot-nosed college brat; Roger Sanders was another) at one point and he had a fair amount to say about it and how it drove his design.

Since I have no real expertise to share, at least the type AJ is looking for, and my experience is basically anecdotal, I'll bow out of this one. Better things to do... It's almost 3 pm here and I am still trying to finish lunch. Sometimes lab work can be a pain when the tests interrupt life (or at least lunch).

 

[AJ Soundfield]

Since I have no real expertise to share, at least the type AJ is looking for

Me neither, although as I stated previously, I stayed at a Holiday Inn Express last night.
Which is also why I wanted to hear an explanation from the horses mouth.
So while Amir hurls spaghetti at the wall, I emailed both of the good Dr's, Toole and Olive.
I'll let you know Don. Cool story about the B&W laser analysis. Guessing you have Maggies?

cheers

AJ

 

[amirm]

Which is also why I wanted to hear an explanation from the horses mouth.
So while Amir hurls spaghetti at the wall, I emailed both of the good Dr's, Toole and Olive.

Wonder how they resist the urge to whack you on the back of the head!

 

[Thomas savage]

Wonder how they resist the urge to whack you on the back of the head!

Would you propose this action might stimulate a audible cranial resonance?

 

[amirm]

Would you propose this action might stimulate a audible cranial resonance?

Given the ample amount of dead space in AJ's brain, for sure. Recommend some fiberglass to fill the many voids.

 

[TitaniumTroy]

Hey Don, any thoughts on Analysis Audio, and Apogee via Graz, vs Magnepan? Regarding overall sound quality, even if it's just a guess or just going by what others have said?

 

[AJ Soundfield]

Still not clear?

Oh I'm clear.
Dr Toole showed me the impedance measurement of the ML they tested (and quite a bit of discussion how they were measured, with averaging, etc, etc, etc.). Lo and behold, the wrinkle in the impedance that corresponded to the mode seen in the FRs was there.
So, I (we) quickly agreed he is correct, those are indeed "resonances" buried in the interference laden FRs. No question. Absolutely no doubt why those MLs faired very poorly in the listening tests, mono or stereo!
He also informed me that ML got a serious injection of engineering talent at some point...which brings us to newer stuff, like the Montis that I linked...without panel mode resonances. the measurements ther show acoustic interference common to all large full bandwidth diaphragms.
So Troy, to address your question, will some panel speakers exhibit resonances? Yes. But not all. Not sure what Amir said he heard at shows, but take it with a grain of salt unless there are corresponding measurements.
Lastly, MLs are not planars. Maggies are. Different kettle of fish, though similar somewhat. Both are largish dipolar line sources and as a result, will exhibit some similar characteristics. YMMV regarding sound preferences

cheers,

AJ

 

[DonH56]

MLs are planar speakers, but not planar-dynamic. Electrostatic field rather than magnetic field (though technically both use EM fields). Although you could argue the curvature makes them "non-planar". They act like line sources as the wavelengths get shorter than their physical dimensions. Which means from the lower midrange on up.

All panels will exhibit resonance modes, just as all cones will exhibit beak-up; the trick is to control and constrain them so they don't become overtly audible.

TitaniumTroy, I have never heard Analysis speakers, though I suspect their Apogee-like look is not a coincidence. Wisdom Audio is another dipole but I have not heard them; I think Amir may have.

Apogee's I have heard, long time ago, were very impressive but a very hard load. Mostly resistive, like Magnepan, but their pure ribbon design made for sub-1 ohm loads (later increased to an ohm or two, still awfully low). They suffered from modal distortion less than Maggies but I am not sure how much that impacted the sound people heard. They had lower measured distortion and hair better impulse response. I thought they sounded a bit better but can't be sure it was not my own bias. They did not sound better enough for me to spend the money on them, and an amp to drive them, but for a while they were king of the planar dynamic speakers.

Bob Carver's Amazing Loudspeaker used a similar structure but paired with conventional woofers. As much as I respect his designs, the Amazing just never hit a home run for me. It sounded great, yes, but integration was a bit tricky, it did not souind (or measure) as clean to me as the Apogee, and was not enough better than my Maggies to justify a swap. We had a pair in or store for a while but I never had them at home IIRC.

Eminent Technology is another more recent contender. I have heard the LFT-8, might have been the A version, but not the latest LFT-8B. I have never measured them. They use mid/tweeter drivers that look more like the old Infinity EMIM/EMIT drivers (I had an IRS2 for a while). I like their upper end, but frankly was underwhelmed by their bass. The hybrid design provides more and deeper bass than most Maggies but I thought it sounded a little boomy and "loose" compared to other speakers. I suspect a pair mated with my Rythmik subs would sound great.

I have listened to various Martin Logan and Soundlab speakers over the years in addition to Quads, KLH, Infinity Servo-Statik, Acoustat, Beveridge, etc. ESLs. ML had major integration issues IMO in the 90's that seem to have been resolved in more recent products, bearing in mind I have very little actual listening experience with them. The big Soundlabs always sounded very impressive to me but the deep bass seemed sloppy and loose and measurements confirmed that. However, I have not heard them in a while, and others with more recent experience have said they are much better, and/or I did not have adequate amplification (Mark Levinson and Krell mainly, briefly Threshold and Perreaux, but it was probably 90-something last I heard Soundlabs). My opinion of most ESLs has not really changed over the decades; gorgeous midrange, treble hard to drive and can get beamy/harsh, bass a mixed bag but panels (dynamic or ESL) and deep bass is a poor match for many reasons.

All IME/IMO - Don

 

[amirm]

Oh I'm clear.
Dr Toole showed me the impedance measurement of the ML they tested (and quite a bit of discussion how they were measured, with averaging, etc, etc, etc.). Lo and behold, the wrinkle in the impedance that corresponded to the mode seen in the FRs was there.

This is the reason I asked you if you know how impedance measurements are done and what went into the graph from stereophile. That highly smooth graph should have been a good hint. Using an audio analyzer for current measurement to detect small differences requires a different setup than just showing the large resonances for cone drivers.

 

[amirm]

So Troy, to address your question, will some panel speakers exhibit resonances? Yes. But not all. Not sure what Amir said he heard at shows, but take it with a grain of salt unless there are corresponding measurements.

Instead of being rude and and speculating, you can just ask me. Here is my observations that Troy was quoting from me: http://www.whatsbestforum.com/showt...pressions-of-Martin-Logan-Neolith-at-CES-2016

And this part in the specific:

"Now as it turned out, this was not such a bad thing as I heard what I could swear was resonances of the panel in one loud note. So I walk up to the speaker and put my hand on the surroundings of the panel. The amount of vibration transmitted to it was insane. At first I thought it was bass induced. So I put my hand on the box at the bottom. There was still significant vibrations there but at lower level. So clearly it was not the source.

I slid my hand on the panel surrounds and by that I mean the red part of the frame starting at the bottom going all the way to the top. The vibrations kept increasing. At the top of the panel frame, it was incredible amount. I actually thought the panel was moving back and forth. I am exaggerating but you get the idea.

In my view, this means that panel does not have sufficient structural support. It is exerting forces on that thin frame which due to being cantilevered, gets worse and worse. Every one of those vibrations acts as a dirty loudspeaker generating its own sound. "

Do you have personal experience with this speaker and can add something on that basis?