The two most common mistakes are bass and treble

Within a specific power envelope ?

The other problem you deal with is of course that you get thermal compression if the speaker lacks efficiency. You can’t solve this by getting a more powerful amp

No, you can solve it by various aspects of speaker driver and crossover designi. I don‘t think anyone could state that the ATC SCM50 (85db/W/m) suffers unduly from thermal compression. There are many other manufacturers who are cognisant of the impacts of thermal compression and design their drive units accordingly.

We make amps from 30 to 500 watts. And certainly, anyone *could* state that the ATC SCM50 (85db/W/m) suffers unduly from thermal compression.


You can't solve thermal compression via crossover design or internal amplifiers for the loudspeaker system. Thermal compression is a function of the voice coil itself.

Rather than explain how this works I've linked a Wiki page below.


This is not mysterious; thermal compression has been a known thing for a long time. https://en.wikipedia.org/wiki/Power_compression

@clio09 I wasn't really saying you were wrong, nor did I feel that you were saying that I was either, but I felt from your comments that I should flesh some things out a bit more. I agree with everything Roger states from your comment above. Despite conversations like this nothing is going to change; although people reading this might be a tad more careful.


Older solid state amps tend to run too little feedback which is why they can sound bright and harsh. Driving higher impedances helps with this simply by reducing distortion, which is the source of the brightness.

I think Roger didn't like that we run no feedback in our OTLs but the reason was I knew that if we used feedback, the phase margins of the amps would never permit enough to be applied. That's actually true of every tube amp. So to prevent the dreaded brightness coloration we ran no feedback at all and relied on other means to keep distortion at bay.

To Atmasphere: I’m on my third pair of low impedance, super-inefficient Magnepans. I’ve had many other brands of speakers - Altec, Allison, Advent, AR, B&W, Definitive Technology, Elac, KLH, Klipsch, and Wharfdale.
In my opinion, Magnepans are the most neutral speakers made, as well as fantastic bargains.

You can imagine with Manapan in my town and being on a first name basis with the owners (one of them called me last summer when my stolen bicycle turned up on Offerup, for which I am grateful) that I’ve heard them a lot . ’Some of my best friends own them’ although a trope is actually also true. They are good speakers no doubt. Imagine though what they might sound like if higher impedance: They would be less sensitive to speaker cables. The amps playing them would have less distortion. In a nutshell, you’d experience more impact in the bass (owing to the speaker cables not robbing you of impact) and smoother, more detailed sound through the mids and highs (owing to lower distortion from the amps).

I think @erik_squires is correct that planars suffer less thermal compression than lower efficiency cone drivers.

In my collection, my best speakers are not the high efficiency Audio Note AN-e or my DIY single driver using an Audio Nirvana driver, they are my Spendor 1/2e, Quad ESL, Acoustat Model 2, and Analysis Audio Omega. None of which should be confused with high efficiency.

@clio09 Its always important to be careful about conflating personal anecdote with how things are: just because you have one efficient speaker that is bested by others of lessor efficiency does not mean that what I’m saying is incorrect. I have a similar anecdote- a set of desktop Coral loudspeakers that I bought at a garage sale for $5; they are 98dB but sound pretty mediocre; the cabinets resonate and they lack definition and bandwidth, entirely because they simply weren’t built to be anything other than cheap. Obviously the Audio Notes are not that but you get my point.


The Quads and Acoustats are not as inefficient as they appear on paper; generally speaking add 6dB to any planar’s measured values to sort out where they actually sit (some amps have trouble driving ESLs which is unrelated to their efficiency). This is because when measured, the microphone is placed at 1 meter. At that distance, a good deal of the output of the speaker is not picked up by the mic. You have be back from the speaker a good 10 feet to really understand its efficiency. ESLs do not suffer thermal compression as they have no voice coils.


Efficiency and impedance are important, but like class A vs AB in amplifiers, a lot has to do with execution. As an amplifier manufacturer that makes both class A tube amps and class D amplifiers, I see how distortion plays out in systems all the time. In solid state, you lose power as the impedance is increased but you also lose distortion. Now if sound quality is your goal then your amplifier dollar will be best served be a speaker of higher impedance; if *sound pressure* is your goal then you have a weak (3dB) argument for 4 ohms.


Higher efficiency speakers require greater precision in their construction, which makes them a lot more expensive. They were the only game in town when tubes were King; speakers had to make the most of amplifiers whose power is expensive. OTOH when solid state became practical, amp manufacturers realized the benefit of no output transformer and no filament circuit- so they could make an amp of the same power that cost 10% of the cost of a tube amp and sell it for 90% of the price of that tube amp. Speaker manufacturers saw this and realized that if they went to 4 ohms and less efficiency the lower cost to build the speaker meant they could make more money too. Solid state amplifier power is cheap, so this has worked great market-wise but its been one of the barriers between sounding like real music as opposed to a really good stereo.

Not saying you are incorrect, but give me an idea of a couple of common inefficient speakers, and a couple you think do a great job.

What is your cut-off for efficiency and impedance, more or less? What is good, and what is bad?

What a joke about not purchasing speakers with sensitivity ratings lower than 90 or lower than an 8 ohm load. In this 1 sentence, you pretty much eliminated every "best" speaker on the market. Most of these "best" speakers have sensitivity ratings in the middle to upper 80's and some go down below 4 ohms (some in the 2 ohm range). All this means, if you buy the best speaker, you also need to buy a better/best amp to drive them.

A speaker that is low efficiency (less than 88db if 8 ohms) is simply not 'the best' plain and simple. It might be very good, but not the best. Its a common misconception that when the speaker is inefficient it trades off efficiency for transparency. This idea is simply false. One of the most transparent speakers I've heard (and this is compared to Wilson, Magico and others) is 98dB (and also is flat to 20Hz), made by Classic Audio Loudspeakers. Transparency and efficiency are unrelated.

IMO/IME when the speaker gets below 89 dB you are into a situation known in the radio world as 'gold plated decibels'. You also have to pay attention to impedance; if the speaker is 89dB and 4 ohms its actual efficiency is 86dB, because 2.83 volts into 4 ohms is 2 watts, not 1. Think about it this way: in most average rooms to work with a speaker that is 89dB and 8 ohms you'll need about 250 watts to really play whatever you want without clipping the amp. But if the speaker is 95dB you'll only need 60 watts to do the same thing. Its far easier to find a good sounding 60 watt amp than it is to find one that makes 250 watts. As the power goes up, the field narrows dramatically!


Amplifiers might be able to drive 4 ohms and less, and they might be able to double power well below that, but because they can do that isn't the same as saying that amplifier is sounding its best. This is all about distortion: its distortion that causes us to hear differences between amplifiers. The brightness coloration of solid state is caused by distortion- the higher ordered harmonics, even though the THD might be only 0.01%. The ear uses the higher ordered harmonics to sense sound pressure and so is keenly sensitive to these harmonics.


Its a simple fact that all amplifiers have less distortion into higher impedance. You can see it in their specs and you can hear it as well. If you want a solid state amp that suffers from brightness to sound a bit smoother and more detailed, have it drive a higher impedance.


Not all solid state amps have brightness- this coloration occurs when the amp had insufficient feedback which is describing most of the amps built in the last 70 years. You can get away from this by having no feedback at all, or you can get around it by having really a lot. But its important to understand that you can't just go out and 'buy a better amp'; there are not that many really powerful amps that sound like music.


The other problem you deal with is of course that you get thermal compression if the speaker lacks efficiency. You can't solve this by getting a more powerful amp, it gets worse the more power you put into the speaker.

P05129 that’s what I mean- so many people try to speak with authority and absolutes but they don’t know jack other than what they know and like.

I apologize if I sound like this- I'm coming from the perspective of engineering and as an amplifier manufacturer of the last 45 years.

Hey Ralph, I finally got the speakers. Roger has labeled them 645-8's. I put a picture of them on my system page. I have a new Cosmos coming from Sota and a Schroder CB to put on it following which it will be time to think about amplifiers:-)

@mijostyn Sound Labs, like most speakers with rear-firing information, need at least 5 feet from the wall behind them. I can't tell if that is what you have or not.

Inefficient and low impedance speakers are IMO/IME a mistake.


All amps make more distortion playing into lower impedance. If you want the amp to sound smoother and more detailed, set it up to drive a higher impedance.


Inefficient speakers need really powerful amps to play properly, and they suffer thermal compression. Really powerful amps that actually sound like real music are not that common and likely pretty expensive, when the whole thing could have been solved without any loss of bandwidth or resolution by simple getting a speaker that's easier to drive.


Speaker cables get far more critical on low impedance speakers too. You can easily spend $$$$s on speaker cables, but if you have a higher impedance speaker the cables simply won't bring as much to the table.