[mobbaddict]

I was looking for concrete information about the advantages of multi balanced armatures designs for IEMs versus single armature ones when I came across this video posted by Knowles, which explains a lot of things:



I thought I would share it here since it's very educational. The video contains a lot of interesting inputs on the question:

• Etymotic ER4S are still the benchmark for neutrality in a single armature design
• AKG N5005 are apparently the best matching IEM for the Harman target curve
• Knowles' main argument for multi BA design is lower THD
• Other arguments are related to specific preferences/purposes (so non applicable for a reference sound)
• A lot of other design factors can impact sound quality such as dampers, tubing, venting, etc...

I'm very interested to hear other potential arguments for multi BA designs. Does a lower THD say anything about a potential superior resolution?

 

[KeithPhantom]

I'm very interested to hear other potential arguments for multi BA designs. Does a lower THD say anything about a potential superior resolution?

Harmonic distortion has to be pretty bad to be easily heard in most of the frequencies. With transducers, it is pretty hard to hear because it is masked by the signal, being other types of distortion more important (IMD and crossover). Everything under 1% should be fine for music listening. THD is important, but, it isn't everything that defines fidelity, things such as FR are even more important when defining this term.

 

[castleofargh]

IMO the main argument for multidriver is that it makes tuning a given FR much easier. I think, that alone would be enough to convince most manufacturers to go for it.
As shown in the video, THD will usually change with SPL. It's all a matter of how loud something will be played and when the distortions will become noticeable.

I'm not a fan of multi BA IEMs, but it's mainly because some tend to be huge and I love tiny IEMs. And because I don't want my frequency response to be at the mercy of the amp I use, which is even more likely with multi driver designs. I happen to use a bunch of different devices depending on circumstances, so I much prefer an IEM with a pretty stable impedance curve if I have a choice.

Just like I avoid very sensitive IEMs nowadays so that slightly noisy amps won't bother me too much. Meaning that stacking multiple driver for the same crossover section to lower distortions and increase max SPL, is usually a bad thing from my specific point of view. It's objectively good for the IEM, but if it makes half my amps unusable because of how loud the background hiss becomes, or because the impedance goes too low for those amp designs and they hiss even more or distort like crazy or whatever, then it's not an objective improvement for the complete playback system.
Obviously I try to follow those as vague rule of thumb, but in practice it's hard to give a general position. Of course I have some sources that have low enough impedance and low noise, that will handle most IEMs just fine. And of course there are some single driver IEMs with stupidly low and sometimes widely changing impedance over freq. Some can have really high sensi too. So people should probably shop for their own needs.

Otherwise, I'm of the opinion that tuning is tuning. EQ, dampers, vents, creating specific acoustic chambers inside the IEM to get specific resonance, using crossovers, using tubes of different lengths to adjust phase, etc, I'm fine will all of them so long as the result isn't dumb. TBH I never understood why someone would be totally against any given option if it helps reaching a certain goal.
Have I contradicted myself enough? ^_^

 

[mobbaddict]

Harmonic distortion has to be pretty bad to be easily heard in most of the frequencies. With transducers, it is pretty hard to hear because it is masked by the signal, being other types of distortion more important (IMD and crossover). Everything under 1% should be fine for music listening. THD is important, but, it isn't everything that defines fidelity, things such as FR are even more important when defining this term.

That's what I would assume indeed...

IMO the main argument for multidriver is that it makes tuning a given FR much easier. I think, that alone would be enough to convince most manufacturers to go for it.

Yes that seems to be the main argument. And it does explain why Etymotics never felt the need to go for multi BA designs since they're all about reference sound. One might argue that the single driver design is also part of their marketing positioning and that they probably have nothing to win competing with the hundreds of multi BA designs available on the market.

I think one similar example was Phonak which launched a similar single design 10 years ago that competed with the ER4 series, then launched a dual driver design that took a completely different approach with a V-shaped signature. I would Etymotics to take a similar move if they ever went for a multi BA design.

I'm wondering how something like the PP8 competes with Etymotics, since they have a similar neutral approach but use 8 armatures.

 

[castleofargh]

A few manufacturers are against crossover. Ety is among them. Or at least was, as some things have changed since the brand was sold. So IDK what the future holds.
Sennheiser was also pretty vocal about not using crossover. But then they're also not big on BA drivers ^_^.

I don't think it's hard to objectively outperform an er4. In some ways, other Ety products already do it, but they rank their products in term of how close to their target FR they get, while I don't.

 

[Blackwoof]

Yes that seems to be the main argument. And it does explain why Etymotics never felt the need to go for multi BA designs since they're all about reference sound. One might argue that the single driver design is also part of their marketing positioning and that they probably have nothing to win competing with the hundreds of multi BA designs available on the market.

People said the ER4S can't do bass because of it being single driver, Yet the ER4XR has 3db more bass than the SR & the ER3XR even higher. I've seen people get ER4SR's that rate 0.4% at 97db from the white sheets etymotic give in the ER4SR/ER4XR box. The above THD comment is pretty much why the ER4XR average 0.8 ~ 1.35% at 97db.

Those same sheets also show the ER4 stop at 17KHz before rolling off, Meaning it had more bandwidth than most multi BA set ups. The ER4PT/ER3SE(No EQ) sound brighter than the SE425 i had.

Edit; To add more 17KHz seems to be from tests that the adult hearing limit. And the ER4SR/ER3SE can handle +5 ~ 20db bass boost like a planar if the shelf is 100Hz, Which seems to be from them really deep fit having less SPL air volume than a Headphone/Speaker would. Which i think Etymotic here in 2004 said the same with the older Etymotic ER4S, Want bass? use a low shelf EQ.

 

[dazzerfong]

Those same sheets also show the ER4 stop at 17KHz before rolling off, Meaning it had more bandwidth than most multi BA set ups. The ER4PT/ER3SE(No EQ) sound brighter than the SE425 i had.

If you want to read into it, Etymotics measures only up to 12.5 KHz. https://www.etymotic.com/technology/hwmra. Got a source on your 17KHz? Interested to see which GRAS they used to achieve it.

Also, Shures are notoriously dark and bassy, especially their lower BA stuff. If you're hearing 'bright', it's not going to be 10 kHz+: it'll more be around 7-10 kHz. After that point, it's less heard and more felt.

I'm not a fan of multi BA IEMs, but it's mainly because some tend to be huge and I love tiny IEMs. And because I don't want my frequency response to be at the mercy of the amp I use, which is even more likely with multi driver designs. I happen to use a bunch of different devices depending on circumstances, so I much prefer an IEM with a pretty stable impedance curve if I have a choice.

Just like I avoid very sensitive IEMs nowadays so that slightly noisy amps won't bother me too much. Meaning that stacking multiple driver for the same crossover section to lower distortions and increase max SPL, is usually a bad thing from my specific point of view. It's objectively good for the IEM, but if it makes half my amps unusable because of how loud the background hiss becomes, or because the impedance goes too low for those amp designs and they hiss even more or distort like crazy or whatever, then it's not an objective improvement for the complete playback system.

The annoying thing is that it seems to be only the expensive multi-BA designs that address this crossover issue. That being said, if done correctly, it works well: I've seen properly-compensated crossover networks not skip a beat even with a 75 ohm output impedance.

 

[Vamp898]

I was looking for concrete information about the advantages of multi balanced armatures designs for IEMs versus single armature ones when I came across this video posted by Knowles, which explains a lot of things:



I thought I would share it here since it's very educational. The video contains a lot of interesting inputs on the question:

• Etymotic ER4S are still the benchmark for neutrality in a single armature design
• AKG N5005 are apparently the best matching IEM for the Harman target curve
• Knowles' main argument for multi BA design is lower THD
• Other arguments are related to specific preferences/purposes (so non applicable for a reference sound)
• A lot of other design factors can impact sound quality such as dampers, tubing, venting, etc...

I'm very interested to hear other potential arguments for multi BA designs. Does a lower THD say anything about a potential superior resolution?

You can do things that would otherwise not be possible.

You can use different materials for a given FR range, for example Sony uses an physical LPF for the Woofer and an magnesium diaphragm for the Super Tweeter.

So the Woofer can produce an bass that pretty much can't be distinguished from an DD, but it can only do this, it's incapable in doing high frequencies.

On the other hand, the magnesium diaphragm Super Tweeter could never produce such an bass, but delivers an almost flat frequency response up to 40khz

There is an easy to understand picture here about the (now antique) XBA-1/2/3/4