Exactly, you (your amp) need to double the power to get another 3dB out of the speaker. If you start at, say, 85 dBper watt per meter, (a rather inefficient loudspeaker) then, when you want to reproduce a full orchestral cresento at, say, 110 dB(still one meter away from the driver), then you'll need: 1watt-85dB, 2watts-88dB, 4w-91dB, 8-94, 16-97, 32-100, 64-103, 108-106, 216watts-109dB! And that's only at 1 meter (1.1 yards) away from the driver! Now, if your head is at an equivalent distance of 1meter from both loudspeakers, you'll get double the sound pressure which is another 3dB. That means you'll get 112dBs from both loudspeakers for 216 watts per channel of amp power. Or, you'll get 109dBs (your goal) from both speakers at one meter away, for 108 watts per channel. This sounds easy but, if you (normally) have your listening position at, say, 4 meters (around 13feet)away, then this 109 dBs from both loudspeakers drops dramatically to(!!!) 97 dBs! And that's because every doubling of the distance from the source leads to a drop of 6dBs! (so, from 1 to 2 meters, drop from 109 to 103 dB and then from 2 to 4 meters away, another drop to 97 dB). So, if you get 97 dB at 13ft away for 108 watts per channel, then you'll need...1728 watts, per channel, to reach 109 dBs at your listening position! So, do yourself a favor: BUY A TRULY EFFICIENT LOUDSPEAKER! (and listen at a closer distance away). Efficient loudspeakers don't come without flaws, though. Most of them are plagued by distortions and poor frequency response charts. The sad truth is that you'll still need A LOT of power to reproduce those very low and barely audible (hence, needing a lot of amplification) frequencies. You'll need a sub (or two) but you'll be lucky by the fact that low and very low frequencies get trapped by your room (and, also, amplified by adjacent room boundaries) and so the loss by the distance is not as great as with mid and high frequencies. Also, big planar loudspeakers or line arrays/sources (different than usual point sources), have a much lower than 6dBs per double-distance loss, so they sound louder in a room than one would expect from their efficiency measurement. (Of course, these come with another set of flaws, as well...). Anyway, be ready to make WISE COMPROMISES, taking in account, your room and your tastes and your wallet.
This is a good explanation of why so much power and efficient speakers are needed for live events that are in large arenas or outdoors. Some of the audience can be over 50 meters away and there also seems to be the unfortunate need to blast everyone with music that is as loud as possible. However, most home reproduction is in a relatively small room and the sound reflects so doesn't reduce at an anechoic chamber rate.
That's why I said there is more but I tried to keep it short. What I describe refers to either an open space or a fully anechoic chamber. A real world room contributes to the preservation of higher levels of sound pressure by keeping a good fraction of it through reflections. (But this is done very inconsistently throughout the frequency spectrum). In a real, medium sized room, not overly treated with absorbing materials, for the same example, one could only need about a quarter of the theoretically estimated power (still a good over 400watts). Other factors have to be taken into account, like the frequency range where the sound pressure level peaks. If this happens to be a frequency range at which the impedance of the loudspeaker drops low, this would require from the amp to provide a lot more watts than initially estimated. (Good point about outdoor use of efficient speakers).
yes, Paul. Double the power to gain 3db of sound. This is important at the listening position/distance. The standard measurements are ok if you listen at 1m away from the speakers. :)
One thing that struck me about this was it was measuring efficiency at a specific frequency, power and distance, but do speakers vary in efficiency depending on the frequency. So one speaker might be more efficient at 1 Khz and another more efficient at 2 Khz. If so is the 1m, 1Khz, I watt test a rule of thumb that gives you a rough idea of how loud a given set of speakers might get with a given amp. In addition is the possible differences in efficiency at difference frequencies part of what causes speakers to charachterised as "Bright" or "flat" sounding?
Thanks for the explanation. I would have liked you to explain what sensitivity means to the consumer for people shopping speakers. Is a speaker at 85db inferior to a 90db speaker? Are there tradeoffs or should we be shopping always for a higher sensitivity as a rule? I guess I'm asking -- how important is this spec when evaluating a speaker?
What matters more is cone materials, surround, damping etc, efficiency can be good or bad. My ~87dB/w Isobariks sound much better than my 99dB/w array speakers, though I'd imagine a klipsch loudspeaker to sound better than both at 95 or whatever. Though around 90 is a good start.
Efficiency of a speaker doesnt tell anything about the sound quality for example. It just gives a tendency if you need a more powerful amp if you want to hear loud, or not. A Vanlue of 85 db/1W/1m up to 90 is normal in Hifi. Below you should think about power, above it is less important. See the comment of "Full Ranger". Highest example. 1 Speaker, 216 Watt = 109db in 1 meter distance to the speaker. You can add 3db because 2 speakers involved. so 112 db at 1 meter! You are sitting maybe 4 meters away. outside, without reflections - double distance = half loudness (-6db). inside a room its more! But... 1 meter = 112db 2 meters = 106db 4 meters = 100db. Thats very close to discotheque dancefloor level! hope that can help... cheers
Very high efficiency loudspeakers usually have more distortion and worse frequency response charts than lower efficiency ones. On the other hand, it's not quite so fair to compare their frequency response charts, because, if, for instance, you press the low efficiency speaker to reach the sound pressure level of the higher efficiency one, you don't know what kind of curve you'll get. As a rule, a low efficiency loudspeaker, no matter how good quality it may be, will require from you A LOT of power. And if you need a lot of GOOD power, then you'll also need a lot of MONEY. If it is a usual, electrodynamic one, with a fairly stable impedance along the frequency spectrum, it may save you some. So, a low efficiency loudspeaker will, finally, be more, or even much more expensive than you think. On the other hand, a good quality (low distortion, well made) high efficiency one, will save you money from the amp. If it has a really high efficiency (say 95-97 dB SPL per watt per meter), then you may enjoy the very best of amplification and overall sound around, at a very good price (because you'll need about 3 to 10 good watts to make it really sing). If you go for horns (which may be even more efficient), be prepared for "horny" distortions and rather poor imaging (but also a lot of excitement, for as long as you have some hearing left).
I wonder what the difference in ratings would be in subwoofers if we used say 50 or 60hz instead of 1k.. I own different subwoofers of the same diameter and power handling but one has a higher efficiency than the other.. however, the less efficient one has an xmax of 23.5mm whereas the more efficient one’s xmax is only 10mm.. the less efficient sub can produce lower tones much louder in the same enclosure(s) than the more efficient one and their fs is within half a hz of each other.. obviously moving more air makes sense, but would like to see the difference in efficiency rating at a lower frequency. Guess I’ll have to dig out my old db meter and measuring tape lol
A friend and I built a mad 15" rear-loaded horn using some Eminence plans back in the 90s. The driver alone is rated 99dB / 1W / 1m so with the horn it is maybe 102dB! I could run that thing off a headphone amp. I wouldn't, though. It sounds like ass.
you are right, 3 db increase is the speaker playing twice as loud, but the hearing needs 10db increase to feel like its twice as loud. you still feel twice the punch from bass tho, thats why huge speakers can give you a physical punch from low bass at low volume and small speakers can play low bass with noe punch at all.
Chris Bishop Double the power is not double the volume. It takes 10db to be perceived as twice as loud. So to be clear 3db is twice the power. 10db is twice the loudness.
carlos oliveira I'm fully aware of "perceived" level changes. That's not a scientific measurement, that's "psychoacoustics" ... The relative loudness that we perceive is a subjective psychological phenomenon, not something that can be objectively measured.
Chris Bishop Well then how do you explain you are wrong? 3db power increase is not double the volume as you have stated. Volume or loudness is always perceived
carlos oliveira Doubling of the volume (loudness) should be felt by a level difference of 10 dB − acousticians say. Doubling the sound pressure (voltage) corresponds to a measured level change of 6 dB. Doubling of acoustic power (sound intensity) corresponds to a level change of 3 dB. 6 dB represents a doubling of sound pressure. 6 dB represents a doubling of voltage (RMS amplitude). 6 dB less SPL means a doubling of distance. 3 dB SPL increase requires double the power. 3 dB represents a doubling of sound intensity.
One simple way to test it is to set 1 speaker at a certain level, then add a second speaker, it won't be twice as loud, but it will be twice the power.
A good explanation of efficiency thank you, but I’m still not sure how that equates to sound quality? I’ve recently got obsessed with headphones, and a lot of the best sounding (dynamic driver) ones seem to be really inefficient, is it the same with loudspeakers?
Hi Paul I have had about three deferent amplifiers, starting with a Rega Elicit R rated at 105 watts per channel in to 8 ohms, a Naim supernait 2 rated at 80 watts per channel in to 8 ohms and a Hagel 360 rated at 250 watts per channel. Out off the three amplifiers, the Naim was the most powerful but it had the lowest power rating, I think if I turned it up to much the speaker cones would fly out. My question, is one watt out of any amplifier the same, or is there more to the way the power is delivered and can this affect how speakers are tested for dB rating if an amplifier gave out a louder watt, are manufacturers over claiming the amplifiers output? I enjoy your videos, hope you keep uploading more thanks.
Jonathan Bishop Firstly some manufacturers rate their amps conservatively so a rated 80 watt amp?may actually put out 110 watts or more. Also amps have varying levels of headroom. 2 db is typical but some designs can have 3 or even 6 db for brief periods. Also some speakers may have a nominal impedance of 8 ohms but can have dips under 2 ohms at certain frequencies. If this occurs in the upper bass area then a high current amp will take it in stride and sound much more powerful than an amp that struggles with a load lower than say 4 ohms. Very few speakers have a relatively flat impedance. Some speakers are known to require high current amps. All others will sound wimpy. High current amps can generally double the wattage with each halving of the impedance. There are even more reasons why amps of the same rated wattage can sound like they are not, but you get the idea. All amps are not created equal even when their ratings say they are.
All these comments saying how great the explanation was, and I'm here sighing and shrugging. Guess I expected more than the base explanation someone should already know. Like what if anything does it tell us about them other than they get loud? Does that increase also add detail or is that a separate issue? Many questions.
Generally, when a louspeaker goes loud, also its distortion increases. You are more likely to hear low level detail at a louder level, but, generally, to hear low level detail you need both loudness AND low distortion.
A 10dB increase in level is about what most people would refer to as "twice as loud" but requires a ten fold increase in power...thus the logarithmic scale. So if 1 watt drives a certain speaker to 85dB you need about 10 watts to drive the same speaker to 95dB. 105 dB would take 100 watts and if you wanted 115dB you'd need 1000 watts! Whew! Not many speakers or amps could take or deliver that kind of wattage. Now take a typical PA (Public Address) Loudspeaker. It's not uncommon to find a two way unit with say a 12" woofer and horn loaded tweeter that will give up around 98dB with only 1 watt of input. So now 10 watts will give us 108dB and 100 watts about 118dB. However more efficiency doesn't mean better sound QUALITY just more quantity for a giving wattage. Want to know everything about sound systems.? "The Yahama Sound Reinforcement Handbook" ...about 20 bucks.
The sensitivity is only part of the story, it doesn't tell you it's maximum volume level, because speakers also have power ratings, supply too much power and you will cook the voice coil, even worse, is clipping the signal because your amp is under powered for the speakers efficiency for the volume you're trying to get to. All of this has no correlation to sound quality, I've heard efficient and not so efficient speakers sound beautiful, they will need very different amps to be able to perform at the same volume level, so a speaker with 3db less efficiency(1/2) will require an amp with 3db more power(2x) to reach the same output.
Exactly, you (your amp) need to double the power to get another 3dB out of the speaker. If you start at, say, 85 dBper watt per meter, (a rather inefficient loudspeaker) then, when you want to reproduce a full orchestral cresento at, say, 110 dB(still one meter away from the driver), then you'll need: 1watt-85dB, 2watts-88dB, 4w-91dB, 8-94, 16-97, 32-100, 64-103, 108-106, 216watts-109dB! And that's only at 1 meter (1.1 yards) away from the driver! Now, if your head is at an equivalent distance of 1meter from both loudspeakers, you'll get double the sound pressure which is another 3dB. That means you'll get 112dBs from both loudspeakers for 216 watts per channel of amp power. Or, you'll get 109dBs (your goal) from both speakers at one meter away, for 108 watts per channel. This sounds easy but, if you (normally) have your listening position at, say, 4 meters (around 13feet)away, then this 109 dBs from both loudspeakers drops dramatically to(!!!) 97 dBs! And that's because every doubling of the distance from the source leads to a drop of 6dBs! (so, from 1 to 2 meters, drop from 109 to 103 dB and then from 2 to 4 meters away, another drop to 97 dB). So, if you get 97 dB at 13ft away for 108 watts per channel, then you'll need...1728 watts, per channel, to reach 109 dBs at your listening position! So, do yourself a favor: BUY A TRULY EFFICIENT LOUDSPEAKER! (and listen at a closer distance away). Efficient loudspeakers don't come without flaws, though. Most of them are plagued by distortions and poor frequency response charts. The sad truth is that you'll still need A LOT of power to reproduce those very low and barely audible (hence, needing a lot of amplification) frequencies. You'll need a sub (or two) but you'll be lucky by the fact that low and very low frequencies get trapped by your room (and, also, amplified by adjacent room boundaries) and so the loss by the distance is not as great as with mid and high frequencies. Also, big planar loudspeakers or line arrays/sources (different than usual point sources), have a much lower than 6dBs per double-distance loss, so they sound louder in a room than one would expect from their efficiency measurement. (Of course, these come with another set of flaws, as well...). Anyway, be ready to make WISE COMPROMISES, taking in account, your room and your tastes and your wallet.
Nice job explaining that Full Ranger.
Thanks, Spiceman50s. There's more, of course, but I tried to keep it short.
This is a good explanation of why so much power and efficient speakers are needed for live events that are in large arenas or outdoors. Some of the audience can be over 50 meters away and there also seems to be the unfortunate need to blast everyone with music that is as loud as possible. However, most home reproduction is in a relatively small room and the sound reflects so doesn't reduce at an anechoic chamber rate.
Charts don't tell the whole story especially if there are only tests in an anechoic chamber and not any done in a real world enviroment.
That's why I said there is more but I tried to keep it short. What I describe refers to either an open space or a fully anechoic chamber. A real world room contributes to the preservation of higher levels of sound pressure by keeping a good fraction of it through reflections. (But this is done very inconsistently throughout the frequency spectrum). In a real, medium sized room, not overly treated with absorbing materials, for the same example, one could only need about a quarter of the theoretically estimated power (still a good over 400watts). Other factors have to be taken into account, like the frequency range where the sound pressure level peaks. If this happens to be a frequency range at which the impedance of the loudspeaker drops low, this would require from the amp to provide a lot more watts than initially estimated. (Good point about outdoor use of efficient speakers).
I continue to enjoy and appreciate your creativity in moving around the facility to tape your segments Paul.
yes, Paul. Double the power to gain 3db of sound. This is important at the listening position/distance. The standard measurements are ok if you listen at 1m away from the speakers. :)
One thing that struck me about this was it was measuring efficiency at a specific frequency, power and distance, but do speakers vary in efficiency depending on the frequency. So one speaker might be more efficient at 1 Khz and another more efficient at 2 Khz. If so is the 1m, 1Khz, I watt test a rule of thumb that gives you a rough idea of how loud a given set of speakers might get with a given amp. In addition is the possible differences in efficiency at difference frequencies part of what causes speakers to charachterised as "Bright" or "flat" sounding?
Thanks for the explanation. I would have liked you to explain what sensitivity means to the consumer for people shopping speakers. Is a speaker at 85db inferior to a 90db speaker? Are there tradeoffs or should we be shopping always for a higher sensitivity as a rule? I guess I'm asking -- how important is this spec when evaluating a speaker?
What matters more is cone materials, surround, damping etc, efficiency can be good or bad. My ~87dB/w Isobariks sound much better than my 99dB/w array speakers, though I'd imagine a klipsch loudspeaker to sound better than both at 95 or whatever. Though around 90 is a good start.
Efficiency of a speaker doesnt tell anything about the sound quality for example.
It just gives a tendency if you need a more powerful amp if you want to hear loud, or not.
A Vanlue of 85 db/1W/1m up to 90 is normal in Hifi.
Below you should think about power, above it is less important.
See the comment of "Full Ranger". Highest example.
1 Speaker, 216 Watt = 109db in 1 meter distance to the speaker.
You can add 3db because 2 speakers involved. so 112 db at 1 meter!
You are sitting maybe 4 meters away. outside, without reflections - double distance = half loudness (-6db).
inside a room its more! But...
1 meter = 112db
2 meters = 106db
4 meters = 100db.
Thats very close to discotheque dancefloor level!
hope that can help...
cheers
Thanks. That does help!
Very high efficiency loudspeakers usually have more distortion and worse frequency response charts than lower efficiency ones. On the other hand, it's not quite so fair to compare their frequency response charts, because, if, for instance, you press the low efficiency speaker to reach the sound pressure level of the higher efficiency one, you don't know what kind of curve you'll get. As a rule, a low efficiency loudspeaker, no matter how good quality it may be, will require from you A LOT of power. And if you need a lot of GOOD power, then you'll also need a lot of MONEY. If it is a usual, electrodynamic one, with a fairly stable impedance along the frequency spectrum, it may save you some. So, a low efficiency loudspeaker will, finally, be more, or even much more expensive than you think. On the other hand, a good quality (low distortion, well made) high efficiency one, will save you money from the amp. If it has a really high efficiency (say 95-97 dB SPL per watt per meter), then you may enjoy the very best of amplification and overall sound around, at a very good price (because you'll need about 3 to 10 good watts to make it really sing). If you go for horns (which may be even more efficient), be prepared for "horny" distortions and rather poor imaging (but also a lot of excitement, for as long as you have some hearing left).
Great video-thanks for all that. Been watching your stuff for a while. Hoping to see you @ the next RMAF!
Some of the bigger Avantgardes are actually up in the 109dB/W range
I wonder what the difference in ratings would be in subwoofers if we used say 50 or 60hz instead of 1k.. I own different subwoofers of the same diameter and power handling but one has a higher efficiency than the other.. however, the less efficient one has an xmax of 23.5mm whereas the more efficient one’s xmax is only 10mm.. the less efficient sub can produce lower tones much louder in the same enclosure(s) than the more efficient one and their fs is within half a hz of each other.. obviously moving more air makes sense, but would like to see the difference in efficiency rating at a lower frequency. Guess I’ll have to dig out my old db meter and measuring tape lol
A friend and I built a mad 15" rear-loaded horn using some Eminence plans back in the 90s. The driver alone is rated 99dB / 1W / 1m so with the horn it is maybe 102dB! I could run that thing off a headphone amp. I wouldn't, though. It sounds like ass.
Hi Paul.
Technically, a 3 dB increase in power is a double in volume:
10 * log10 (2) = 3.01 dB, where 2 is the double in power (volume)
you are right, 3 db increase is the speaker playing twice as loud, but the hearing needs 10db increase to feel like its twice as loud. you still feel twice the punch from bass tho, thats why huge speakers can give you a physical punch from low bass at low volume and small speakers can play low bass with noe punch at all.
Chris Bishop Double the power is not double the volume. It takes 10db to be perceived as twice as loud. So to be clear 3db is twice the power. 10db is twice the loudness.
carlos oliveira
I'm fully aware of "perceived" level changes. That's not a scientific measurement, that's "psychoacoustics" ...
The relative loudness that we perceive is a subjective psychological phenomenon, not something that can be objectively measured.
Chris Bishop Well then how do you explain you are wrong? 3db power increase is not double the volume as you have stated. Volume or loudness is always perceived
carlos oliveira
Doubling of the volume (loudness) should be felt by a level difference of 10 dB − acousticians say.
Doubling the sound pressure (voltage) corresponds to a measured level change of 6 dB.
Doubling of acoustic power (sound intensity) corresponds to a level change of 3 dB.
6 dB represents a doubling of sound pressure.
6 dB represents a doubling of voltage (RMS amplitude).
6 dB less SPL means a doubling of distance.
3 dB SPL increase requires double the power.
3 dB represents a doubling of sound intensity.
One simple way to test it is to set 1 speaker at a certain level, then add a second speaker, it won't be twice as loud, but it will be twice the power.
Ok but what makes speaker A more sensitive than speaker B is it cone mass ?
Ha ha... Paul always makes me laff !!! Loved the 1K tones !!
A good explanation of efficiency thank you, but I’m still not sure how that equates to sound quality? I’ve recently got obsessed with headphones, and a lot of the best sounding (dynamic driver) ones seem to be really inefficient, is it the same with loudspeakers?
Interesting stuff, Anyone know anything about Westrex 2080G speakers?, I've had a set for years as my 'daily drivers', I really like them, Cheers, J.
The old Infinity Perfect subs sound amazing and are 96db. Too bad nobody makes such good sounding, high efficient subs/speakers anymore.
Hi Paul
I have had about three deferent amplifiers, starting with a Rega Elicit R rated at 105 watts per channel in to 8 ohms, a Naim supernait 2 rated at 80 watts per channel in to 8 ohms and a Hagel 360 rated at 250 watts per channel. Out off the three amplifiers, the Naim was the most powerful but it had the lowest power rating, I think if I turned it up to much the speaker cones would fly out. My question, is one watt out of any amplifier the same, or is there more to the way the power is delivered and can this affect how speakers are tested for dB rating if an amplifier gave out a louder watt, are manufacturers over claiming the amplifiers output? I enjoy your videos, hope you keep uploading more thanks.
Jonathan Bishop Firstly some manufacturers rate their amps conservatively so a rated 80 watt amp?may actually put out 110 watts or more. Also amps have varying levels of headroom. 2 db is typical but some designs can have 3 or even 6 db for brief periods. Also some speakers may have a nominal impedance of 8 ohms but can have dips under 2 ohms at certain frequencies. If this occurs in the upper bass area then a high current amp will take it in stride and sound much more powerful than an amp that struggles with a load lower than say 4 ohms. Very few speakers have a relatively flat impedance. Some speakers are known to require high current amps. All others will sound wimpy. High current amps can generally double the wattage with each halving of the impedance. There are even more reasons why amps of the same rated wattage can sound like they are not, but you get the idea. All amps are not created equal even when their ratings say they are.
All these comments saying how great the explanation was, and I'm here sighing and shrugging. Guess I expected more than the base explanation someone should already know. Like what if anything does it tell us about them other than they get loud? Does that increase also add detail or is that a separate issue? Many questions.
Generally, when a louspeaker goes loud, also its distortion increases. You are more likely to hear low level detail at a louder level, but, generally, to hear low level detail you need both loudness AND low distortion.
A 10dB increase in level is about what most people would refer to as "twice as loud" but requires a ten fold increase in power...thus the logarithmic scale. So if 1 watt drives a certain speaker to 85dB you need about 10 watts to drive the same speaker to 95dB. 105 dB would take 100 watts and if you wanted 115dB you'd need 1000 watts! Whew! Not many speakers or amps could take or deliver that kind of wattage.
Now take a typical PA (Public Address) Loudspeaker. It's not uncommon to find a two way unit with say a 12" woofer and horn loaded tweeter that will give up around 98dB with only 1 watt of input. So now 10 watts will give us 108dB and 100 watts about 118dB. However more efficiency doesn't mean better sound QUALITY just more quantity for a giving wattage. Want to know everything about sound systems.? "The Yahama Sound Reinforcement Handbook" ...about 20 bucks.
The answer is correct: 3 dB more is double the amp output in watt. 10 dB more requires 10 times the power output. 20 dB 100 times and so on
It's all about Ohm's law.
en.wikipedia.org/wiki/Ohm%27s_law
The sensitivity is only part of the story, it doesn't tell you it's maximum volume level, because speakers also have power ratings, supply too much power and you will cook the voice coil, even worse, is clipping the signal because your amp is under powered for the speakers efficiency for the volume you're trying to get to.
All of this has no correlation to sound quality, I've heard efficient and not so efficient speakers sound beautiful, they will need very different amps to be able to perform at the same volume level, so a speaker with 3db less efficiency(1/2) will require an amp with 3db more power(2x) to reach the same output.
J-man72 b Don't confuse sensitivity with efficiency.
Hi Paul.
I was wondering how does age affect a driver? Let's say 20 years or older .
Amazingly explained. Thank you Paul.
GREAT explanation! Thanks so much.
Might want to convert your graveyard into a museum.