• Check out the Annual AK Secret Santa exchange Program, signup before November 23rd in the AK Charities & Christmas forum!

A Sensitive Topic or an Efficient Post?

ken kantor

Fowl Humor
When Efficiency, When Sensitivity?

OK, after a bit of research, here is what I have come up with.

In the US, the convention is to use the term "Sensitivity" synonymously
with "Voltage Sensitivity." This is what I was advocating. However, in
Europe the tendency is to be more rigorously correct, avoiding the use
of the term "Sensitivity" by itself, in favor of "Voltage Sensitivity" or
"Power Sensitivity." In any case, Sensitivity implies the input is a
reference with a dimension such as voltage or power, while the output
is expressed as Sound Pressure Level, (dBspl).

In contrast, Efficiency is a dimensionless ratio, correctly stated as a
percentage. It is incorrect to us the term "Efficiency" if one is listing
an SPL output for a given input, even if the input is referenced to power.
In this case, what is actually being stated is simply the "Power Sensitivity."

Therefore, I retract my previous definition equating sensitivity only to
voltage and efficiency only to power. I replace it with:

1- Efficiency refers only to power, and is a dimensionless ratio.

2- Efficiency can never be stated using SPL, Volts, Ohms,
or any other dimension.

3- If an input and/or output dimension is stated, then the term
Sensitivity applies; either Voltage or Power Sensitivity,
as the case may be.

4- If the term "Sensitivity" is used alone, it is taken to mean Voltage
Sensitivity.

Some quotations:

"...What is important, therefore, is not efficiency but sensitivity
defined as the sound-pressure level produced by a loudspeaker at lm
by an input voltage of 2.83V..."

Atkinson in AES

"The overall sensitivity is defined as the ratio of the reproducer
output voltage to the recorder input voltage, at some specified
reference frequency."

McKnight in AES

"Sensitivity is a measure of a speaker's output volume level for every
volt of input..."

Ward in Sound on Sound

-----

We now return me to my regularly scheduled life.

-k
 
Last edited:
Register to hide this ad
Ken, thanks for doing the research and clearing that up. I was confused after that last thread.
 
Hi,

Practically you only need to know a few basic principles :

Speakers impedances referred to are the nominal overall impedance.

1) Real efficiency is meaningless due to speaker impedance variation.
(It can be stated as a % efficiency for a single driver meaningfully.)

2) All speakers are measured with a fixed voltage input, not power.

3) Sensitivity is for 2.83V @ 1m, the equivalent of 1W/8ohms/1m.

4) Speaker nominal impedance, 4, 6 or 8ohm does not effect sensitivity.
2.83V into 4ohms is 2W and 1.5W into 6ohms, more power than 8 ohms.

5) Consequently a 4ohm 89dB speaker is 3dB less "efficient" than an 8 ohm.

6) Sensitivity /(efficiency) is often stated as dB/Watt/1m. This number
is the same for a 8 ohm speaker but should be 3dB lower for a 4 ohm
speaker, it often isn't in a specification, the 2.83V number is given.

rgds, sreten.
 
Hi,

Practically you only need to know a few basic principles :

Speakers impedances referred to are the nominal overall impedance.

1) Real efficiency is meaningless due to speaker impedance variation.
(It can be stated as a % efficiency for a single driver meaningfully.)

2) All speakers are measured with a fixed voltage input, not power.

3) Sensitivity is for 2.83V @ 1m, the equivalent of 1W/8ohms/1m.

4) Speaker nominal impedance, 4, 6 or 8ohm does not effect sensitivity.
2.83V into 4ohms is 2W

5) Consequently a 4ohm 89dB speaker is 3dB less "efficient" than an 8 ohm.

6) Sensitivity /(efficiency) is often stated as dB/Watt/1m. This number
is the same for a 8 ohm speaker but should be 3dB lower for a 4 ohm
speaker, it often isn't in a specification, the 2.83V number is given.

7) You often can't trust the numbers given .......

rgds, sreten.
 
When Efficiency, When Sensitivity?

OK, after a bit of research, here is what I have come up with.

In the US, the convention is to use the term "Sensitivity" synonymously
with "Voltage Sensitivity." This is what I was advocating. However, in
Europe the tendency is to be more rigorously correct, avoiding the use
of the term "Sensitivity" by itself, in favor of "Voltage Sensitivity" or
"Power Sensitivity." In any case, Sensitivity implies the input is a
reference with a dimension such as voltage or power, while the output
is expressed as Sound Pressure Level, (dBspl).

In contrast, Efficiency is a dimensionless ratio, correctly stated as a
percentage. It is incorrect to us the term "Efficiency" if one is listing
an SPL output for a given input, regardless of whether the input is
referenced as power or voltage. In this case, what is actually being
stated is "Power Sensitivity."

Therefore, I retract my previous definition equating sensitivity only to
voltage and efficiency only to power. I replace it with:

1- Efficiency refers only to power, and is a dimensionless ratio.

2- Efficiency can never be stated using SPL, Volts, Ohms,
or any other dimension.

3- If an input and/or output dimension is stated, then the term
Sensitivity applies; either Voltage or Power Sensitivity,
as the case may be.

4- If the term "Sensitivity" is used alone, it is taken to mean Voltage
Sensitivity.

Some quotations:

"...What is important, therefore, is not efficiency but sensitivity
defined as the sound-pressure level produced by a loudspeaker at lm
by an input voltage of 2.83V..."

Atkinson in AES

"The overall sensitivity is defined as the ratio of the reproducer
output voltage to the recorder input voltage, at some specified
reference frequency."

McKnight in AES

"Sensitivity is a measure of a speaker's output volume level for every
volt of input..."

Ward in Sound on Sound

-----

We now return me to my regularly scheduled life.

-k
The first thing that most normal people struggle with are the "English" interpretations of the words themselves. Science has never been very good at using common words to describe events or measurements, and while I'd love to produce examples for the moment I'm simply brain dumping.

Sensitivity is most commonly recognized as say eyes sensitive to light, or the sensitivity of a solar cell to light (but also thought of as efficiency for the amount of electrical generation for given amount of light exposed to). How sensitive one is to a given statement or word. Sensitivty is more often thought of as a reaction to something as opposed to a degree, though we might say that person is "really sensitive" which is more subjective than objective.

Efficiency, on the other hand is more often used to reference, in terms of ration usually, how, shall we say, economical something is, i.e. the less fuel a vehicle uses to travel a certain distance or consumed over time (like a boat as an example) is a reflection of how efficient it is. Efficiency is often more objective than subjective, making it more suitable for scientific reference.

It's not difficult to confuse or substitute one for the other depending on a given interpretation of the root word itself. To say one word is better or more meaningful is really a function of the definition of the root words themselves.

Until someone comes up with clear definitions of the words "Efficiency" and "Sensitivity" such that they clearly mean specific things in all cases, this subject will forever be challenged.

Good grief how tall is this soap box? Someone got a ladder. ;)
 
Am I to assume then, that my arguments had some degree of validity after all? If so, Ken I applaud you for researching this further and even going on record to state this. I bet many people would not have done so.:thmbsp:

I know you know a lot more about speakers than myself, so I felt a little dumb disagreeing with you. I'm glad is this cleared up now, or at least I think it is.:scratch2:
 
I guess somebody should point out something to anyone that stumbles upon this unaware of previous discussions: These definitions (so to speak) of the two terms are not followed on spec. sheets, in general. Not because some other definition is substituted, though - it's just pretty much random.

Ken, thanks for putting in the effort!
 
Am I to assume then, that my arguments had some degree of validity after all? If so, Ken I applaud you for researching this further and even going on record to state this. I bet many people would not have done so.:thmbsp:

I know you know a lot more about speakers than myself, so I felt a little dumb disagreeing with you. I'm glad is this cleared up now, or at least I think it is.:scratch2:

Wait, did I say I was wrong??? I take it all back!!!! :(

Seriously, your arguments always did have some validity. My debating certain terminology was never meant to be a wholesale dismissal. Anyway, I shall have to reread your posts in light of my latest, hopefully improved, understanding.

-k
 
Wait, did I say I was wrong??? I take it all back!!!! :(

Seriously, your arguments always did have some validity. My debating certain terminology was never meant to be a wholesale dismissal. Anyway, I shall have to reread your posts in light of my latest, hopefully improved, understanding.

-k

Well the only things I had said from the first post from, I don't know, several weeks ago when this all started was that efficiency to me was expressed as a percentage, and that sensitivity is expressed in SPL regardless of whether 1 watt or 2.83v was used. And then we disagreed from that point on it seemed, until this thread. The main disagreement I thought was mine and your definition of efficiency.

Maybe I'm confused.:scratch2:
 
Well the only things I had said from the first post from, I don't know, several weeks ago when this all started was that efficiency to me was expressed as a percentage, and that sensitivity is expressed in SPL regardless of whether 1 watt or 2.83v was used. And then we disagreed from that point on it seemed, until this thread. The main disagreement I thought was mine and your definition of efficiency.

Maybe I'm confused.:scratch2:

Don't be confused. Let's just walk hand in hand toward the sunset. Until the next thing we get into a cat fight about....

-k
 
1) Real efficiency is meaningless due to speaker impedance variation.
(It can be stated as a % efficiency for a single driver meaningfully.)

Both efficiency and sensitivity are functions of frequency.

3) Sensitivity is for 2.83V @ 1m, the equivalent of 1W/8ohms/1m.

While I was taught to use dB @ Watt/Ohm/meter, as can be seen on the AR spec that I posted, the terminology is factually incorrect. Watt/Ohm/meter is a statement of "Power Sensitivity." It is not a dimensionless ratio, so it cannot be called "Efficiency," in the strict use of the term. As I said, I have used it thousands of times, so fully understand its common usage and meaning. Live and learn.

-k
 
I've always tried to specify exactly what I was talking about, in terms of sensitivity, for just the reason that there seems to be some difference in the definition of these terms, in different times and places.

For example- I would generally say "this speaker is xxdB @1w @ 1meter, sensitivity", and specify the impedance of the speaker as well. Or, "this speaker is xxdB @ 2.83v, @1m"... also specifying the impedance.

This can help to alleviate some of the confusion, as the effective sensitivity of the speaker is influenced greatly by impedance, while efficiency, ostensibly, is not (i.e, Hoffman's Iron Law- the size of the enclosure and bass extension completely determine the efficiency).

The measure of 2.83v has been convenient, in that most speakers are rated at 8 ohms, and 2.83v into 8 ohms equals one watt of power. However, if the speaker is 4 ohms, it only takes two volts to reach 1 watt... and differs accordingly, for other impedance, too...

One other variable I've seen, is the DISTANCE that the SPL measurement is taken at, to rate the sensitivity! For example- many JBL speakers were rated at distances GREATER than 1 meter distance- say, a typical measurement might be "78dB SPL @ 1w @ 15 feet", (which is the rating for the well-known L26 system, IIRC). Fortunately, those measurements can also be converted to "standard" 1 meter distance measurements, using the inverse square law (the acoustic power decreases as the SQUARE of the distance from the source increases). In short, if the distance is doubled, the sound pressure level goes down by 6dB, and so on. Takes some fiddling with log conversions on a calculator, but it's readily convertible!

In short, it's good to be as exact as possible, in giving these ratings. Make sure the reader has enough information to do the calculations themselves. With the following parameters stated:

1) impedance of the speaker in question,
2) the sensitivity rating at either 2.83v or 1 watt, and
3) the distance from the speaker the measurements are made at,

all the other calculations can be done, readily, to compare one speaker to another.

Regards,
Gordon.
 
There is one other calculation that can be done- I decided to split it into a separate post from the above, to avoid confusion, since it's a bit complicated itself. That is, the conversion from sensitivity to efficiency. It's not as easy as it might first seem!

In theory, a perfect speaker (omnidirectional at every frequency, flat response from DC to light) would be capable of an output level of 112dB @1 watt @ 1 meter distance. That would be 100% efficiency, in essence... no loss anywhere.

But, there's several factors in that statement. First is the "omnidirectional" requirement. A speaker that is NOT omnidirectional- and every known speaker has directional restrictions of some sort- they "beam" (say, a cone driver driven to above the frequency where the cone diameter is 1/2 wavelength of the sounds produced) or are purposefully directional (such as a horn, which may only operate into a small conic section, instead of full spherical space) will produce MORE effective SPL into that restricted area, than it would if its energy was distributed evenly into all directions. In essence, it's possible to "concentrate" the energy of a speaker into a smaller radiating area... the total energy is the same, but the area is smaller- so, the intensity within that smaller area, HAS to be greater (by conservation of energy!).

So, for a given speaker, to find the efficiency, you need:

1) the impedance of the speaker
2) the sensitivity at 1w or 2.83v
3) the distance of the measurement, AND
4) the DISPERSION AREA ("conic area", "dispersion angle" or such) of the speaker.

The dispersion "concentration factor" is pretty easy to calculate. You just need to find out what effective portion of a sphere, the radiating area is... take that, and multiply the radiated power (which can be found by converting decibels to acoustic watts, given that 112dB @1w@1m into spherical space is one acoustic watt) by that factor. For example, if the speaker is operating into a 90 degree rectangular corner space (say, a woofer in a triple corner), that's 1/8 of a sphere- the total acoustic output power at one electrical watt input, should be multiplied by 1/8, to get the equivalent power into a full spherical space. With an input of one electrical watt into the speaker in question, that fraction would be the effective efficiency.

And truthfully- you'd also want to factor in the frequency response. If the speaker varies in output with frequency, then it technically is NOT 100% efficient. However, we generally just assume it's flat within the needed passband (the range of frequencies we need), and ignore drop-offs outside that range...

Yeah, it's kind of complicated... but at least the math is well known, and do-able, if you have the inclination to go through it...

Regards,
Gordon.
 
Last edited:
Back
Top Bottom