Question about speaker specifications

C

cptslow

Audioholic Intern
Hi Guys,
I am planning to upgrade my bookshelf speakers , I am moving to a bigger apartment soon and also I am expecting some change back from uncle sam after i get done with my taxes, so i thought i would invest on a nicer pair of speakers.
I am looking at the specifications of speakers from several manufactures, and i see the frequency response being noted. I had a few question hoping to find answers.
-- How is the frequency response measured ?
--Is a sine wave used to measure the response or are the speaker connected to real music input.
I am guessing they have specifice test room that they place the speakers in to carry out the test.
--Is a microphone placed in front of the speaker to measure the response.?
--How far and where is the sound levels measured ?

I really appreciate any input about this.

--Thanks
 
Seth=L

Seth=L

Audioholic Overlord
Hi Guys,
I am planning to upgrade my bookshelf speakers , I am moving to a bigger apartment soon and also I am expecting some change back from uncle sam after i get done with my taxes, so i thought i would invest on a nicer pair of speakers.
I am looking at the specifications of speakers from several manufactures, and i see the frequency response being noted. I had a few question hoping to find answers.
-- How is the frequency response measured ?
--Is a sine wave used to measure the response or are the speaker connected to real music input.
I am guessing they have specifice test room that they place the speakers in to carry out the test.
--Is a microphone placed in front of the speaker to measure the response.?
--How far and where is the sound levels measured ?

I really appreciate any input about this.

--Thanks
The most simple answer is that most if not all speaker manufacturers measure their speakers differently.

From my limited understanding ideally measurements should be done (outside) mic'd at 1 meter from the front of the speaker (I don't know of any speaker manufacturer that does this). Speaker manufacturers usually do this in "a room", sometimes at arbitrary distances that aren't 1 meter or directly in front of the speaker, or use multiple measured responses from around the room to make a much nicer (on paper) response, or just give no parameters at all like Definitive Technology's almost worthless published specifications. If any of this is outlandishly incorrect, someone please correct me.

A good few speaker manufacturers use Anechoic chambers (an acoustically deadened room) to perform their measurements. This is clearly superior to just putting a speaker in a random room that won't even closely resemble the room in your home that you would be using the speakers in. This will measure how the speaker behaves on it's own, how the drivers react with the cabinet, crossover, other drivers in the cabinet, etc... The caveat is that you still won't know exactly how the speaker will behave in your room. Often times a speaker that measures say 50hz-20Khz +/-3dB 1 watt 1 meter in their chamber may have greater bass extension but reduced high frequency output in your listening room.

If you're genuinely curious about how a speaker performs based on measured response I recommend looking to 3rd party measurements in order to do your comparisons, preferably from the same source using the same conditions.
 
Swerd

Swerd

Audioholic Warlord
How is the frequency response measured? Is a sine wave used to measure the response or are the speaker connected to real music input?
A combination of many sine wave tones, called pink noise, is used. Pink noise has a frequency spectrum where each octave carries an equal amount of acoustic power. Music would not be used.
I am guessing they have specifice test room that they place the speakers in to carry out the test.
Yes and no. There are specific rooms, called anechoic rooms, built to have no reflective surfaces. They are ideal for this purpose, but they are expensive to build and few of them exist. It is also possible, with the appropriate computer software and a typical home computer, to perform useful measurements in a normal room. This is often called simulated anechoic response. Edit: This computer-simulated anechoic response is often called a quasi-anechoic response.
Is a microphone placed in front of the speaker to measure the response?
Yes, a test microphone is placed on a stand. These test microphones are specially designed to have as broad and flat a frequency response as possible.
How far and where is the sound levels measured?
They are placed 1 meter away from the speaker. Typically they will be placed directly in front of a speaker (on-axis) and several defined locations off-axis, such as 15°, 30°, 45°, 60°, etc.

The results can be graphed with frequency on the horizontal axis and loudness (sound pressure level or SPL) on the vertical axis. Several traces can be combined onto one plot showing the responses on-axis and off-axis.

Good examples of such measurements can be viewed here http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=16&Itemid=18. The Canadian National Research Council has an anechoic room and is well respected for its accurate speaker measurements.
 
Last edited:
Swerd

Swerd

Audioholic Warlord
As Seth=L explained, some speaker makers will state the measured performance as a range, such as 50 Hz-20 KHz ± 3 dB, where the microphone was 1 meter away and the signal was delivered at 1 watt (2.83 volts for an 8 ohm load).

But frequency response curves, such as the two shown below, will tell much more useful information. Which would you rather listen to? (Hint: flat over as wide a frequency range as possible is better)

B&W CM1 (see http://www.soundstagenetwork.com/measurements/speakers/bw_cm1/)


or a NHT Classic Three (http://www.soundstagenetwork.com/measurements/speakers/nht_classic_three/)
 
Last edited:
C

cptslow

Audioholic Intern
Thank you for the response. Swerd..i am looking at the frequency response of the NHT, the curve looks a lot flater as compared to the CM9. the CM9 are a more expensive speakers as compared to NHT, but still i see a big dip in the curve around 2k and after that the curve is lets say not flat at all. Does this mean that i may hear certain tones or notes in the music less loud as compared to others ? The frequency response of the CM9 not being flat and huge dip at 2k, would that be because of the poor crossover/filter design ?
 
Swerd

Swerd

Audioholic Warlord
Many people say they have trouble understanding the impact of the details shown in these frequency response curves. It does take some practice and effort to learn what visible features of these curves correspond to what a speaker sounds like.

But if you compare curves of two different speakers, measured by the same method, as is the case with the Canadian NRC measurements, it can be easy to guess which might sound better in any room.

Another good source of reliably done speaker measurements is the collection of speaker reviews done by Stereophile http://www.stereophile.com/equipment-reviews. Usually the measurements are done carefully by the same methods and allow comparison of different speakers. I don't pay attention to the written comments made by the subjective reviewer, who rarely, if ever, does the speaker measurements.
 
Swerd

Swerd

Audioholic Warlord
Thank you for the response. Swerd..i am looking at the frequency response of the NHT, the curve looks a lot flater as compared to the CM9. the CM9 are a more expensive speakers as compared to NHT, but still i see a big dip in the curve around 2k and after that the curve is lets say not flat at all. Does this mean that i may hear certain tones or notes in the music less loud as compared to others ? The frequency response of the CM9 not being flat and huge dip at 2k, would that be because of the poor crossover/filter design ?
Where is the frequency response curve of the CM9? Do you have a link?
 
Swerd

Swerd

Audioholic Warlord
Thank you for the response. Swerd..i am looking at the frequency response of the NHT, the curve looks a lot flater as compared to the CM1. the CM1 are a more expensive speakers as compared to NHT, but still i see a big dip in the curve around 2k and after that the curve is lets say not flat at all. Does this mean that i may hear certain tones or notes in the music less loud as compared to others ? The frequency response of the CM9 not being flat and huge dip at 2k, would that be because of the poor crossover/filter design ?
The B&W CM1 is a small 2-way speaker with a crossover at 4 kHz. (I think that's too high, but that's another story.) The big dip, centered around 2 kHz, is probably not caused by the crossover. I don't know if it was intended by the designers or not, but it is clearly large enough (more than 5 dB) to hear in music. Depending on the music, it might make it sound "laid back", "less than bright", or "dull". It will swallow details in music, and it may negatively impact on imaging by the speaker.

In some other British speakers, such a dip in the upper mid range was intentional, and often called "the BBC Dip". If not so deep, such a dip can create speakers with a "polite" or "refined" sound as opposed to a "bright" or "in your face" sound.

As you noted, price can have little to do with it.
 
Swerd

Swerd

Audioholic Warlord
You should also note that the CM1 has an exaggerated bass response. See the hump from 70 to 150 Hz. This was more than likely intended by the designers, giving greater bass to a small bookshelf speaker.

This bass hump is as much as 4 dB greater than the sound in the fairly flat region from 200 to 1000 Hz and should be noticeable. I can't tell from looking at the curve whether it merely sounds noticeable or objectionable. It might be enough added emphasis to create muddy sounding bass.
 
Last edited:
C

cptslow

Audioholic Intern
You should also note that the CM1 has an exaggerated bass response. See the hump from 70 to 150 Hz. This was more than likely intended by the designers, giving greater bass to a small bookshelf speaker.

This bass hump is as much as 4 dB greater than the sound in the fairly flat region from 200 to 1000 Hz and should be noticeable. I can't tell from looking at the curve whether it merely sounds noticeable or objectionable. It might be enough added emphasis to create muddy sounding bass.
But how do the designers design a driver that has a higher response for a specific frequecy ( in the case the 70-150 hz range).
 
Seth=L

Seth=L

Audioholic Overlord
But how do the designers design a driver that has a higher response for a specific frequecy ( in the case the 70-150 hz range).
I don't think you'll find that drivers are often designed for that range. A speaker system may be tuned to that range in a very odd 3 way design.
 
M

markw

Audioholic Overlord
One thing to keep in mind when analyzing these charts, particularly when looking at world class speakers: They are most likely designed for a particular sound profile.

Why? while you may think you can tell what sounds good, one doesn't really know what sounds good to someone without listening.

While you might think there's a drastic difference between the NHT and B&W speakers, consider that both are loved by many, many people. Personal preference takes precedence over what a graph can tell you.

Simply put, trust your ears rather than your eyes for your final judgment when it comes to judging and selecting speakers.
 
Swerd

Swerd

Audioholic Warlord
But how do the designers design a driver that has a higher response for a specific frequecy ( in the case the 70-150 hz range).
It wouldn't be in the design of the driver itself. The response in the lower range of a woofer is determined by the cabinet design, specifically the volume of the box and the dimensions (length & diameter) of the port tube. With different dimensions (a larger box), this hump could be flattened, but the overall bass response of the speaker would be noticeable less.

markw is right when he said this:
One thing to keep in mind when analyzing these charts, particularly when looking at world class speakers: They are most likely designed for a particular sound profile.
I prefer to believe that the marketing people at B&W had an identifiable company product-line sound in mind. And their decision outweighed that of the speaker designers who might have had other ideas.

Similarly, NHT has a company identity of speakers with an overall neutral tone balance and flat frequency response.

My cup of tea is definitely the NHT type of sound. Different strokes for different folks. If you can read the graphs, and correlate their patterns to sounds you already have heard, you can identify speakers you might like faster.
 
Swerd

Swerd

Audioholic Warlord
BSA provided a link for the B&W CM9 speaker. Thanks.

The CM9 is a floor-standing 3-way speaker with 2 woofers, a mid range, and tweeter, with crossover frequencies of 350 Hz and 4 kHz. It is part of the same speaker series as the smaller CM1 bookshelf speaker shown above.

The measurements were done by Sound & Vision and show:
"the quasi-anechoic (employing close-miking of all woofers) frequency response of the CM9 L/R (purple trace), ASW 10CM subwoofer (blue trace), CM Centre 2 center channel (green trace), and CM5 surround (red trace). All passive loudspeakers were measured with grilles at a distance of 1 meter with a 2.83 volt input and scaled for display purposes.

The CM9’s listening-window response (a 5-point average of axial and ±15° horizontal & vertical responses) measures +0.83/–5.12 dB from 200 Hz to 10 kHz. The –3 dB point is at 60 Hz, and the –6 dB point is at 49 Hz."​

Note that these curves appear to be more smoothed than those from the Canadian NRC. S&V makes its measurements with the speaker grilles attached, unlike the Canadian NRC or Stereophile.

Interestingly, the CM9 also appears to have a broad dip in treble response, beginning above 1.5 kHz, centered at 5-6 kHz, and extending all the way to 9 or 10 kHz. The CM2 (green trace) and CM5 (red trace) both share this general feature. It wouldn't be a stretch to say that this treble dip is an intended feature of the entire CM series.

 
Last edited:
C

cptslow

Audioholic Intern
Thank you guys for all the details,I really appreciate it.
 
H

Hobbit

Senior Audioholic
Interestingly, the CM9 also appears to have a broad dip in treble response, beginning above 1.5 kHz, centered at 5-6 kHz, and extending all the way to 9 or 10 kHz. The CM2 (green trace) and CM5 (red trace) both share this general feature. It wouldn't be a stretch to say that this treble dip is an intended feature of the entire CM series.
Would you consider this trend, or feature, between speakers in the same product line to be typical throughout the industry?

Are there any rules or guidelines about speaker size vs. room size? I keep reading about bass extension in small to medium rooms and large speaker placement problems in small rooms.
 
Swerd

Swerd

Audioholic Warlord
Would you consider this trend, or feature, between speakers in the same product line to be typical throughout the industry?
It certainly is true for B&W. Many other speaker makes and product lines seem to have an identifiable sound, but I couldn't speculate how widespread this is. Even within a company as large as B&W, and their well-known sound features or flaws (depending on your preference), there seems to be big differences between the sound characteristics of their 600 and CM series and the much higher priced and much better sounding (and measuring) 800 series. To me, it's almost as if the 800 series are made by a different company.
Are there any rules or guidelines about speaker size vs. room size? I keep reading about bass extension in small to medium rooms and large speaker placement problems in small rooms.
None that I have enough confidence in to repeat to others :).

How bass sounds in a room isn't a simple matter of a speaker's size and it's bass extension as measured under optimum conditions. Just the same, many people do choose to use larger size and more amp power to attempt to solve problems. Brute force can work, but it costs money and it won't necessarily solve the problem.

You have to think about reflections off of walls, ceiling, and floor, and the locations of the listener and the speaker or sub woofer relative to those reflecting surfaces. How this behaves varies with the wavelength and the room dimensions. This is a complex topic and I won't attempt to try and explain here. Read this for starters, read this: http://en.wikipedia.org/wiki/Loudspeaker. Scroll down to the paragraphs on Listening Environment, Placement, and Directivity to get a feel of the complexity. Here is one short paragraph under Placement:

"In a typical rectangular listening room, the hard, parallel surfaces of the walls, floor and ceiling cause primary acoustic resonance nodes in each of the three dimensions: left-right, up-down and forward-backward. Furthermore, there are more complex resonance modes involving three, four, five and even all six boundary surfaces combining to create standing waves. Low frequencies excite these modes the most, since long wavelengths are not much affected by furniture compositions or placement. The mode spacing is critical, especially in small and medium size rooms like recording studios, home theaters and broadcast studios. The proximity of the loudspeakers to room boundaries affects how strongly the resonances are excited as well as affecting the relative strength at each frequency. The location of the listener is critical, too, as a position near a boundary can have a great effect on the perceived balance of frequencies. This is because standing wave patterns are most easily heard in these locations and at lower frequencies, below the Schroeder frequency – typically around 200–300 Hz, depending on room size."​

If you listen to some bass heavy music, and simply walk around the room you'll find locations where the bass seems louder, and other locations where it is lacking. A simple solution to puny bass is to move the speakers or the listening position. Of course, in many homes this isn't always possible.

This is the problem that room EQ software such as REW, Audyssey, or YPAO try to solve. Its a great idea, but they have had variable success. Different people love or hate the results. Stay tuned as the software improves.
 
C

cptslow

Audioholic Intern
How bass sounds in a room isn't a simple matter of a speaker's size and it's bass extension as measured under optimum conditions. Just the same, many people do choose to use larger size and more amp power to attempt to solve problems. Brute force can work, but it costs money and it won't necessarily solve the problem.
I am trying to understand what bass extension actually means.From my limited understanding/knowledge here is what i am thinking.
lets say for example ,I have a 6" driver and the driver has a -3db cutoff freq of 60 Hz, meaning the 6" cone will not move much for freqecies below 60 Hz. how can we have a extented bass response if the cone movement is not producing any noticiable SPL ?
 

Latest posts

newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top