Distortion in loudspeakers

[DonH56]

The problem with dipoles' rear wave is not just the delayed sound, it is also the comb filter effects.

 

[March Audio]

So...

Music is distortion. Musical instruments get their "tone" from varying levels of harmonic distortion.

How to measure whether what we play is distorted is not so easily determined as the basic THD measurement.

IMD uses two sine tones.

THD + N (noise) adds calculation for sound that is not in the harmonic series.

So, make of that what you want, it's something I know.

I'll show something else in a minute.

.....mmmmm..... that's great but that doesn't mean THD is not a useful tool. Just because music is not pure sine waves doesn't make analysis of component harmonic distortion invalid. A component shouldn't generate additional harmonics that are not in the original signal.

The question is how this is perceived by the subject, how significant it is and what levels become noticeable.

 

[watchnerd]

Why even talk about THD when IMD is the real bad guy?

 

[Blumlein 88]

I have 2 sets of Martin Logans, and 3 different dynamic speakers lying around (2 sets of Dynaudios, both active and passive, one set of little JBLs actives), so this is not coming from a lack of comparisons.

I didn't say dynamic limitation (somebody else did), I said somatic feel.

I'll pick a classical example, since that's your genre:

If I play Dvorak's 9, New World, 4th movement, I get more of a somatic effect from dynamic speakers than I do from my ML's.

The ML's have a bigger, more lifelike soundstage, but the dynamic speakers have more impact in the orchestral power zone (upper bass / lower mid).

Just posting off the top of my head. I have never decided which it is. Subjectively the panels sound like what they are in the upper bass lower mids which is soft and billowy lacking genuine firm impact. When considering the lower frequencies are partly cancelling each other out it sort of makes sense. The harder they move air the more the front and back wave cancel. So they would be soft in a sense. On the other hand, that type speaker with the cancellation of dipole nature does not excite half the room modes. Are box speaker monopoles just exciting more room modes giving a heavier thump and thwack to the low end? Or some mixture of both effects?

 

[RayDunzl]

Why even talk about THD when IMD is the real bad guy?

IMD 83 dB listening position

MartinLogan

JBL LSR 308

 

[Blumlein 88]

IMD 83 dB listening position

MartinLogan

View attachment 5326

JBL LSR 308

View attachment 5327

I notice the 1900 hz value is very low on the ML and about 20 db higher on the 308. I believe that would suggest a bit of asymmetry on the part of the 308 that is missing in the ML.

 

[RayDunzl]

Are box speaker monopoles just exciting more room modes giving a heavier thump and thwack to the low end?

I don't know, but in this room, I think I measured (a while back) about a 9dB difference between the direct and reflected sounds between panels and monopoles. Here's a quick sweep right now:

The ML dipole (blue) excites a reflection off the front wall at 7ms, and the front wall again at about 25ms. Other reflections are quite low in amplitude.

The JBL monopole (red) excites many reflections.

The 1~2ms stuff is the couch the microphone sits upon.

 

[RayDunzl]

The problem with dipoles' rear wave is not just the delayed sound, it is also the comb filter effects.

Maybe so, but I don't see where that puts them at any particular disadvantage. Just, different.

MartinLogan (blue) and JBL (red). dipole effect prominent at 220 and 330Hz (I think).

 

[Blumlein 88]

I was looking at the fact F2 plus the difference (500 hz) is 1900 hz. So this is one of your lower level sum and difference frequencies. When you don't see a similar level at F1 minus the difference which is the 400 hz level it usually is due an assymetry in the driver which isn't terribly surprising for a cone in a box.

 

[watchnerd]

Just posting off the top of my head. I have never decided which it is. Subjectively the panels sound like what they are in the upper bass lower mids which is soft and billowy lacking genuine firm impact. When considering the lower frequencies are partly cancelling each other out it sort of makes sense. The harder they move air the more the front and back wave cancel. So they would be soft in a sense. On the other hand, that type speaker with the cancellation of dipole nature does not excite half the room modes. Are box speaker monopoles just exciting more room modes giving a heavier thump and thwack to the low end? Or some mixture of both effects?

Here's my completely out-of-my-ass amateur hypothesis:

Real life instruments with a lot of energy in the upper bass / lower mids (bass, drums, timpani, lower piano registers, etc.) do excite room nodes in real life. We also experience (at least sometimes) those instruments not just as sound, but as physical experience - surfaces can vibrate and resonate around us, in addition to the room nodes. If we were cats, I'd say we experience this with our whiskers.

Thus, the monopoles excite more room nodes, it matches closer to our real life expectation.

In contrast, instruments that are in the upper mids / treble are things we "hear" as opposed to "feel", and they lie within our most sensitive hearing range, and also in the area where our lizard brain impulse response sensitivity to snapping twigs is most acute. In this area, electrostats, with their faster impulse response and lower inertia can sound more realistic.

So here's the weird paradox in my SWAG hypothesis:

If I'm correct, the whole hybrid electrostat hierarchy (pay more money, get a bigger panel with lower crossover to the woofer, more full range, and a bigger woofer) is wrong. Instead of trying to make panels bigger to make them reach lower, we should use them from the midrange on up, maybe even in a 3-way system.

 

[RayDunzl]

With no smoothing, maybe the dipole comb is more evident - the 220 330 440 580 series - but so narrow...

I don't see any advantage to JBL here...

 

[RayDunzl]

Instead of trying to make panels bigger to make them reach lower, we should them from the midrange on up, maybe even in a 3-way system.

Mine is a two way cross to a 12" sealed woofer at 180Hz, and I've added the four 15" CheezeWoofers to move more air down low low.

 

[watchnerd]

Mine is a two way cross to a 12" sealed woofer at 180Hz, and I've added the four 15" CheezeWoofers to move more air down low low.

I've never asked, but is there a high pass filter between the CheezeWoofers and the ML's? Or are the ML's still running full range?

 

[Blumlein 88]

Here's my completely out-of-my-ass amateur hypothesis:

Real life instruments with a lot of energy in the upper bass / lower mids (bass, drums, timpani, lower piano registers, etc.) do excite room nodes in real life. We also experience (at least sometimes) those instruments not just as sound, but as physical experience - surfaces can vibrate and resonate around us, in addition to the room nodes. If we were cats, I'd say we experience this with our whiskers.

Thus, the monopoles excite more room nodes, it matches closer to our real life expectation.

In contrast, instruments that are in the upper mids / treble are things we "hear" as opposed to "feel", and they lie within our most sensitive hearing range, and also in the area where our lizard brain impulse response sensitivity to snapping twigs is most acute. In this area, electrostats, with their faster impulse response and lower inertia can sound more realistic.

So here's the weird paradox in my SWAG hypothesis:

If I'm correct, the whole hybrid electrostat hierarchy (pay more money, get a bigger panel with lower crossover to the woofer, more full range, and a bigger woofer) is wrong. Instead of trying to make panels bigger to make them reach lower, we should use them from the midrange on up, maybe even in a 3-way system.

That was basically Mr. Sanders idea, and he is not alone in that. Makes perfect sense, except those have been the least satisfying panels I have heard. Never gave all that much thought to why. In the ML line the wider panels have sounded better to me than the narrower panels. That is nothing definitive. It may just mean no one has done it right as it does make plenty of sense. Of course I did have an active DIY friend who thought the best rig would use huge panel woofers (big enough cancellation of front to back happened below 50 hz) coupled with very nice controlled directivity horns for mid and treble. Minimize the room influence and maximize what was on the recording to our ears.

 

[watchnerd]

IMD 83 dB listening position

I don't know how far you are from the speakers -- what do you think the SPL is at 1 m?

If I think about IMD in a speaker, I would expect it to get hyperbolic towards max volume as vibration, thermal compression, crossover distortion, etc kicks in.

 

[RayDunzl]

I've never asked, but is there a high pass filter between the CheezeWoofers and the ML's? Or are the ML's still running full range?

The ML are are running full range.

The CheezeWoofers have a lowpass, and for sure start dropping around 90/100Hz (forgot to mention they have a low-pass cutoff knob when I described them elsewhere).

I balanced the woofer and each of the subs to play at the same SPL level (without DRC), and then use DRC to knock down the lows to be flat like the rest of the spectrum .

They're stacked next to the mains currently, so, it's a duplication of effort that results in lower distortion in the bass.

I have a phase anomaly around 45Hz, somehow the room puts the waves at the listening position 180 degrees out of phase between the left and right. The room is rectangular but the left rear third has no walls and open to the dining/kitchen area.

Gold - subs + mains stereo
Green - subs only - stereo
Read and Purple - left/right subs only mono

You can see the bass is flat individually at 45Hz, but when combined, dig a hole. It isn't interference with the mains.

Distortion for the Gold trace (mains plus subs stereo). It measures lower distorion if I do steady tones than using a sweep, but 2% down low isn't bad.

 

[watchnerd]

That was basically Mr. Sanders idea, and he is not alone in that. Makes perfect sense, except those have been the least satisfying panels I have heard. Never gave all that much thought to why. In the ML line the wider panels have sounded better to me than the narrower panels. That is nothing definitive. It may just mean no one has done it right as it does make plenty of sense. Of course I did have an active DIY friend who thought the best rig would use huge panel woofers (big enough cancellation of front to back happened below 50 hz) coupled with very nice controlled directivity horns for mid and treble. Minimize the room influence and maximize what was on the recording to our ears.

In my case, I think I'm going to take the third route, that I've yet to explore: compression drivers + waveguides.

On paper, compression drivers can have distortion specs much better than direct radiators (not as good as stats, though), but with higher SPLs than either. I'm hoping the new 7-series JBLs give me a bit of the best of both worlds.

 

[RayDunzl]

I'm hoping the new 7-series JBLs give me a bit of the best of both worlds.

The cross is at 1.9khz on the 705p and 1.7kHz on the 708p.

On the 4367 it is 700Hz, and on the M2 800Hz.

I like my 180Hz cross.

 

[watchnerd]

The cross is at 1.9khz on the 705p and 1.7kHz on the 708p.

On the 4367 it is 700Hz, and on the M2 800Hz.

I like my 180Hz cross.

I'm not worried about the 1.9khz on the 5" of the 705P.

I am a bit worried about the 1.7khz on the 708P because it's right at the edge of the beaming frequency for an 8" driver.

 

[Cosmik]

On the general theme of speaker distortion, I have seen mentioned elsewhere mechanical "hysteresis" and "memory" effects of the surrounds and spider. These have been raised as possible reasons why amplifier damping factor is important, not just for frequency domain reasons, but actual distortion. Does anyone have a view on this?