Loudspeaker sensitivity and dynamics: are the two inexorably linked?

@ditusa —

+1

@deep_333 wrote:

... Since high fidelity gobs of power is quite affordable these days, it is largely silly to chase high sensitivity speakers any more.

If only it was that simple. 

@bobbydd wrote:

I’m not referring to maximum loudness or volume, rather that the music sounds less reproduced and more that the instrumentation and vocals are more real sounding through higher sensitivity speakers.

Is this a real phenomenon? Or is it more the particular gear I’ve experienced?

Oddly it’s rare to read such fine expressions uttered here, what sounds "less reproduced." It says a lot without stepping into the realm of pretending what’s heard is a facsimile of a real, live acoustic event, and yet it’s at the heart at what can be more readily offered with the attainment of certain physical attributes of a speaker, of which dynamic capabilities are a core aspect and intricately linked to both high sensitivity and prodigious air radiation area.

It’s also about how one assesses and is habitually exposed to ’dynamics;’ I’ve heard quite a few low eff. speakers that, on the face of it, sounded rather dynamic, but when compared to larger and more dynamically capable speakers (because such, factually, there are) it becomes obvious that the latter is somewhat more relaxed yet visceral, effortless and "liquid" sounding in its dynamic portrayal, which to me can be condensed into a more singular impression as a "less reproduced" presentation.

I’ll concede to poster @mijostyn’s findings on at least very large, high-passed (and properly subs augmented) ESL’s that can be dynamically astute, but they also have plenty of displacement to yield while being transiently excellent (with narrow dispersion) - a powerful combo on top of being a crossover-less speaker plane. I do believe the Soundlab’s aren’t that inefficient but rather in the 90dB range? So hardly a typical representative of a low eff., direct radiating speaker - of limited size, no less.

@timlub --

Good post, in which you did however only touch upon the amp side of things and not the speakers, which with low efficiency in particular will potentially further complicate issues with thermal compression/modulation as heat build-up in the voice coils, even if it’s only for short periods of time as a rather "dynamic" phenomena.

I might also add that removing the passive crossover between the amp and drivers for active configuration will have the amp(s) seeing a much easier load, with the better driver control and sonic benefits this entails, while also thriving on load independency between the amp channels coupled to their respective driver segments and limited frequency spectrums. So, a combination of high efficiency, good power handling, active config. and sufficient quality power will be among the core parameters to ensure the most optimal outset for dynamic prowess, as well as other aspects in sound reproduction.

As the article supplied by poster @ditusa points to, not everyone is in the need of the same effective headroom (if there even is any to speak of) for their specific requirements and setup context, and so to their needs may be dynamically well served with smaller, low efficiency speakers. That’s however also a clear indication of what's relative to the individual benchmark one sets about to work from, and the importance that is assigned to this specific area.

Tom Danley on headroom and power compression (excerpt):

Well before a speaker burns out, it is severely compromised in it’s performance.

With the heating of the voice coil, one finds the SPL decreases relative to the expected level with increasing power, also the systems tuning / frequency response changes at the same time, for the same reason.

For modern drivers, this power compression begins about 1/8 of the drivers rated power, if rated using the AES procedure.

Honestly there is so much BS regarding specs and such in commercial sound that to add a sense of realism or something to that mess, we have a 3rd party laboratory specify usable rated power. Hifi, don’t get me started.

https://www.avsforum.com/threads/danley-dts-10-super-spud-diy-kit.1189404/post-17409024

@timlub wrote:

I was really trying to deal with the idea of dynamics only. I did bring up efficiency a few efficiency issues, but tried to keep it in an understandable format.  To those that don't have a better idea of sensitivity vs efficiency.  The short answer:

Sensitivity measures the volume out vs the power in.  Example 90db output with 2.83v input or 90db output with 1watt of input @ 8 ohms. 

Efficiency measures the amount of output vs the amount of input in power or how much power is lost due to heat dissipation.  Example would be 100% of power in, 90% of power out, 10% loss.  

I'm fully aware of the distinction between sensitivity and efficiency. Now and then may use both terms in the same post, but without getting into numbers I take most get the basic idea on how high sensitivity and high efficiency relate in opposition to low one and the other. 

I tried to express in simple terms what for most is the culprit when it comes to dynamics and transparency.  We can discuss crossovers or thermal compression in another thread

I fail to see the need for that when what's brought up is interconnected. 

@tvrgeek wrote:

Thermal compression happens to all drivers, high sensitivity or low. How much depends on how well the motor can dissipate heat.

It happens to all drivers regardless of sensitivity, yes, but obviously at different stages; all things being equal a 10dB gain in sensitivity means a given SPL has less heat build-up in that voice coil than the lesser sensitive variant being it handles 10x less power. To boot the more sensitive driver, typically a pro segment ditto, has better power handling with a larger diameter voice coil and usually better gap cooling, so it’s not even an "all things being equal" scenario but rather one that generally favors the more sensitive driver as well with regard to heat dissipation.

For example, fero-fluid may dissipate heat better and have lower thermal compression ( re increasing with temp) but it is not without other issues that can effect perceived dynamics.

True, on both accounts, but here the basis of comparison is other tweeters with similar-ish sensitivity, where ferrofluid-using variants may see an advantage in regards to power compression "fatigue" for a given SPL. (At a panel discussion some 30 years ago with the late Hother Bak of Dali, he argued ferrofluid could have resonant damping qualities in the voice coil, but they ultimately worked towards avoiding ferrofluid in their tweeters due to, as you put it, "other issues")

Same goes for the flexibility of the suspension, that can actually get "looser" with heat for less compression. Drivers are a combination of parameters, some working in opposite directions giving a better all things considered performance than any one part looked at in isolation.

This is an interesting field of discussion. Question is if there’s basis for speculating into mechanical "compression" at lower SPL’s with more stiffly suspended pro drivers. They are, some of them, thought to be less adept sonically at lower volumes, only to open up with more prodigious SPL’s - which is their intended target range anyhow. Myself I’ve invested some effort into selecting my particular pro cinema speakers and knowing their drivers in these regards, but usually high efficiency speakers are known for their great low-SPL capabilities, coming-to-life and overall attentiveness here. Some guys, like Devon Turnbull of Ojas, are keen on selecting rather low power handling, high efficiency pro driver designs of yore with their lighter voice coil assembly, overall moving mass and higher compliance, matching them up with low powered and topologically simple tube amps. The rationale likely being (if it doesn’t simply come down to a preference based on listening only): why all that power handling (and amp power capacity) from pro drivers with their high efficiency for use in a domestic environment? Devon finds some of the modern pro drivers "lethargic" sounding compared to the older designs he prefers, and while in certain contexts at least his views here are not without merit, I don’t entirely agree with him.

However, all things considered, thermal compression is most likely a wash between technologies. Other factors already mentioned ae greater.

It’s not that complicated. Pragmatically speaking sensitivity and power handling are the primary factors, and ones that have direct implications on dynamic capabilities. And as already mentioned, passive crossovers is a factor as well.

@britamerican wrote:

I think you are barking up a tree there @alexberger . No way crossover parts are heating as much as voice coil. Inductors probably not at all and what are we paying all that money for fancy resistors for??

The relevance here, coming down to practical usage, is that passive filter components heat up sufficiently to impact filter parameters, thus affecting the sound of the speaker to deviate from its intended design goals at changing, progressive SPL's. Surely this is not desirable, and that's just the workings within the crossover itself, apart from the fact that it's a "roadblock" by its mere existence; it effectively acts as a bottleneck between the amp and speaker, presenting a more challenging load to the amp that then has negative repercussions in its handling of the speaker and its drivers. Throwing gobs of money at fancy components doesn't really "add up" to anything, but rather potentially minimizes a nuisance. 

@steve59 wrote:

Paul Klipsch gets quoted for saying he just wished somebody would make a quality 5 watt amplifier, (hopefully I didn’t butcher that quote too bad) I’m not sure how the math works, but I would think the difference in sensitivity can be evened out by using more powerful amps to compensate. Say, a pair of kilowatt amps for 86 db speakers could have the same dynamics as a horn design being driven by your favorite 12 watt tube amp?  Having no educated qualification I would think heat would have to be addressed when designing for maximum spl? 

Say you have properly sized all-horn speakers with a sensitivity sitting at no lower than ~105dB's (likely higher from the mids on up). That's a some 20dB discrepancy in sensitivity here compared to your 86dB speaker example. 12 watts on those horns, in theory, should be SPL-equalled by shoving ~1.2kW's into the low eff. speakers. Practically: good luck with that. Those horns by and large will likely cruise fairly effortlessly along with 12 watt peaks, whereas the low eff. speakers fed with +1kW's will be at the end of their ropes (if they haven't already passed unto those eternal audio fields) with power compression screaming its ugly face in heavy measure.

Look at the Tom Danley quote above - if power compression starts creeping in at 1/8 the rated AES power handling it suddenly makes sense to work towards maintaining prodigious headroom. That is, how to achieve any notable headroom with 86dB sensitive speakers and a fair max. peak SPL requirement at the LP of, say, 105dB's - without any effects of beginning power compression? The speakers would have to take no less than 1-1.5kW rated power just to avoid the beginning effects of power compression. Good luck with that, not to mention the amount of clean amp power that would require - with plenty of power to spare. 

Many get by with much less power handling and capacity while feeling no shortage of dynamic prowess at the LP with lower eff. speakers, and if their max. SPL requirement is no higher than in the 90-95dB range it certainly lessens the need for higher sensitivity, power handling and capacity. That however is not to say more headroom won't make a difference here, certainly to those who knows the difference it can make, and to whom it's an important and desirable trait.