Frequency vs dB

ajithlal

Member
Joined
Jul 15, 2008
Messages
61
Points
8
Location
Palakkad,Kerala
Frequency 28-24k @ -6dB
Freq. 35Hz-20kHz +/- 3 dB
Frequency Response (??3dB) 32Hz ?? 40kHz

These are from specs of diff l.speaker brands.

Could any one explain how will this affect in perfomance?

Regards
Ajith
 
Last edited:
First let us quickly understand what a how a speaker functions.

Speakers consist of one or more driver units in a box. The driver is constructed of a metal frame to which is attached a cone, made of paper or plastic and occasionally metal. At the rear end of the cone is attached a coil of wire (the "voice coil") wound around an extension of the cone, called a "former". The two ends of the voice coil are connected to the crossover network, and the crossover network is connected to the speaker binding posts on the rear of the speaker enclosure. The voice coil is suspended inside a permanent magnet so that it lies in a narrow gap between the magnet pole pieces and the front plate. The voice coil is kept centered by a "spider" that is attached to the frame and to the voice coil. A rear vent allows air to get into the back of the driver when the cone is moving, but a dust cap on the cone keeps air from getting in through the front. A rubber surround at the outer edge of the cone allows for flexible movement. In the case of a tweeter, the cone is very light, perhaps made of silk, and is glued directly to the voice coil. It isn't attached to a frame or rubber surround because it needs to be very low mass in order to respond quickly to high frequencies.

When the musical electrical signal from the amplifier passes through the voice coil, the voice coil turns into an electromagnet. Depending on which way the current is travelling in the voice coil, the north and south pole of the magnetic field will be at one end of the voice coil or the other. The permanent magnet has a north and south pole as well. Its magnetic field will push the coil (and the attached speaker cone) outward if the north and south poles of the two magnetic fields are lined up together (north to north, and south to south), or pull the voice coil inward if they are lined up oppositely (north to south, and south to north). So, as the electrical signal from the amplifier, which is a representation of the original musical waveform, passes through the voice coil, and changes direction with the musical waveform, the voice coil and attached speaker cone are driven outward and pulled inward in time with the music. The speaker cone pushes or pulls air in the room, which translates to increases or decreases in air pressure at your eardrums, and there you have it: music.

Speakers have created air pressure. Human ears do just the reverse. Human ears have diaphragm that are like microphone. When you put pressure on the diaphragm, it vibrates and creates electrical signals that are sent to neurons in our brains. These neurons convert the signals to what we perceive as sound.

The human ear hears sound as it is sensitive to pressure in the air. It does not have a flat spectral response. In other words, it's efficiency of detection or reaction to sound pressure), is a function of the frequency or wavelength of the sound signal. Sound pressure is often frequency weighed such that the measured level will match the perceived level. When weighed this way, the measurement is referred to as a sound level.

When you say perceive, you are referring to a comparison or a ratio. To do this you must have a reference level. The commonly used reference sound pressure in air is 20 micropascals (20 µPa) (rms), which is usually considered the threshold of human hearing (roughly the sound of a mosquito flying 3 m away). The issue is this is a very small unit (as it is 2 ten billionths of an atmosphere) and difficult to represent. Thus to measure sound pressure levels, a logarithmic ratio is used.

Thus sound pressure level (SPL) or sound level is a logarithmic measure of the rms sound pressure of a sound relative to a reference value. It is measured in decibel (dB).

The formula for calculating differences in sound pressure is

Lp=10Log10 (P~2 rms / P~2ref) = 20log10 (Prms/Pref) = dB

Where where p(ref) is the reference sound pressure and p(rms) is the rms sound pressure being measured. [Please see below for a representation of the formula is correct mathematical form.]

soundpressurelevelformu.png


Most measurements of audio equipment will be made relative to this level, meaning 1 pascal will equal 94 dB of sound pressure.

For instance, suppose we have two loudspeakers, the first playing a sound with power P1, and another playing a louder version of the same sound with power P2, but everything else (how far away, frequency, etc) kept the same.

The difference in power between the two is defined to be

10 log (P2/P1) dB where the log is to base 10.

If the second produces twice as much power than the first, the difference in dB is 10 log (2/1) = 10 log 2 = 3 dB.

If the second had 10 times the power of the first, the difference in dB would be 10 log (10/1)= 10 log 10 = 10 dB.

Remember that decibels measure a ratio. 0 dB occurs when you take the log of a ratio of 2. So 0 dB does not mean no sound, it means a sound level where the sound pressure is equal to that of the reference level. This is a small pressure, but not zero. It is also possible to have negative sound levels: - 20 dB would mean a sound with pressure 10 times smaller than the reference pressure, ie 2 µPa.

Let us see how a speaker specification is shown. If you look at the B&W 683, there are two rows of information.

Frequency response 38Hz - 22kHz ±3dB on reference axis (EFFICIENCY)
Sensitivity 90dB - (2.83V, 1m) (SENSITIVITY)

Both use the same unit of measurement - dB. Well let us not get confused. Decibel or dB is a unit of measurement that is used in both sound signal and electrical signals. Unfortunately since the unit of measurement is the same, a loudspeaker's sensitivity appears to be universally confused with its efficiency.

Sensitivity is strictly defined as how much acoustic power the loudspeaker puts out for how much electrical power it is being driven with. For example, if you feed a loudspeaker with 10 electrical watts, how many acoustic watts of sound does it produce? The answer is "not many," a typical moving-coil loudspeaker being about 1% 'efficient'.

Efficiency, on the other hand, is expressed as sound-pressure level produced by a speaker at a specific distance, 1m, for 1W input; ie, in decibels watt meter (dB/W/m) or simply dB. This is not that simple as it sounds as a loudspeaker's efficiency is dependent upon both impedance and frequency.

To simplify matters, the impedance of a loudspeaker is always assumed to be 8 ohms at 1000 Hz, at 2.83V, As per ohms law, (Power = V x V/R = (2.83 x 2.83)/8 = 1W.), you are now feeding 1 watt to the loudspeaker.

So there we have it. The base frequency for measuring the efficiency of a loudspeaker is 1000Hz or 1KHz.

The efficiency of loudspeakers are measured in what is called a anechoic chamber. This is a room where sound-absorbing materials on every surface soak up every sound emitted by the speaker. The room is therefore removed from the picture and the only sound that reaches the measuring microphone is therefore that from the speaker. A SPL meter is kept 1 meter from the speaker. The loudspeaker is first fed with a continuous sound signal at a frequency of 1000Hz. The sound pressure created by the loudspeaker is measured and this becomes the reference level sound pressure. Then various sound frequencies are fed to the speaker and their sound levels are measured. Sound pressure will increase or decrease at various frequencies as compared to 1000Hz. The differences are specified as ± dB.

A loudspeaker is accepted as efficient if the sound pressure level variation is equal to or below 4 for the stated frequency range. Thus a speaker specified as 20-20,000Hz ±2.5dB is more efficient than a speaker specified as 20-20,000Hz ± 3.5dB.
 
Last edited:
See the data below:
Frequency 28-24k @ -6dB
Frequency Response (??3dB) 32Hz ?? 40kHz.It doesnt have upper limit. Then how will we interpret it? Is there any influence for sensitivity on efficiency?
 
See the data below:
Frequency 28-24k @ -6dB
Frequency Response (??3dB) 32Hz ?? 40kHz.It doesnt have upper limit. Then how will we interpret it? Is there any influence for sensitivity on efficiency?

A speakers specification will be, amongst other things, given as follows.

Frequency response 38Hz - 22kHz ±3dB on reference axis
Frequency range -6dB at 30Hz and 50kHz
Sensitivity 90dB spl (2.83V, 1m)

I have already explained frequency response and sensitivity. The frequency response is 'overall;. In other words that represents the frequencies that the loudspeaker can handle.

Frequency range measures the frequency range within which the loudspeaker will not distort the sound beyond a certain limit. This is measured in amplitude and also represented by decibel or dB.

When you increase or decrease the volume, you do not want the loudspeaker to change the 'mix' of tones thereby destroying the timbre of voices and instruments. A music consists of number of voices and instruments each playing at a different frequency. If you record a music, with all these frequencies at the same volume, an ideal loudspeaker should reproduce the music without any change in the volume of individual frequencies across the full range of frequencies that the loudspeaker is capable of reproducing. This is called a loudspeaker having a flat response. Such loudspeaker do not exist. So sound engineers have settled for a amplitude figure that they think is tolerable and this is -9dB. In other words, any frequency reproduced more than 9dB down from the its original frequency will matter little to the sound. The speaker specs above say that this speaker loses 6 dB (or -6dB) on any one or more frequency at 30Hz and at 50Hz. What this means is that at 30Hz, some of the high frequency sounds may be reproduced with 6 dB less amplitude and vice versa. That is why this amplitude drop is shown as a negative figure. (-6dB)

As long as the loudspeakers frequency range covers the frequency response, and the amplitude drop is between 0 to -9dB, the loudspeaker will sound good in terms of imaging and delivering all the frequencies of the recording.

I have chosen these specific terms as different speaker manufacturers show these specification in different ways and the reader gets confused. The specs as shown above are used by reliable manufacturers such as B&W, Polk Audio, etc.
 
A speakers specification will be, amongst other things, given as follows.

In other words, any frequency reproduced more than 9dB down from the its original frequency will matter little to the sound.

Small correction. The anove statement should read

In other words, any frequency reproduced not more than 9dB down from the its original frequency will matter little to the sound.

Sorry for the slip up. English is fun and, at the same, difficult. If you miss a word, the meaning of the sentence is completely different Sigh !!!!!
 
But one practical advise..do not go purely by the specs.
-3 dB at 35 hz and at 40Hz for another speaker may not mean much, as more than 50% of what you hear is due to the placement of the speaker and due to room influence (which can add between 3 - 6 db easily to your response)

also 6 db at 28 means nothing in comparison as based on the design of the speaker, the 3 dB point may be 30 Hz or even 40 Hz !!


Funnily these specs are not even taken under standardized conditions as there is no standardisation .
some manufacturers do it in a Anechoic chamber while others do it in different rooms. some folks keep it on stands while others may hang it at a specific height from the ceiling etc etc....


in the end the room you are going to place the speaker in and the synergy it has with the amplifier driving it is what its sound is going to be defined by...not the specs.

so its better to leave the specs as an indicator of range of the speaker and nothing more .
 
But one practical advise..do not go purely by the specs.
-3 dB at 35 hz and at 40Hz for another speaker may not mean much, as more than 50% of what you hear is due to the placement of the speaker and due to room influence (which can add between 3 - 6 db easily to your response)

also 6 db at 28 means nothing in comparison as based on the design of the speaker, the 3 dB point may be 30 Hz or even 40 Hz !!


Funnily these specs are not even taken under standardized conditions as there is no standardisation .
some manufacturers do it in a Anechoic chamber while others do it in different rooms. some folks keep it on stands while others may hang it at a specific height from the ceiling etc etc....


in the end the room you are going to place the speaker in and the synergy it has with the amplifier driving it is what its sound is going to be defined by...not the specs.

so its better to leave the specs as an indicator of range of the speaker and nothing more .

In general what you say is correct. But nowadays, more and more speaker manufacturers are using sophisticated manufacturing techniques, making their own drivers, and designing high end corssovers. Most of the measurements published by respectable manufacturers are not done by themselves but by independent testing houses to which the loudspeakers are sent. Such testing houses with Anechoic chamber are present in US, in Canada, and in Europe. The Canadian testing site is owned by a research wing of the Canadian government and cannot be pressurised to give false information.

Thus the specifications published by manufacturers such as B&W, Wharfedale, Polk, etc., will not be false. Usually the speakers manufactured by such companies will deliver good results if you understand what the specs mean.

This thread was to understand what these specifications mean and should help in shortlisting speakers that you want to audition.

Ultimately, as I have been saying repeatedly, only a personal audition is the final decision maker. But that does not mean you can ignore specifications.
 
Most of the measurements published by respectable manufacturers are not done by themselves but by independent testing houses to which the loudspeakers are sent. Such testing houses with Anechoic chamber are present in US, in Canada, and in Europe. The Canadian testing site is owned by a research wing of the Canadian government and cannot be pressurised to give false information.

Thus the specifications published by manufacturers such as B&W, Wharfedale, Polk, etc., will not be false. Usually the speakers manufactured by such companies will deliver good results if you understand what the specs mean.


You are correct, paradigm, Energy and PSB (I think Mirage also does it) are 3 manufacturers who do get speakers tested there in an anechoic chamber (ie National Research Council or NRC) . they i guess publish their specs to 2 db wlimits. although testing results of an Anechoic chamber in what it means to the final consumer can be extensively debated ;) as it is of use (Actually great use) during its design. and these are truly very well measuring speakers


On the other hand Proac, does not even give db limits for their specs..aparently it is a 10db limits !. but again they are fantastic speakers. the reason given is that room parameters are so varying that specs mean nothing much (Which is what incidentally even Bose says !)

Finally polk does publish specs with 3dB limits and Mathew Polk is a true scientist in audio there is no disputing his results

but all three cases above have specs measured unders different considitions and give theior specs differently !

to reiterate i am not disputing the correctness of the Specs..only that the content of what they give does not as mean much

In case manufacturers do give us a response curve and a Impedence/phase curve and possibly even the dispersion data, we would be able to read far more into the spec (Although personal audition is still the most important)
eg the stereophile measurements are really nice
Stereophile: B&W Nautilus 805 loudspeaker

soundstage network also does do a good measurement, although the methodology they follow is again different:eek::D
 
any frequency response above 20Khz or so actually goes waste, and i'd rather think it is useless to buy speakers with frequency response above that because human ear can only listen upto 20Khz, CDs have recordings upto 20.5KHZ only. the lower end of frequency reponse is more important, generally a frequency response on lower end of 40Hz will do the job for most of the audio listeners (2ch & 5.1/7.1), but in order to get those deep (i call it Ear Filling, Clothes Shivering) effects you'll need speakers that go even lower to say 30 Hz or so (in case you buy a subwoofer, it'll go even below that to say 24hz).

Sensitvity is a more simple issue, higher the sentivity of speaker less is the power going to be required by it to achive certain levels of loudness as compared to a speaker with compartively low sensitivity.
generall speakers with sensitivity above 90db are considered good, pro grade speaker have sensitivty above 100db.
Klipsch are extremely sensitive and powerfull speakers in my opinion.

ps: frequncy and db's are not interrelated
 
generall speakers with sensitivity above 90db are considered good, pro grade speaker have sensitivty above 100db.

Are you sure about the above ? is there any link on the above?

I checked up on some of the known pro models like Dynaudio, Harbeth etc and did not find a corroboration. IN fact there are very few speaker models above 100dB and fro what i know they are not in the Pro range
 
What he probably means is transducers used for open air concerts or those used as guitar amplifiers and so on. Most of these are horn loaded and have 100+ dB of sensitivity. Not the most accurate sounding speakers in the book though.
 
You might be right. seems to be more of a switch between Pro Audio and PA audio ..
 
any frequency response above 20Khz or so actually goes waste, and i'd rather think it is useless to buy speakers with frequency response above that because human ear can only listen upto 20Khz, CDs have recordings upto 20.5KHZ only. the lower end of frequency reponse is more important, generally a frequency response on lower end of 40Hz will do the job for most of the audio listeners (2ch & 5.1/7.1), but in order to get those deep (i call it Ear Filling, Clothes Shivering) effects you'll need speakers that go even lower to say 30 Hz or so (in case you buy a subwoofer, it'll go even below that to say 24hz).

A standard music CD has audio signals recorded as a sequence of 16-bit numbers with 44.1 kHz sampling frequency. Theoretically, it allows recording and consequent playing back of the audio signals with frequency range from 0 up to 22.05 kHz. In practice, most DACs cut off at 18-19 kHz without significant distortions. But moderns ADC and DAC now operate at 24 bits and upto sampling frequencies of upto 96Khz, though many go upto 192Khz. This means that you can store and play data upto rougly 48KHz or even 96KHz.

Current CD recording are capable of having data that are recorded as low as 10Hz and upto 48KHz. Many CDPs and amplifers are also commonly mentioning specs from 10Hz to 50 or 60 KHz. SACD recordings can span between 10Hz and thoretically 100KHz. But SCAD players currently can play upto 80 or 90KHz.

Even the best made floor standers usually do not go below 20Hz. This is because such a wide span of frequency response will affect mid and high frequency. That is the reason that pre-amplifiers and receivers have external (outside the loudspeaker) filters that remove frequencies below 40 to 100Hz and route them to specialised loudspeakers that can handle just such low frequencies. These are, of course as you know, called sub woofers.

Modern sub woofers can go low as 10Hz, at least in specification.

But remember, if you play too much importance to low frequencies, you are adding to too much compression, and a gradual distrotion of music. THX set the low frequency cutoff ar 80Hz. Why did they do this? THX looked at a large number of modern 5.1 soundtracks and guess what they found in the LFE channel? Not much data even in the region of 80 Hz - 120 Hz, leave alone less than that. Dolby Digital's LFE channel stops at 120 Hz (does not roll-off), so content creators almost always roll-off their stuff, usually somewhere around 80 Hz. So what are you amplifying at anything less than 80Hz? Literally nothing.

If you go to a club or a discotheque, the speakers are large, and the wattage are in 500s or more, This is to make the music sound powerful and superficially more dynamic, lending a sense of drive and impact. Such mixes are notorious in applying high amounts of compression. Low frequencies also travel a lot by getting reflected by the any object that is there in its path. This adds to the compression.

How long can you hear such music? Is there any clarity at all? Is there any imaging?

I once was standing on the road and listening to set of artists playing drums of various kinds. No mike, no amplifier, nothing. Pure natural sound. And the astonishing thing was I have never heard that sound on any amplifier system. A tabla that you hear directly without any amplification, will sound completely different when you play it through your system. There will be an addition of low frequency making the tabla sound's seem bigger, and in some cases, maybe even a reverberation.

In my mind this is the best test of a system. Do the voice and instruments sound natural? Get a friend of yours to play a tabla, violin, a veena or a sitar with out any amplification. Register that sound in your mind. When you hear your system next, close you eyes and compare it to a tabla or a violin you have heard without any amplification. If it sounds the same or even similar, you are very fortunate to have an excellent system.

Sensitvity is a more simple issue, higher the sentivity of speaker less is the power going to be required by it to achive certain levels of loudness as compared to a speaker with compartively low sensitivity.
generall speakers with sensitivity above 90db are considered good, pro grade speaker have sensitivty above 100db. Klipsch are extremely sensitive and powerfull speakers in my opinion.

Though you are right that sensitivity and power are inversely related, it is not that easy to make drivers of high sensitivity. For a driver to be very sensitive, the cone needs speacilsed material such as silk and thin paper. The issue with these material is that their life time is short. Played for some time at high volumes, they start losing their rigidity and start vibrating and creating unnecessary sounds.

Manufactures such as B&W use speacialised material such as Kevlar. Irrespective of how much ever you fine tune the material, it has a certain rigidity below which it cannot go. So the manufacturer has to play a fine balance between life of the cone and its sensititivity. Generally, we do not change speakers very often. We expect them to work for at least 10 years if not more.

Sensitivity, as I said before, is measured in terms of dB/watt/meter or dB/2.83 volts/meter which means that a certain loudness (1 dB or decibel) will be achieved with a standard signal level (2.83 volts or 1 watt @ 8 ohms) at a standard distance from the speaker (1 meter). Conventional "dynamic" drivers can offer reasonable efficiency and good dynamic performance, usually in the range of 85-90 dB/2.83 volts /meter. Anything below or above that will introduce distortion.

Highly sensitive speakers such as horns can reach levels above 100 dB/watt/meter, and this type of speaker were used in the early days as amplifiers only had about a 10 watt output.

If you have a high powered amplifier (100 watts per channel or more into 8 ohms), sensitivity is not a major concern. It become important only when your amplifier is pumping out something between 20-40 watts. Technically, Speakers with ratings above 90dB/2.83 volts/meter are considered highly sensitive, ratings between 86 - 90 dB/2.83 volts/meter are not as sensitive, and speakers with ratings of below 85 dB/2.83 volts/meter are considered insensitive.

They are a rare and expensive breed now, and used with specialised amplifiers that pump out less than 20 watts. Amplification is so easy nowadays that speaker manufacturer maintain the sensitivity around 90dB and focus more on other issues such as crossover, frequency range and respose. The main attempt will be to deliver flat frequency response. This becomes difficult if the sensitivity is too high.

Sensitivity ratings, looked at alone, can be somewhat misleading because the measurement will depend on the impedance variations of the speaker across its audio spectrum. Speaker sensitivity ratings should be viewed in combination with other specification of the speaker, not in absolute terms.

I also read your link on Speaker sensitivity. It just re-emphasises what I have written. As I have said above, 'LOUDNESS" has no value. There is only a certain amplitude that human ears can sustain. Anything above that is just noise. And, in any case, specialised amplifiers that have low amplification are rare. Majority of the speaker manufacturers we are interested in are in the more humble and general market.

ps: frequncy and db's are not interrelated

[/QUOTE]

I do not understand what you are saying here. dB is a just a unit of measure and is used for measuring a number of items including sound and electrical signals. Frequency by itself usually is not enough to give any information. That is the reason loudspeakers sepcify frequency response with pressure and frequency range with amplitude. Both pressure and amplitude are respresnted by dB as their unit of measure.
 
A beautiful, well-constructed speaker with class-leading soundstage, imaging and bass that is fast, deep, and precise.
Back
Top