Sensitivity Ratings for Speakers

0 Members and 1 Guest are viewing this topic. Read 5211 times.

Rick Craig

  • Industry Participant
  • Posts: 3680
  • Selah Audio
    • http://www.selahaudio.com
Sensitivity Ratings for Speakers
« on: 23 Feb 2016, 03:01 pm »
I thought this might be beneficial to post, especially for those looking to make future purchases. It's an excerpt of my reply in an email. It's not intended to single out the mentioned manufacturer; in fact, there are far worse examples that could've been mentioned.

The speaker in the picture I sent is 91.5db in sensitivity. We follow the NRC (widely recognized as one of the top speaker testing places in the world) guidelines and specify the sensitivity as an average over 300hz-3K. SoundStage magazine uses the NRC and I found two Magico speakers that were tested by them. The V2 was factory rated at 89db but the NRC number was 86db. The S5 was listed as 89db by Magico  and rated 87db at the NRC. This isn't unusual; in fact, there are examples where the manufacturers were overstating a speaker's sensitivity by 4-6db or even more.

The main problem is that there's no standard (such as the FTC power ratings for amplifiers) so manufacturers often take the liberty to quote a number that benefits them the most for sales. Sometimes it's either at 1K or whatever frequency has the greatest number. This makes it difficult for buyers to make comparisons, especially in cases where one speaker is honestly portrayed and the other overstated. The speakers I rate at 87db would typically be listed at 90db or higher by many manufacturers.

Guy 13

Re: Sensitivity Ratings for Speakers
« Reply #1 on: 23 Feb 2016, 03:48 pm »
I thought this might be beneficial to post, especially for those looking to make future purchases. It's an excerpt of my reply in an email. It's not intended to single out the mentioned manufacturer; in fact, there are far worse examples that could've been mentioned.

The speaker in the picture I sent is 91.5db in sensitivity. We follow the NRC (widely recognized as one of the top speaker testing places in the world) guidelines and specify the sensitivity as an average over 300hz-3K. SoundStage magazine uses the NRC and I found two Magico speakers that were tested by them. The V2 was factory rated at 89db but the NRC number was 86db. The S5 was listed as 89db by Magico  and rated 87db at the NRC. This isn't unusual; in fact, there are examples where the manufacturers were overstating a speaker's sensitivity by 4-6db or even more.

The main problem is that there's no standard (such as the FTC power ratings for amplifiers) so manufacturers often take the liberty to quote a number that benefits them the most for sales. Sometimes it's either at 1K or whatever frequency has the greatest number. This makes it difficult for buyers to make comparisons, especially in cases where one speaker is honestly portrayed and the other overstated. The speakers I rate at 87db would typically be listed at 90db or higher by many manufacturers.

 :thumb:

srb

Re: Sensitivity Ratings for Speakers
« Reply #2 on: 23 Feb 2016, 03:59 pm »
Really no reason to always entirely quote the preceding post just to add an in-agreement thumbs up emoji (other than to provide visual distraction or take up space).

paul79

  • Industry Participant
  • Posts: 901
Re: Sensitivity Ratings for Speakers
« Reply #3 on: 23 Feb 2016, 04:05 pm »
The above post, and this post, are taking up space.........................

Rick Craig

  • Industry Participant
  • Posts: 3680
  • Selah Audio
    • http://www.selahaudio.com
Re: Sensitivity Ratings for Speakers
« Reply #4 on: 23 Feb 2016, 04:07 pm »
I forgot to add that my response was to a current Magico owner who has a tube amp and is looking for a speaker with higher sensitivity. That's why I looked up the Soundstage information to see what Magico models they had tested.

srb

Re: Sensitivity Ratings for Speakers
« Reply #5 on: 23 Feb 2016, 04:19 pm »
Yes, I've had a few speakers with quoted "higher" sensitivity than the speakers they replaced, but actually played at a lower volume (-3dB to -5dB) using the same amplifier and volume control position.

Steve

undertow

  • Full Member
  • Posts: 895
Re: Sensitivity Ratings for Speakers
« Reply #6 on: 4 Mar 2016, 09:03 pm »
No question most speakers are not only over rated in sensitivity, but several other factors as well.

Most of these so called high end companies like to build 86 db speakers, but that rating is also at 1000 khz. Guess what the low frequencies normally suffer even more. They have a wide bandwidth of lets say 20 hz, to 20 khz, but fact is they are not even 80 db at 30 hz, and require hundreds of watts to even achieve that at normal listening levels. Even 40 hz in most cases they drop off like a rock, and create very unbalanced speakers which is why you always see complaints about a lot of harsh mids or highs, and nearly non-existent lows in some speaker designs. But that is mostly what a lot of the off the shelf modern drivers will produce.

Yeah if you have 1000 watt mono blocks pumping high volume gain music thru them you might get that 20 hz to light up a little in a conventional full range speaker, but it is far more work, and cost than necessary, and likely at that point your just creating as much distortion as you are true useable output.

This is why my opinion you need to buy speakers more like 98 db and they only go down to 50 hz or so used with a 40 or 50 watt amp, then put your money into a good sub if you need true subsonic levels with authority. Rarely can you get these big passive towers to really have any efficiency down low easily, or cheaply enough to make them worthwhile in a real world environment vs. testing on a waterfall plot.

There are many "Wonder" speakers out there that claim to touch 32 hz with 95 to 99 db efficiency, and they very well may, but not unless your pushing them at ear bleeding levels in a perfect acoustic room does this frequency response become very useable or fill a room somewhere in their conventional designed house for enjoyable listening.

If you hear a high powered P.A. system live they have some ridiculous sensitivity with horns etc... up to 106 db, but only rate down to 60 hz yet produce far more useable bass at high power which put most peoples 18 hz subs to shame.

Hence many people in this hobby will be disappointed with a lot of the results they expect from various claims. There is a big difference between measured "Electrical" sensitivity, and true "Acoustic output" sensitivity.
« Last Edit: 4 Mar 2016, 11:45 pm by undertow »

JLM

  • Full Member
  • Posts: 10668
  • The elephant normally IS the room
Re: Sensitivity Ratings for Speakers
« Reply #7 on: 4 Mar 2016, 11:40 pm »
Two more considerations:

1.)  What is the speaker's impedance vs. frequency curve like?

2.)  How much power can they handle?  The often stated peaks of 105 dB for classic/jazz are dismissed by most who say they never listen that loud, but 105 dB is instantaneous peaks and are needed to overall provide realistic, life-like performance.  And isn't that what high-end audio is all about?  Average levels will be much lower, in line with levels we "normally" listen at.  Point being, an 85 dB/w/m speaker, in an "average" room may have to handle 1,000 - 2,000 watts (30 dB of gain) to reach those peaks.  How many of us own amps that big or speakers that are rated to handle that?

Rick Craig

  • Industry Participant
  • Posts: 3680
  • Selah Audio
    • http://www.selahaudio.com
Re: Sensitivity Ratings for Speakers
« Reply #8 on: 7 Mar 2016, 08:49 pm »
No question most speakers are not only over rated in sensitivity, but several other factors as well.

Most of these so called high end companies like to build 86 db speakers, but that rating is also at 1000 khz. Guess what the low frequencies normally suffer even more. They have a wide bandwidth of lets say 20 hz, to 20 khz, but fact is they are not even 80 db at 30 hz, and require hundreds of watts to even achieve that at normal listening levels. Even 40 hz in most cases they drop off like a rock, and create very unbalanced speakers which is why you always see complaints about a lot of harsh mids or highs, and nearly non-existent lows in some speaker designs. But that is mostly what a lot of the off the shelf modern drivers will produce.

Yeah if you have 1000 watt mono blocks pumping high volume gain music thru them you might get that 20 hz to light up a little in a conventional full range speaker, but it is far more work, and cost than necessary, and likely at that point your just creating as much distortion as you are true useable output.

This is why my opinion you need to buy speakers more like 98 db and they only go down to 50 hz or so used with a 40 or 50 watt amp, then put your money into a good sub if you need true subsonic levels with authority. Rarely can you get these big passive towers to really have any efficiency down low easily, or cheaply enough to make them worthwhile in a real world environment vs. testing on a waterfall plot.

There are many "Wonder" speakers out there that claim to touch 32 hz with 95 to 99 db efficiency, and they very well may, but not unless your pushing them at ear bleeding levels in a perfect acoustic room does this frequency response become very useable or fill a room somewhere in their conventional designed house for enjoyable listening.

If you hear a high powered P.A. system live they have some ridiculous sensitivity with horns etc... up to 106 db, but only rate down to 60 hz yet produce far more useable bass at high power which put most peoples 18 hz subs to shame.

Hence many people in this hobby will be disappointed with a lot of the results they expect from various claims. There is a big difference between measured "Electrical" sensitivity, and true "Acoustic output" sensitivity.

Most of the higher sensitivity pro-style drivers don't have the tonal quality of the better drivers typically used in home audio. That's partially what pushed me towards doing line arrays because I can get the higher output without the problems that often plague the higher sensitivity drivers.

undertow

  • Full Member
  • Posts: 895
Re: Sensitivity Ratings for Speakers
« Reply #9 on: 7 Mar 2016, 08:50 pm »
Most of the higher sensitivity pro style drivers don't have the tonal quality of the better drivers typically used in home audio. That's partially what pushed me towards doing line arrays because I can get the higher output without the problems that often plague the higher sensitivity drivers.

Agreed. I was not suggesting that using pro drivers was the answer just illustrating what is accomplished in the real world.

G Georgopoulos

  • Restricted
  • Posts: 1253
Re: Sensitivity Ratings for Speakers
« Reply #10 on: 8 Mar 2016, 12:35 am »
Most of the higher sensitivity pro-style drivers don't have the tonal quality of the better drivers typically used in home audio.

 :lol:

Rick Craig

  • Industry Participant
  • Posts: 3680
  • Selah Audio
    • http://www.selahaudio.com
Re: Sensitivity Ratings for Speakers
« Reply #11 on: 8 Mar 2016, 02:27 am »
Two more considerations:

1.)  What is the speaker's impedance vs. frequency curve like?

2.)  How much power can they handle?  The often stated peaks of 105 dB for classic/jazz are dismissed by most who say they never listen that loud, but 105 dB is instantaneous peaks and are needed to overall provide realistic, life-like performance.  And isn't that what high-end audio is all about?  Average levels will be much lower, in line with levels we "normally" listen at.  Point being, an 85 dB/w/m speaker, in an "average" room may have to handle 1,000 - 2,000 watts (30 dB of gain) to reach those peaks.  How many of us own amps that big or speakers that are rated to handle that?

105db peaks shouldn't be a problem with a decent size amp and quality speakers with 85db sensitivity. Short-term peaks are possible with both well in excess of continuous wattage ratings but not necessarily needing to be 1K-2K watts.

JLM

  • Full Member
  • Posts: 10668
  • The elephant normally IS the room
Re: Sensitivity Ratings for Speakers
« Reply #12 on: 8 Mar 2016, 10:39 am »
Tube amp fans boast about the soft/friendly clipping characteristics of tubes and how they can be driven well into measured distortion with little effect to the sound produced.  OTOH the "common knowledge" is that solid state clips hard and offers little extra headroom.  From my experience this seems to be true and while I rarely listen loud, I don't want to risk ruining my speakers.

So Rick, I suppose it varies by subsets of tube and solid state designs, but what is a fuller, more accurate description of amps ability to handle peaks?

Rick Craig

  • Industry Participant
  • Posts: 3680
  • Selah Audio
    • http://www.selahaudio.com
Re: Sensitivity Ratings for Speakers
« Reply #13 on: 8 Mar 2016, 05:08 pm »
Agreed. I was not suggesting that using pro drivers was the answer just illustrating what is accomplished in the real world.

I forgot to add that moving a huge amount of air at say 50-60hz can be pretty impressive.