|
|
Post by Ex_Vintage on Apr 27, 2020 21:35:40 GMT -5
Since this is a standard definition (dB SPL at 1 meter for 2.83V). I have an observation. I have some ELAC UF5 mains rated at 4 ohms and 85dB and recently purchased a pair of Klipsch RP-600M bookshelves rated at 8 ohms and 96 db. One would think that given the same source input level, the same level settings for the pre-amp and main amp that there would be a noticeable "loudness" difference, but I cannot say that has been my observation. I have exercised both speakers over a variety of music tracks and as I stated, I cannot say that I sense a real difference in sound level. This seems to disagree with the general statement that a high sensitivity speaker would require less amplifier to drive them. Thoughts?
|
|
|
|
Post by frenchyfranky on Apr 27, 2020 22:18:28 GMT -5
For sure you can reach the same listening output level at approx. the same volume setting level but your power amplifier will give a lot more output wattage with the 4ohms 85dB speakers for achieving this result.
|
|
|
|
Post by AudioHTIT on Apr 27, 2020 23:54:06 GMT -5
The reason the standard evolved from 1 watt @ 1 meter, to 2.83 volts a 1 meter (1 watt @ 8 ohms), was to eliminate speaker impedance from the calculation. But I agree, and would expect to hear a ‘more than noticeable’ difference between speakers with greater than 10 dB higher sensitivity (the subjective difference for ‘twice as loud’). If you are using the same electronics at the same levels, I’m puzzled as well, or the ratings are faulty. The best way to know for sure would be to do your own measurements.
|
|
|
|
Post by 405x5 on Apr 28, 2020 7:40:46 GMT -5
Since this is a standard definition (dB SPL at 1 meter for 2.83V). I have an observation. I have some ELAC UF5 mains rated at 4 ohms and 85dB and recently purchased a pair of Klipsch RP-600M bookshelves rated at 8 ohms and 96 db. One would think that given the same source input level, the same level settings for the pre-amp and main amp that there would be a noticeable "loudness" difference, but I cannot say that has been my observation. I have exercised both speakers over a variety of music tracks and as I stated, I cannot say that I sense a real difference in sound level. This seems to disagree with the general statement that a high sensitivity speaker would require less amplifier to drive them. Thoughts? Did you swap and test both pair in the same physical location? |
|
|
|
Post by Ex_Vintage on Apr 28, 2020 7:55:21 GMT -5
Since this is a standard definition (dB SPL at 1 meter for 2.83V). I have an observation. I have some ELAC UF5 mains rated at 4 ohms and 85dB and recently purchased a pair of Klipsch RP-600M bookshelves rated at 8 ohms and 96 db. One would think that given the same source input level, the same level settings for the pre-amp and main amp that there would be a noticeable "loudness" difference, but I cannot say that has been my observation. I have exercised both speakers over a variety of music tracks and as I stated, I cannot say that I sense a real difference in sound level. This seems to disagree with the general statement that a high sensitivity speaker would require less amplifier to drive them. Thoughts? Did you swap and test both pair in the same physical location? Exactly. I have a default "start up" volume on the XMC-1 of -30. I have a streamer, Bluesound Node 2i, that I can adjust the volume on. With the streamer app set all the way up and with the speakers in exactly the same locations, I can't tell a significant difference in overall loudness. The ohm rating of the speakers should not matter in this because the output voltage of the amplifier should be the same and the difference in sensitivity (dB) should be noticeable. |
|
|
|
Post by Boomzilla on Apr 28, 2020 8:08:30 GMT -5
Klipsch probably exaggerates their speakers' sensitivity. Also, if the difference in acoustic output is LESS than 6dB (a doubling of perceived loudness), you may not notice it.
And unless you're putting a VOM across the speaker terminals and measuring the AC voltage going to each speaker, identical volume knob settings are unreliable due to differences in speaker impedance.
Finally, if you want to measure the relative volumes of the two speakers rather than relying on your ears, free apps are available for cell phones that will show the actual SPL generated. Note that the cell phone measurements may not be absolutely accurate but they DO NOT NEED TO BE. The cell phone will be equally inaccurate when measuring both speakers. It's the difference in SPLs that you want to know - not the actual loudness.
|
|
|
|
Post by 405x5 on Apr 28, 2020 8:44:43 GMT -5
Did you swap and test both pair in the same physical location? Exactly. I have a default "start up" volume on the XMC-1 of -30. I have a streamer, Bluesound Node 2i, that I can adjust the volume on. With the streamer app set all the way up and with the speakers in exactly the same locations, I can't tell a significant difference in overall loudness. The ohm rating of the speakers should not matter in this because the output voltage of the amplifier should be the same and the difference in sensitivity (dB) should be noticeable. Unfamiliar with the two speakers being compared for sensitivity, I just took a close look at both. Too radically different in driver complement, internal cabinet volume and room boundaries to draw the conclusions you’re getting at. |
|
|
|
Post by leonski on May 4, 2020 12:02:24 GMT -5
The ONE data point I do have Agrees that Klipsch may beef up their sensitivity rating. The Forte III is spec'd at 99db but measured by Stereophile was OVER 3db less.
Just like many other specs, this one has a couple different ways to measure. The 2 most common would be 'In-Room' and 'Anachoic'. In room may measure higher due to
reflections. Stereophile measures in-room BUT time-limits the measurement to minimize reflections.....
You could also take the speaker OUTDOORS to measure, which is what some DIY types end up doing.
Also? 2.83volts into a 4ohm speaker is 2 watts. NOT 1.
The ONLY way to know how good your measureing system IS is to measure it. I'm not going to go into detail, but it is an involved process to take a LOT of measurments of the SAME thing
followed by 'crunching the numbers'. This is called a 'capability study'.
Boom refers to 'equally inaccureate' when talking about measuring with a cell phone. I think it is right to talk about Repeatability. When given a given input, what is the OUTPUT of the measuring system?
And more importantly, How Much does it vary from test to test?
|
|
DYohn
Emo VIPs 
Posts: 18,489
|
Post by DYohn on May 4, 2020 12:53:26 GMT -5
I wrote this for my car audio forum back in 2006. It seems appropriate to repost in this thread.
The sensitivity of a loudspeaker is the medium sound pressure between 125 Hz to 4 kHz using a constant voltage at a specific distance, measured in dB per watt and meter. Usually 2.83 volts is used as this corresponds to 1 watt @ 8 ohms, with the SPL measured at 1 meter.
For instance, the sensitivity of an 8-ohm loudspeaker might be listed as 92 dB / 2.83 volts / 1 meter. In this case the power (P = V squared / R) = 2.83 X 2.83 / 8 = 1 watt. In a 4-ohm loudspeaker 2.83 volts will generate 2 watts. To get the actual reference value at 1 watt, you have to subtract 3 dB (power cut in half) from the sensitivity measured; if the speaker was measured at 2-ohms, you subtract 6db, etc. Some loudspeaker manufacturers fail to include this correction. Some cheat it by measuring at 0.5 meter. Some intentionally fudge the numbers or the test voltage. Also since the value is measured at frequencies higher than normal subwoofer usage bands, it may or may not be 100% accurate below 80Hz. Good manufacturers will do it right and give you accurate numbers, and often will include specifics about the test such as if a subwoofer was measured using standard frequency sweeps or using sub-woofer sweeps below 80Hz.
But there are other factors contributing to the overall system efficiency and resulting system output levels. The speaker enclosure adds several SPL components, especially in a ported system. The vehicle has what’s called “cabin gain” (or "room gain" in a listening space that is not a car) made up from the many resonances set up in the closed space. These function as a resonance chamber similar to what happens inside a loudspeaker enclosure. These might increase overall loudspeaker-room system sensitivity value by as much as 6db or even more, and increase the overall system efficiency from the lowly speaker by 1 or 2% to 50% or 60% or more. In-room or in-vehicle placement and measurements often mean much more to what you hear than simple system ratings.
A calculation beginning with the speaker’s sensitivity rating and working up by adding +3db for every doubling of input power will only tell you part of the story. (Reminder that the raw SPL difference between 1 watt and 1000 watts is +30db.) It only indicates what the speaker would do in a free air anechoic test chamber at 1 meter. A better method is to test the system after installation and determine the overall system sensitivity including all space variables. Then a calculation to determine required wattage to achieve a particular dbSPL value will have much more meaning. Also realize that SPL decreases with distance at a rate of -6db for each doubling of distance. So if the output is 87 db at one meter, at 2 meters the output will be 81 db. It is critical to know where the test mic will be positioned in relation to the speakers to calculate a predicted output level.
By the way, loudspeaker efficiency (eta-naught) is a measure of how much incoming electrical energy is converted into acoustic energy. Often speaker manufacturers will not give you this value, but for good quality loudspeakers a typical value might be between 0.2 % and 2% (only 0.2-2% of the amplifier signal is actually being converted into sound.) The rest is being wasted as heat. Some high efficiency speakers might get as high as 10% or 15%, and some super-high efficiency systems like compression drivers designed for pro audio can be almost 100% efficient (100% efficiency is 112db/watt-meter.) You can convert the listed sensitivity number into an efficiency using this formula: Efficiency = 10E(Sensitivity in dB - 112)/10
|
|
|
|
Post by Ex_Vintage on May 4, 2020 13:11:41 GMT -5
Also? 2.83volts into a 4ohm speaker is 2 watts. NOT 1. The ONLY way to know how good your measureing system IS is to measure it. I'm not going to go into detail, but it is an involved process to take a LOT of measurments of the SAME thing followed by 'crunching the numbers'. This is called a 'capability study'. Boom refers to 'equally inaccureate' when talking about measuring with a cell phone. I think it is right to talk about Repeatability. When given a given input, what is the OUTPUT of the measuring system? And more importantly, How Much does it vary from test to test? I realize all of the specifics of doing Gauge R&R studies and establishing Cpk numbers for measurement system analysis, but this is fairly straight forward. In controlling (or not changing) virtually all of the relevant variables (input levels, room reflections) the only thing that is changed is the speakers. The pre-amp / amplifier are a voltage source and were set at the same levels for both speakers. The difference in stated sensitivity of 11dB should produce a noticeable sound level difference. I guess my only take away is that many statements have been made about a more sensitive / efficient speaker requiring a lower wattage amplifier, and that is not my observation. |
|
|
|
Post by leonski on May 4, 2020 13:31:44 GMT -5
I wrote this for my car audio forum back in 2006. It seems appropriate to repost in this thread. The sensitivity of a loudspeaker is the medium sound pressure between 125 Hz to 4 kHz using a constant voltage at a specific distance, measured in dB per watt and meter. Usually 2.83 volts is used as this corresponds to 1 watt @ 8 ohms, with the SPL measured at 1 meter. For instance, the sensitivity of an 8-ohm loudspeaker might be listed as 92 dB / 2.83 volts / 1 meter. In this case the power (P = V squared / R) = 2.83 X 2.83 / 8 = 1 watt. In a 4-ohm loudspeaker 2.83 volts will generate 2 watts. To get the actual reference value at 1 watt, you have to subtract 3 dB (power cut in half) from the sensitivity measured; if the speaker was measured at 2-ohms, you subtract 6db, etc. Some loudspeaker manufacturers fail to include this correction. Some cheat it by measuring at 0.5 meter. Some intentionally fudge the numbers or the test voltage. Also since the value is measured at frequencies higher than normal subwoofer usage bands, it may or may not be 100% accurate below 80Hz. Good manufacturers will do it right and give you accurate numbers, and often will include specifics about the test such as if a subwoofer was measured using standard frequency sweeps or using sub-woofer sweeps below 80Hz. But there are other factors contributing to the overall system efficiency and resulting system output levels. The speaker enclosure adds several SPL components, especially in a ported system. The vehicle has what’s called “cabin gain” (or "room gain" in a listening space that is not a car) made up from the many resonances set up in the closed space. These function as a resonance chamber similar to what happens inside a loudspeaker enclosure. These might increase overall loudspeaker-room system sensitivity value by as much as 6db or even more, and increase the overall system efficiency from the lowly speaker by 1 or 2% to 50% or 60% or more. In-room or in-vehicle placement and measurements often mean much more to what you hear than simple system ratings. A calculation beginning with the speaker’s sensitivity rating and working up by adding +3db for every doubling of input power will only tell you part of the story. (Reminder that the raw SPL difference between 1 watt and 1000 watts is +30db.) It only indicates what the speaker would do in a free air anechoic test chamber at 1 meter. A better method is to test the system after installation and determine the overall system sensitivity including all space variables. Then a calculation to determine required wattage to achieve a particular dbSPL value will have much more meaning. Also realize that SPL decreases with distance at a rate of -6db for each doubling of distance. So if the output is 87 db at one meter, at 2 meters the output will be 81 db. It is critical to know where the test mic will be positioned in relation to the speakers to calculate a predicted output level. By the way, loudspeaker efficiency (eta-naught) is a measure of how much incoming electrical energy is converted into acoustic energy. Often speaker manufacturers will not give you this value, but for good quality loudspeakers a typical value might be between 0.2 % and 2% (only 0.2-2% of the amplifier signal is actually being converted into sound.) The rest is being wasted as heat. Some high efficiency speakers might get as high as 10% or 15%, and some high efficiency systems like compression drivers designed for pro audio can be almost 100% efficient (100% efficiency is 112db/watt-meter.) You can convert the listed sensitivity number into an efficiency using this formula: Efficiency = 10E(Sensitivity in dB - 112)/10 Somewhere I've run across a calculator to convert sensitivity to efficiency and reverse. Speakers ARE awful! All I was pointing out in my remarks about measurement is that what you measure ain't necessarily so. MANY factors interfere, from operator to system variability. I would also only measure ONE speaker at a time. Going thru the whole Gage R+R thing for any reason short of either ISO or manufacturing necessity is a LOT over the top..... When I did such things for my work, I always had a problem securing HELP to do the measurments. You need 3 people to do the measurments. As for speakers rated far differently having similiar perceived outputs? If you believe the sensitivity ratings...and they are say.....greater than 6db apart, and you can't tell 'em apart? I'd say you have either a measrement or technique problem. Careful detailing of the system may help. And don't forget that 2 speakers even with 'identical' sensitivity may sound different AND will have enough electrical differences to perhaps make a difference when doing listening tests......Not to mention how the amplifier drives the 2 different loads. NONE of which resemble a resistor. I think most would agree that manufacturers numbers are to be taken as 'advisory' not carved in stone. |
|
