speaker sensitivity, spl [Archive] - Audio & Video Forums

PDA

View Full Version : speaker sensitivity, spl



MomurdA
03-02-2005, 11:14 AM
Sorry if this question has been asked before, but i couldnt find it while searching. If a speaker has a sensitivity of 95db@1w@1m, that means that from one meter away with only one watt going to it it is 95db? And that an increase in ten decibels requires twice as much power to that speaker? The reason im asking is because of my magnepans. They are about 86db@1w@1m. Instead of hooking them up directly to my marantz receiver, i am using a separate amp which is 100 watts per channel @ 4 ohms rms. Is this enuf power to make my maggies sound like they should?

Pat D
03-02-2005, 11:47 AM
Sorry if this question has been asked before, but i couldnt find it while searching. If a speaker has a sensitivity of 95db@1w@1m, that means that from one meter away with only one watt going to it it is 95db? And that an increase in ten decibels requires twice as much power to that speaker? The reason im asking is because of my magnepans. They are about 86db@1w@1m. Instead of hooking them up directly to my marantz receiver, i am using a separate amp which is 100 watts per channel @ 4 ohms rms. Is this enuf power to make my maggies sound like they should? An increase of 10 dB requires 10 times the power, and generally makes the music sound subjectively about twice as loud. An increase of 3 dB is slightly louder and requires twice the power.

The impedance of most speakers varies considerably with frequency it is impractical to try to figure out how much power they are using. Voltage sensitivity is measured with an input of 2.83 volts (which is 1 watt into an 8 ohm load) as it is more difficult--and not very useful--to actually figure out the efficiency of a speaker . Anyway, here is a link to a description of the type of measurements Soundstage does at the NRC:

http://www.soundstagemagazine.com/measurements/test_loudspeakers.htm

Whether your amp is adequate to drive Maggies depends on how loud you play them.

markw
03-02-2005, 11:51 AM
If a speaker has a sensitivity of 95db@1w@1m, that means that from one meter away with only one watt going to it it is 95db?In theory, yes. That's pretty much the "standard" method mfgrs use to measure this.



And that an increase in ten decibels requires twice as much power to that speaker??No. Applying twice as much power to the speaker would yield a 3 db increase in loudness, and 3 db is not much of a change at all.. To effectively double the loudness would require ten times the power.Here's where it gets interesting.

So, the spl a speaker produces is relative to it's sensitivity and the power drivng it.

Your speakers are 86 db and, in theory, it produces this with one watt. To geta 3 db increase would require we double this to two watts and so on and so on. Let's put this in a table to see it a little better.

86 db with 1 watt
89 db with 2 watts
92 db with 4 watts
95 db with 8 watts
--- somewhere around here the sound would seem about twice as loud.
98 db with 16 watts
101 db with 32 watts
104 db with 64 watts
--- somewhere around here the sound would seem about twice again as loud.
107 db with 128 watts
etc, etc...

As far as to if that's enough for your MMG's well, that depends what type of music you like and how loud you like it. What I can say is that my MMG's, with a powered sub cutting in around 60 hz, did quite well with a NAD 214, which was rated at 80 wpc @ 8 ohms. ...forgot what it was at 4 ohms, but it was somewhat more.

MomurdA
03-02-2005, 12:18 PM
Thank you both for your prompt and informative answers. I thought that i might be wrong about the power to db ratio, its been a few years since i took an electronics course and had forgotten some things.

RGA
03-02-2005, 02:25 PM
You also have to remember that most people listen at distances further than 1 meter...at two meters you need more power as sound dissapates into the room -- and different speakers do this differently.

Ideally, you want a speaker that does not have an total impedence swing of more than 12ohm from minimum to maximum and ideally you don't want a very low minimum impedence because that requires big amplifiers whihc are generally more expensive and can be noisier and of course eliminates a lot of fine lower power amps. If a 4 ohm speaker doesn;t dip below 3 ohms and does not go above say 15 ohm then a lower powered amp rated for 4ohms including SETs should work fine.

If a speaker rated at 8ohm dips to 3 and goes beyond say 25 ohms in the audible range then it's probably an incompetantly designed speaker as this is a design in which the engineers have not covered their bases.

Planars are different --- usually when you sacrifice efficieny in a BOXED speaker you are making this choice for added bandwidth (more deep bass response) AND OR you are going to a smaller cabinet -- Totem tries to get a lot of bass from very small boxes as such they sacrifice sensitivity as well as efficiency and also how LOUD they can play will be lower than big beasts.

E-Stat
03-02-2005, 05:30 PM
You also have to remember that most people listen at distances further than 1 meter...at two meters you need more power as sound dissapates into the room -- and different speakers do this differently.
You have brought up an important point when comparing speakers. Indeed I know few folks who listen at a distance of one meter.

Conventional speakers (including your ANs) drop output by the inverse square law. Double the distance and get one-fourth the output. Planars, like the Magnepan and other line source speakers, however, drop output in a linear fashion.

Using Markw's chart, let's compare a conventional speaker "A" to a line source, "B" having the same efficiency driven by 64 watts:

--------------Speaker A-------------------------------------Speaker B

1 meter -----104 db-------------------------------------------104 db
2 meter------- 98 db-------------------------------------------101 db
3 meter--------92 db---------------------------------------------98 db

3 meters is about my listening distance. In this case, the conventional speaker would require four times the power (256 watts) to play at an equivalent output as the line source to make up for the 6 db loss. Or conversely, they would need to be 6 db more efficient.

Is 100 watts enough power? Like Pat says, it depends. If you listen mainly to popular music which has relatively little dynamic range, then it probably is enough. If, however, you enjoy classical music that typically has wider dynamic range swings, then you may need more. If only for the occasional peaks that require far more power. While the peak requirements of classical are higher, the average power levels, on the other hand, are lower.

rw

RGA
03-02-2005, 06:53 PM
E-Stat

This is why we need t look at a multitude of issues with speaker amp relationships and of course listen. If you know Audio Note at all you'll notice they are a little higher in the power department compared to a lot of SET makers because as sensitive as they are they are still under 100db of the big horns...and An isn;t designed for extremely large rooms other than the e which can do it --- there are limits to how loud they can play...but really because watts need to double for every 3db increase the added watts are often claimed to be useful really only for as you note dynamic range.

It's a dilemma for me though because if high watts are helpful for dynamic range then one should expect the B&W or Paradigm speakers at 90db+ in room response to sound as if they retain superior dynamics at low and higher volume levels than something like my amp and speaker combo when the amp is pumping just 10 watts and a speaker that is really only 90db away from corners. So the impedence issue is the first thing i think of here because it is far smoother on AN than it is in on any B&W Paradigm or any siimilar design to these I've seen. This is where the superiority of stats have always held court at low levels they are superb and you have no desire or need to turn the volume UP to make things out - there is not a single B&W I have ever heard that I can say that about.

but there must be something more to the game than impedence as well -- I'm not an engineer so what I am doing is observing the sound by listening and then trying to back fill information in simpler terms fior myself to sorta "get" what it is Audio Note a Boxed speaker is doing to retain the low level resolution and low level full range dynamics of the electrostats with the ability of some of the big horns to not "lose it" under complex orchestral work at loud levels with such relatively few watts. I am not suprirsed that so many stat owners I've spoken to have liked AN (no it doesn;t totally sound like a stat but it's a lot closer than any B&W like speaker which seem sorta slow and veiled and incredibly boxy sounding in comparison).

I know Peter discussed their speaker design which involves a technical approach related to pressure (which isn't the same as SPL) in their design and if you hear them against typical boxed speakers you will see or hear perhaps what they mean -- For instance if you do a listening panel session and you level match the products the AN will present a full range presentation and seem louder even if put 3-4db lower than other boxed speakers -- now that is cool. None of this is talking about sound quality which I admit is subjective but if a speaker needs to be turned UP to make things out --- it's a bad speaker IMO. Because if it is missing information at low level why would one think it's going to be there when the volume goes up?

it all goes to that 1st watt theory get it right because 10,000 bad ones just makes bad louder. It is no wonder he takes a beating from people who have never heard an entire set-up in action...it is possible that he is in fact correct in his views - the majority buys Bose and Caveliers --- but that doesn't mean their better than Quad, AN, and Ferrari.

I assume you own a stat --- when I heard my first stat --- well I understand why many would not want to go to boxed speakers.

E-Stat
03-02-2005, 07:36 PM
I assume you own a stat --- when I heard my first stat --- well I understand why many would not want to go to boxed speakers.
Yes. After first hearing Tympanis thirty years ago, I bought my first pair of Magnepans. Soon after, I heard a pair of Dayton-Wright electrostats and immediately fell for their purity and coherency. For the past twenty eight years, I've used Acoustat full range electrostats. Now I have a pair of Sound Labs U-1s on order. The sense of perceived "loudness" is very different with some speakers (although not necessarily limited to electrostats). At first blush, they do not *sound* loud, yet they measure otherwise. They are deceptively loud, just like live music.

I have heard few traditional boxed speakers capable of floating a truly open, life-sized voice in air like the best planars.

rw

markw
03-03-2005, 05:48 AM
Indeed I know few folks who listen at a distance of one meter.

Conventional speakers (including your ANs) drop output by the inverse square law. Double the distance and get one-fourth the output. Planars, like the Magnepan and other line source speakers, however, drop output in a linear fashion.

Using Markw's chart, let's compare a conventional speaker "A" to a line source, "B" having the same efficiency driven by 64 watts:

--------------Speaker A-------------------------------------Speaker B

1 meter -----104 db-------------------------------------------104 db
2 meter------- 98 db-------------------------------------------101 db
3 meter--------92 db---------------------------------------------98 db

3 meters is about my listening distance. In this case, the conventional speaker would require four times the power (256 watts) to play at an equivalent output as the line source to make up for the 6 db loss. Or conversely, they would need to be 6 db more efficient.
rwYou've never seen our little library ;). It doesn't really allow much more than 2 meters from those particular speakers (not the maggies).

The 1 meter reading is more of a method of standardizing a measurement procedure and, true, does not take into account the real world situation. I will incorporate this addendum into the next time I give this spiel.

Any thoughts to making something like this into a FAQ since it's a question that's, well, asked quite frequently??

Pat D
03-03-2005, 07:33 AM
You've never seen our little library ;). It doesn't really allow much more than 2 meters from those particular speakers (not the maggies).

The 1 meter reading is more of a method of standardizing a measurement procedure and, true, does not take into account the real world situation. It I will incorporate this addendum into the next time I give this spiel.

Any thoughts to making something like this into a FAQ?
Well, I have a thought before we get to doing a FAQ. The typical listening room provides a reverberant field rather than an anechoic chamber or free field (i.e., outdoors) and the inverse square law doesn't work so well. In other words, the sound doesn't go off into infinity but is reflected around the room until it is absorbed. Thus the intensity of the sound does not decrease by 6 dB with a doubling of distance.

I can verify this very easily with my Radio Shack Realistic Sound Level Meter Cat. No. 33-2050. I played some FM hiss at 72 dB at 10 feet from the speakers, and moved back to 20 feet and got a reading of about 68-1/2. I think in a more enclosed, less open, room, the difference would be about 3 dB.

markw
03-03-2005, 10:45 AM
... That's quite a room dependent variable. Granted, a mention of it's existance would be good but no FAQ can come up with exact numbers for an environment thatis so variable.

At least by keeping to the fairly simplistic numbers and levels that my little chart points out, augmented with E-Stat's input to compensate for distance, would provide at least a rough, workable gauge from which people could extrapolate their own answers, relative to their own particular situation.

Mentioning the effect of reverberant fields and perhaps providing an example would be good, but to try to apply meaningful numbers to a general situation would be quite difficult.

...I just hope E-stat can "prove" his numbers if asked. (private joke ;))

theaudiohobby
03-03-2005, 11:10 AM
I can verify this very easily with my Radio Shack Realistic Sound Level Meter Cat. No. 33-2050. I played some FM hiss at 72 dB at 10 feet from the speakers, and moved back to 20 feet and got a reading of about 68-1/2. I think in a more enclosed, less open, room, the difference would be about 3 dB.

I suppose that you are saying here that as the room gets smaller, the sound would decay less rapidly and vice versa?

E-Stat
03-03-2005, 11:28 AM
...I just hope E-stat can "prove" his numbers if asked. (private joke ;))
Maybe I'm just spouting too much of that theoretical objective stuff. :D

rw

RGA
03-03-2005, 01:35 PM
Yes. After first hearing Tympanis thirty years ago, I bought my first pair of Magnepans. Soon after, I heard a pair of Dayton-Wright electrostats and immediately fell for their purity and coherency. For the past twenty eight years, I've used Acoustat full range electrostats. Now I have a pair of Sound Labs U-1s on order. The sense of perceived "loudness" is very different with some speakers (although not necessarily limited to electrostats). At first blush, they do not *sound* loud, yet they measure otherwise. They are deceptively loud, just like live music.

I have heard few traditional boxed speakers capable of floating a truly open, life-sized voice in air like the best planars.

rw

You need to be very careful not to damge your hearing because with my speakers I am listening to louder levels than I realize and only realize it when I leave the room for a couple minutres come back and say yikes that's really quite loud - My Wharfedales on the other hand can play loud but they seem loud too.

I understand the choice one make for planar loudspeakers. Positioning seems to be the big issue for them but presumably if you get it right you're in luck.

E-Stat
03-03-2005, 02:07 PM
You need to be very careful not to damge your hearing...
Thanks for the concern, but you're talking to a guy in his late forties who hasn't been to a rock concert in decades. I'm also the neighborhood geek who wears ear protection when running the lawn mower, pressure washer, blower, or gas trimmer. I use both earplugs and over-the-head ear protection when I target shoot (especially with the magnum!)


My Wharfedales on the other hand can play loud but they seem loud too.
That's the phenomena I was referring to. While waiting for the Sound Labs to arrive, I'm using my vintage Double New Advents. Similarly, they sound loud when cranked.

rw

Toga
03-03-2005, 04:04 PM
Thanks for the concern, but you're talking to a guy in his late forties who hasn't been to a rock concert in decades. I'm also the neighborhood geek who wears ear protection when running the lawn mower, pressure washer, blower, or gas trimmer. I use both earplugs and over-the-head ear protection when I target shoot (especially with the magnum!)


That's the phenomena I was referring to. While waiting for the Sound Labs to arrive, I'm using my vintage Double New Advents. Similarly, they sound loud when cranked.

rw

E-stat! I'm so sad to hear that you think we're "geeks" :(

I too wear the little green foam "Pura fit-to-be-tried" -28dB inserts for all loud activities. My Colt headset is also added for range shooting, but for more practical reasons than really needing them; The RTO's at the police academy range require visible proof of hearing protection while on deck. In reality, only so much stacking of attenuation can occur before bone conduction renders high numbers a moot point. Then I suppose you could put your head in a vacuum bottle isolation sphere or something - talk about geek haha...

My thinking on the loudness exposure issue is that its probably related to a hearing phenomenon referred to as TLS (Temporary Level Shift). Since it is temporary, its kinda like a frog slowly being heated to death; if you jump out of water that is slowly heated, when you the "frog" reenter later, you notice the change. This is akin to slowly creeping the volume up and your ears slowly reducing in sensitivity. Not as noticeable, but with the TLS fading over several minutes and reentering the room, NOW you can hear the increased level with "fresh ears".

It may be that higher AVERAGE levels cause this more than wide dynamic range with loud peaks, so power compression could be the culprit. As a speaker is played at higher power, the voice coils can shoot to very high resistance levels (more than double), causing reduction in output, and making you turn up the overall level, further increasing the average. With speakers like planars, power compression is lessened by the open spacing of the conductors, and electrostatics have different mechanisms entirely, like transformer heating which have a long time constant. On the other hand, compressed power in a small speaker can make it seem louder because again it has been turned up higher to replace lost output from voice coil heating that has quite a short thermal time constant. Peaks and dips in the response can also make a speaker seem "louder" if there is a rise in the 500Hz to 2kHz range, as the human ear is most sensitive there.

Pat D
03-03-2005, 04:59 PM
I suppose that you are saying here that as the room gets smaller, the sound would decay less rapidly and vice versa?
What I was actually thinking of was that our living room is fairly open. It has an 8 foot wide opening to the kitchen and all that separates it from the rest of the house is a partial half wall (high enough to keep you from falling down the stairs!), whereas in a more enclosed room the room surfaces would reflect more of the sound energy back. My point was that, while an increase of 3 dB requires twice the power and an increase of 10 dB requires 10 times the power, the sound pressure level does not fall off with distance according to the inverse square law in a reverberant field such as most of our listening rooms due to the fact that reflected sound from the room surfaces also reaches our ears (unlike an anechoic or open field situation where get basically the dirrect sound only). It falls off more slowly than that with distance and so in reality the situation is not as bad as E-Stat's figures indicate..

I don't think reverberation time is necessarily longer in a small room because the sound paths are shorter and in a given amount of time the sound would have more chances of being absorbed by the various materials in the room, including furniture, carpets, rugs, curtains, people, etc., since it hit or pass through them more often. Big cathedrals often have long reverberation times of a few seconds, I think. One of my LP jackets contains the information that the reverberation time of the Cathedral of Freiburg is 6 seconds!

This can be quantified but we would need more of an expert for that.

E-Stat
03-03-2005, 06:14 PM
E-stat! I'm so sad to hear that you think we're "geeks" :(
Well, I'm the only person I know who seems to treasure his hearing from regular exposure to loud activities. I plug my ears if I'm with my father-in-law while he's running the circular saw.


In reality, only so much stacking of attenuation can occur before bone conduction renders high numbers a moot point.
For me, it is a quite noticeable difference to not also wear the Colts when I light off the .357.


My thinking on the loudness exposure issue is that its probably related to a hearing phenomenon referred to as TLS (Temporary Level Shift)
While I don't dispute the phenomena you noted, I simply find that some speakers always sound less LOUD than others, despite what the meter says. What I am referring to is independent of the frog acclimation effect.

rw

poneal
03-04-2005, 09:32 AM
QUOTE from RGA -- If a speaker rated at 8ohm dips to 3 and goes beyond say 25 ohms in the audible range then it's probably an incompetantly designed speaker as this is a design in which the engineers have not covered their bases.


************************************************** ******************************************
While the ideal would be to have a perfectly flat response, perfect phase tracking, and linear impedance, generally you can only get two of these close at the expense of the third. Since solid state amps have such low source impedance, fluctuations in impedance will have negligible audible effect on the resultant response. So saying that this would be an incompetent design is a false statement.
************************************************** ******************************************

Pat D
03-04-2005, 08:20 PM
QUOTE from RGA -- If a speaker rated at 8ohm dips to 3 and goes beyond say 25 ohms in the audible range then it's probably an incompetantly designed speaker as this is a design in which the engineers have not covered their bases.


************************************************** ******************************************
While the ideal would be to have a perfectly flat response, perfect phase tracking, and linear impedance, generally you can only get two of these close at the expense of the third. Since solid state amps have such low source impedance, fluctuations in impedance will have negligible audible effect on the resultant response. So saying that this would be an incompetent design is a false statement.
************************************************** ******************************************
I have no idea why one would worry about whether the impedance of a speaker got up fairly high. Many speakers do this. My old Kef 104 speakers got up to 40-70 ohms depending on the setting of the midrange control and only got slightly below 7 ohms as a minimum. They had an excellent impulse response, BTW. An amp has no difficulty at all in driving a highish impedance.

poneal
03-07-2005, 12:22 PM
was a little to quick on that response and had not thought out the process of speaker design and engineering.

RGA
03-07-2005, 07:56 PM
There was a rationale on AA as to why a smaller impedence swing was desirable -- but I don't have the time to try and locate it - I'd avoid loudspeakers with low impedence since there is no value except in causing one to need to buy a more powerful amplifier.

Pat D
03-09-2005, 04:54 PM
There was a rationale on AA as to why a smaller impedence swing was desirable -- but I don't have the time to try and locate it - I'd avoid loudspeakers with low impedence since there is no value except in causing one to need to buy a more powerful amplifier.
No low impedance speakers? I suppose that means you'll give up on recommending all those Dynaudios!

RGA
03-09-2005, 07:01 PM
Yes the Dynaudios are good speakers - but gee they're not in my living room either. The dane 42 is comparable to the AX Two the latter sounds better IMO but it's a race...the AX Two is about $250.00 less money and can run on 8 watts and a non high current amp --- the A42 on the other hand not only costs more but also requires a bigger more powerful (or a more expensive amp) which may cost a buyer, buying new, ANOTHER few hundred dollars --- and that is why I have been recommending the two with a caveat on the Dynaudio...I personally don;t like throwing money at more watts -- I'd rather throw it at better sound.

And Dynaudio is about the only exception under $1k that i've heard and liked -- and I heard and liked it with $8,000.00 worth of front end and amplification -- not sure how many people will pay that.