Sensitivity - is there a trade-off?

[Muzz]

Much as the title suggests, does anyone (and I'm thinking of our more technically adept contributors here) know if a higher sensitivity speaker has a trade-off in another area? This is prompted by my consideration of a replacement speaker for a 12" combo (which has good feedback from folk who've done it, so I'm not fretting too much about the actual speaker's suitability for the cab) which is IIRC 5db higher in sensitivity, and so should produce more volume per watt than the incumbent factory effort. Does this come with a downside?

[Bill Fitzmaurice]

In general higher sensitivity is realized at the cost of low frequency response. That's why guitar drivers tend to have higher sensitivity than bass drivers, which tend to have higher sensitivity than PA subwoofer drivers. However, you can't trust manufacturer sensitivity ratings, as they tend to play fast and loose with the details, like at what frequency it was measured. 105dB/watt seems wonderful, but not if it was measured at 2kHz. The only way to be sure if a driver swap is worthwhile is to use speaker modeling software to compare different options. The fly in that ointment is that it's usually impossible to get the necessary Thiele/Small specs on OEM drivers.

[Dan Dare]

 There is a trade off. Sensitive speakers are difficult to work with and must be treated with kid gloves or they'll burst into tears. As you were.

Edited November 10, 2019 by Dan Dare

[Lozz196]

34 minutes ago, Dan Dare said:

 There is a trade off. Sensitive speakers are difficult to work with and must be treated with kid gloves or they'll burst into tears. As you were.

Snowflake speakers with their sensitivities, can’t they just man up?

[Hellzero]

Yep, beware of specifications, Schroeder, for example has delirious ones. Everything is too much, because with their announced sensitivity (always around 105 dB/1W/1m) you would only need a 50 Watts amp to drive these speakers to total madness. I've tried a few and this is pure snake oil as you need a damn powerful amp to get the amount of sound they pretend out of it.

I've owned the Audiokinesis Hathor 18134 speaker, which was a monster able able to manage 1580 Watts RMS at 4 Ohms at ... 95 dB sensivity. The downside was it needed a very powerful amp with lot of current to move this 18 inches speaker. It did very well with the GSS Sumo 1000.

That's this :

 

[Muzz]

22 hours ago, Bill Fitzmaurice said:

In general higher sensitivity is realized at the cost of low frequency response. That's why guitar drivers tend to have higher sensitivity than bass drivers, which tend to have higher sensitivity than PA subwoofer drivers. However, you can't trust manufacturer sensitivity ratings, as they tend to play fast and loose with the details, like at what frequency it was measured. 105dB/watt seems wonderful, but not if it was measured at 2kHz. The only way to be sure if a driver swap is worthwhile is to use speaker modeling software to compare different options. The fly in that ointment is that it's usually impossible to get the necessary Thiele/Small specs on OEM drivers.

Thanks Bill; the speaker I'm looking at is an Eminence, so I'm hoping it's not all completely made up...

[agedhorse]

As Bill mentioned, low frequency extension is generally one tradeoff, another is power handling. As power handling increases, sensitivity generally decreases. The idea is that the SPL gain from increased power handling exceeds the SPL loss due to higher power handling (which it usually does up until a point). For instance, you may lose a dB in sensitivity to double the power handling.

Regarding sensitivity specs, I typically use the average sensitivity from 50Hz to 5kHz for bass guitar because this seems to reasonably represent the bandwidth that most players perceive as "volume". Unfortunately, there are plenty of examples of sensitivity that are simply impossible to achieve using such metrics. Also, it's important to distinguish sensitivity at 1W/1M from 2.83V/1M because 2.83V/1M with a 4 ohm cabinet will result in numbers 3dB higher due to it now being the equivalent of 2W/1M rather than 1W/1M.

[Bill Fitzmaurice]

1 hour ago, Muzz said:

Thanks Bill; the speaker I'm looking at is an Eminence, so I'm hoping it's not all completely made up...

It's not going to be made up, but even Eminence uses a pass band averaged figure, which doesn't say anything about what it will do where it matters for bass, below 100Hz. Besides, if you don't know what your original driver has for response there's no way to predict how a replacement will compare.

[Phil Starr]

One of the nice thigs about Eminence is that they are forthcoming about their specs and how they are measured. If you go on their website then have a look at the frequency charts, as Bill says they are averaged but have a look at the frequency graph above (roughly) 200Hz and see if there are any big bumps and where they are. Any humps will push the sensitivity up but if they are frequencies you don't want then that isn't very useful and you may be tweaking eq to take them down. Ignore the low frequency response below 200Hz as that is determined by the test conditions they use.

You asked the technical question. Sensitivity depends upon cone area, the strength of the magnet system, the mass of the cone, the resistance of the suspension to cone movement and the amount of the voice coil that is inside the magnetic field. All of these interact with other parts of the performance of the speaker. The cone needs to be heavy for bass but massive is harder to shift, the coil needs to be long for bass but that means it is mainly outside of the magnetic field so some of the power is lost. The mass of the cone and suspension change the tuning of the speaker at bass frequencies and so on. In the end a speaker tuned low and with good extension is going to be less efficient just because of the basic mechanics. You can compensate for this by putting in a really good magnet system to compensate. Hence the rule that you can't have cheap, loud and deep all at the same time