Do deciBells ring a bell ?
published: Nov-13-2014, last edit: Oct-17-2022
1 Bell (named after Alexander Graham Bell) = 10 deciBell (dB)
Forget the Bell, no one uses it. We only use 1/10th of Bell, i.e. deciBell (dB).
In audio a lot of numbers are given in dB (decibel).
The reason for that is rather simple.
It is sometimes easier to calculate with dB numbers as dB numbers are logarithmic and only need to be added or subtracted in most cases.
Our hearing (and eyesight) is also closer to logarithmic behavior so it makes sense to use dB’s in audio and video as it comes closer to how we perceive light and sound intensities.
Double the SPL or amount of light (in a linear scale) is not nearly perceived as a doubling though, just as a substantial increase so the hearing isn’t exactly logarithmic.
A dB value shows a RATIO between 2 values expressed in a logarithmic scale.
This is also the confusing part as a ratio always relates to 2 different values, one must be known… the reference.
In a LOT of cases that reference isn’t very clear or explicitly mentioned and thus some knowledge about those references is needed to understand dB’s.
The biggest drawback of using dB’s (a ratio) is that 0 Volt, 0 Watt etc. can NOT be used in calculations as dividing a number by 0 (zero) will give the answer ‘infinite’.
dB’s can be negative and positive depending on which of the 2 values one considers the reference.
-100 dB simply means the value that is specified is ‘100dB’ smaller than the reference level.
This reference level may be a known voltage or may be a specified output voltage or even a maximum output voltage.
+30 dB means the value of the measured signal is 30dB higher than a reference signal. An amplifier with 30dB gain will thus amplify the input signal 30dB (times)
Voltages, currents, resistances, power levels, SPL etc are noted in linear scale values.
For example 5V is half of 10V, 1A = 4x 0.25A and 10W is 10x as much as 1W.
In the upper example it appears to be very easy to calculate with linear scales so why would one bother using dB’s instead ?
Well the reason for that is the big dynamic range involved with audio.
Voltages could be present ranging from a few nanoVolt to several kiloVolt .
for instance a low output MC cartridge reproducing a signal of -60dB has (0.000.000.2V = 200nV)
An electrostatic loudspeaker membrane can have 1,000V (1kV) on it.
These values are a factor 5,000,000,000 apart.
Very easy to loose one or more 0’s or even misplace an entire x3 multiplication factor.
In dB’s this is ‘only’ 194dB apart so a number that is much easier to work with.
On the left a table with ‘gain’ values on a linear scale on the left colom and the (approximated to create nicely rounded numbers) corresponding dB scale in the right colom. The ‘gain’ numbers are for Voltage, current and SPL NOT for Power values.
A small difference, for example between 8,35V and 8.358V (= 0.01%), can also easily be described and is 0.008dB so small and big differences can easily be ‘captured’ in a dB scale.
Another reason for using dB’s is for calculating total gains.
For example a turntable cartridge is connected to a step up transformer which converts impedances and voltage. It can be said to have a voltage gain. The RIAA stage that follows also has a gain.
That one differs per frequency so we simply use 1kHz as a reference (1kHz is MOSTLY used as a reference frequency for multiple reasons).
If one wants to calculate the total gain we will have to multiply amplification factors when using linear scale values.
As an example a MC step-up trafo of 1:20 is used, followed by the RIAA pre-amp having a gain of 38dB.
The trafo has a gain of 20x. The RIAA amp, if the gain were given in a linear scale, would be around 79x.
The total gain of the system is thus about 1580x.
A low output MC cartridge with 0.5mV output signal will thus be amplified to 0.79V
Same calculations, but using dB’s:
The trafo has a gain of 26dB, the RIAA amp 38dB. The total gain is 26 + 38 = 64dB
So instead of using multiplication factors (in linear scale calculations) in the dB scale we only have to add or subtract. Much easier to do but if we want to know the output voltage of the system we need to convert dB to a linear scale.
BUT… simply adding dB’s isn’t possible in all cases.
For example suppose we have 2 speakers in our room both emitting 90dB SPL.
Simple logic would say 90 dB + 90 dB = 180dB SPL…. right ?
This, however, is not the case as in reality the acoustical ‘energy‘ coming from 2 speakers is merely doubled.
Assuming the speakers have a 90dB/W efficiency and both channels apply 1W, the total output power for stereo is thus 2W. A factor 2 in POWER is slightly over 3dB. The total SPL thus is 93 dB SPL and NOT 180 dB SPL.
This is one of the more difficult and counterintuitive aspects of working with dB’s.
Another example of a difficulty with juggling around dB’s SPL lies in the determination of the actual contribution of a sound ‘source’ we want to measure compared to the total SPL when the surrounding noise is close in level to that of the device being measured.
This involves measuring the background noise (so with the to be measured source turned off), convert that dB reading into a linear scale and then measure the total noise consisting of background + the to be measured noise. That combined reading needs to be converted into a linear scale value as well.
These linear scale values can then be used to calculate what ‘energy’ is added and that outcome has to be converted to dB’s again to know just how much the unknown sound source adds in dB SPL.
Of course there is an easy way and there is a simple Sengpiel Audio online calculator to assist.
I don’t like equasions much but in this case, because to convert to dB, there are needed.
To calculate a dB value we need to know the ‘reference value’ and the ‘difference value’.
We also need to know if we are talking about power (Watts, milliWatts, microWatts) or something like a Voltage, current, resistance or SPL.
One can calculate it yourself using a scientific calculator or simply use one of Sengpiels online calculators.
To calculate the dB difference from 2 different power levels:
xxdB = 10 x LOG10 (VDIF/VREF)
To calculate the dB difference from 2 different voltage/current/SPL etc) levels:
xxdB = 20 x LOG10 (VDIF/VREF)
Reference points:
dB’s also exist with a few ‘fixed’ reference points.
In that case the dB value describes a difference in level compared to that known and fixed reference value.
There are many of those and some of them are relevant to audio such as:
Dbu, dBV, dBm, dBmV, dbμV, dB SPL, dB(A), dB(C), dBFS.
At one time there was the dBv which existed alongside the dBV.
This lead to a lot of confusion so the dBv was replaced by dBu.
Both dBu and dBV have a fixed but different voltage level as a reference.
0dBu = 0.775 Volt regardless of impedance.
The 0.775V is not an accidental choice but is the voltage needed to reach 1mW into 600Ω.
0dBV = 1V regardless of impedance.
The ‘0dB’ mark on professional audio equipment usually indicates a +4dBu output voltage (= 1.23V = + 1.8 dBV)
The ‘0dB’ mark on older home audio equipment usually indicates a -10dBV output voltage (= 0,316V)
‘0 dB’ line level for most DAC’s varies between 1V (0 dBV) and 2V (+6 dBV) depending on what the manufacturers ideas of ‘0 dB’ is.
dBm = is used in fiberoptics, in the telecom world but also for headphones.
For circumstances besides headphones 0dBm = 1mW regardless of impedance. Sometimes dBm is written as dBmW.
For headphones xxdBm, sometimes written as xxdBmW or xxdB/m indicates the SPL level that is reached when 1milliWatt of power is applied.
Usually a 400Hz or 1kHz sinewave.
The voltage needed to reach 1mW is thus dependent on the impedance of the load.
dBu and dBV differ in that these are voltage referenced regardless of the applied load impedance.
To make things worse the efficiency of headphones sometimes is given in dBV.
In that case xxdBV represents the SPL when 1V is applied regardless of the impedance.
1V is NOT the same as 1mW and thus can lead to substantially different dB numbers.
This all becomes even more confusing when manufacturers merely state xxdB efficiency and neglect to mention if it is dBm(W) or dBV.
I have made a headphone power cross reference table that contains the voltages, currents, efficiency numbers (both dBm + dBV) and dB SPL values for most ‘well known’ on- and over-ear headphones.
When dBm is used with headphones the load impedance is determined by the headphone
When dBm is used in telephony the used load impedance is 600Ω
When dBm is used in radio frequencies the used load impedance is 50Ω
Another strange form of referenced dB that is sometimes used is dBFS (dB Full Scale). The Full Scale output voltage is the maximum output voltage just before clipping occurs.
The output voltage of a DAC is generally considered as 0dBFS and thus does not reference to a fixed voltage level but rather the maximum output voltage.
For amplifiers one sees distortion numbers or frequency ranges given at -3dBFS.
This produces low distortion figures in general because near the clipping border the distortion rises fast (exponentially).
Measuring values like S/N ratio and distortion levels at 3 dB below the maximum levels thus produces flattering numbers.
In case of DAC‘s this isn’t the case though as the analog output stages usually have some ‘headroom’ left before the DAC chip reaches its maximum output.
No for all DAC’s though, the Audioengine D3 DAC for instance clips, most likely because of some wrong choices here and there.
In the telecom business dBmV (0 dB = 1mV in 75Ω) and dBμV (0dB = 1 microVolt) are also common because of the extremely low levels present. These are not used for the audio part but can be found in tuners/receivers specifications.
In the world of acoustics dB’s are used a lot.
There is a LOT of science in acoustics and most of it is difficult to fully understand.
Sengpiel Audio is specialised in this kind of stuff and for those that want to dive in deep I can recommend to have a look at their website with online calculators and explanations
dB SPL (Sound Pressure Level) is often used for loudspeakers and headphones.
Mostly the ‘SPL’ part is left out completely and just dB is used.
The reference here is a sound pressure (in air) of 20 microPascals (μPa) = 2×10−5 Pa, this is approximately the quietest sound an ‘average’ human can hear at 1kHz.
There may be some differences between individuals but these are very small.
The well known decibel tables that show in what range certain sounds are as a reference is well known and found everywhere on the Internet.
One should realize that these are average levels and is not the same as peak levels in music reproduction.
Sound pressure level, in some circumstances, is given in dB(A), dB(B) or dB(C).
When one of these suffixes (A, B or C) is used it means the measurement microphone signal has been (high-pass) filtered and is made less sensitive for lower frequencies.
When a 1kHz tone of a certain SPL is measured it should not matter if the switch is in A, B or C setting and should give about the same readings.
Things differ when wideband noise, music or lower sinewave frequencies are measured. The reason for this has to do with the human hearing as it becomes less sensitive for sound pressure as the frequency gets lower.
Below 300Hz this effect is becoming noticeable. Also above 5kHz the human hearing is slightly less sensitive.
This can be seen in the Phon curves (picture from Sengpiel audio website).
To ‘experience’ a 50Hz tone as ‘equally loud‘ compared to a 30dB SPL 1kHz tone that 50Hz tone needs to be around 60dB SPL.
The dB(A) setting should be used to measure background noise values in quiet surroundings below 40dB.
At higher SPL’s the meter will give the wrong numbers in the dB(A) setting as the human hearing becomes relatively less insensitive for lower frequencies when the SPL goes up.
So when measuring background ‘noise’ in a working environment for instance, to check if regulations are met, the ‘A’ filter is applied (a switch on the SPL meter) lower frequencies aren’t ‘weighing’ as much on the entire signal.
A 50Hz tone of 80dB SPL will thus show as ‘50dB SPL‘ on the meter readout in ‘A‘ mode. It will show about 70dB SPL in the ‘B‘ setting, and 80dB in the ‘C‘ setting.
A 1kHz tone of 80dB SPL will always show as ‘80dB SPL’ on the meter readout in all 3 modes.
Most SPL meters also have a dB(C) setting which is ‘almost’ flat between 70Hz and 4kHz and has it’s -3dB points at 30Hz and 8kHz. This setting shows the actual SPL values within that frequency band.
Below 30Hz and above 8kHz all bets are off accuracy wise unless the meter is specified outside the ‘standard’ range.
For measuring the output of speakers/headphones these meters generally don’t have enough bandwidth and incorrectly display the lowest frequencies and the ones above 10kHz. Should you still want to use a dB meter for this the dB(C) setting is the only appropriate one.
In audio reproduction in ‘home conditions‘ our limits in practice will be between 25dB SPL (silence) and 115dB SPL (VERY loud) so roughly 90dB in dynamic range.
Sure… we can even hear below 0dB SPL (at certain frequencies) and handle an SPL well above 120dB (at low frequencies) without experiencing ‘pain’.
BUT we usually don’t have acoustically silent rooms to be able to detect sounds below 20dB SPL.
Most people will rarely listen to music > 100dB SPL but should realize that in order to hear rumbling bass at those frequencies should be able to reach 120dB SPL.
At those higher SPL’s the ‘automatic gain control’ in our ears already lowers the sensitivity of our ears.
Yes, we can hear signals of 0dB SPL (0 Phon) which still has a measurable level of 2×10-5 Pascal.
0dB SPL thus doesn’t mean NO SOUND. There IS still a measurable sound pressure but the average human simply doesn’t hear it any more. As the hearing canal resonantes between 2-4 kHz it is more sensitive than at other frequencies and with specific tests (within the 2-4kHz region) in extreme silent surroundings, some can even hear to down -7dB SPL!
BUT… your typical room in the house isn’t nearly as quiet as those sound proof chambers.
We can also hear well over 120dB such as levels of 150dB SPL (a jet engine at 30 meters) but you really would be hard pressed to NOT cover your ears and even short exposure will ruin your hearing.
The hearing has an ‘automatic gain control’ a muscle called the Tensor Tympani that can make the hearing less sensitive, somewhat similar to the iris in our eyes.
Every one who went to a party/disco/live rockband and spent quite some time there, possibly even near the sound reinforcement/speakers may have experienced how ‘quiet’ the world around them had become when stepping outside into the street.
Cars sound much softer and people around you, that are talking in a normal voice (those that hadn’t been exposed to the loud sounds), will appear to be talking quite softly.
The brain has reduced the sensitivity of the hearing in order to protect it by tensioning the eardrum (making it less flexible) and it takes some time for that muscle to relax again.
Even during the day our hearing is somewhat lowered in sensitivity, but not nearly as much.
Most people have experienced hearing soft sounds in the late hours they usually do not hear during the day.
Think clocks or low level hum or other faint sounds that only become apparent in the small hours.
When we go to sleep the brain also starts to ignore signals coming from the ear.
This differs substantially from person to person. Some wake up from soft sounds, with others you will have to set of a canon to wake them up.
This is a DIFFERENT mechanism though. Sometimes when we wake up you can hear the sounds around you ‘some up in volume’ witin a very short time frame (seconds)
Where does this leave our true dynamic range. Well that is substantially smaller than most people think.
I was curious about dB levels and what things I could and could not hear. To find my personal limits instead of relying on other peoples findings.
This lead to the creation of a switchable attenuator where I can listen to the unattenuated signal and switch over instantly to an attenuated version of that signal.
It consists of 3 switches and calculated resistor ladders (including the input resistance of the amp behind it) as a passive attenuator. 
Using some ‘standard’ steps I was particularly interested in I experimented a lot while listening to music signals (not sine waves or ‘noise’ signals)
I found the following ‘steps’ to be of importance:
0.1 dB is measurable (around 1%) yet considered by most and also as found by me inaudible.
0.2 dB (around 2%) is detectable when switched directly but not as a level difference, rather as ‘fuller sounding’ (when switching from -0.2dB to 0dB).
0.5 dB (around 6%) difference is audible for most, but to some not that audible as a level difference but as ‘fuller’ (+0.5dB) or ‘thinner’ sounding (-0.5dB). Trained ears might detect a small jump in ‘level’.
3 dB (141% or 70.8%) difference is quite audible and is often used in specifications as it is a doubling (or halving) of POWER.
6 dB (200% or 50%) difference is VERY audible and sometimes given as it is a doubling (or halving) of VOLTAGE, CURRENT or SPL.
Around 10 dB is generally considered to be a perceived as a doubling of loudness.
Around -10dB is generally considered to be a perceived as a halving of loudness.
Every 10dB attenuation gave about a similar ‘effect’ of music sounding ‘about’ half as loud compared to the previous setting.
When you play music pretty loud (think 100-110dB peaks) and, instead of completely ‘muting’ the music signal it is lowered by a fixed amount, I found that lowering the output signal of speakers by -70dB left me robbed of any sound. It was equally ‘quiet’ to me as -100, -120 was.
So a reduction of music with peaks of say 110dB (average level around 97 dB = very loud) by 70dB left me (for a short period) with peaks 40dB of SPL which I could NOT hear at all (total silence). Of course the average SPL would have been around 30dB.
After a while I could start to hear some soft sounds again as the ears adjust.
So while this is not really scientifically proven here, I think the hearing has a usable dynamic range for audiophile listening of around 70dB for pretty critical audiophile listening.
It did make me think of the relevance of noise figures and THD components being at -80 dB or even down to -120dB or even further.
I am pretty sure I can’t hear any of those artefacts while listening to music.
loudspeakers, Watts and decibells
Some claim a few Watt is enough others swear by hundreds of Watts….. Who’s right ?
An easier to grasp analogy for understanding what’s involved.
Think of sound as light. Both travel in a straight line and have a certain amount of power applied to it. Both our hearing and eye sensitivity is sort of logarithmic.
Our distance to the speakers is usually determined by the size of the room, just like the distance to the walls will determine how much light falls on each square cm.
Of course in the example below we are speaking of linear scale values which will have to be converted to dB’s.
From 1m to 2 meters is factor 4 reduction in power (4 squares /1 square) and thus -6 dB. A doubling of distance yet again (2m to 4 meter) is again a factor 4 (16 squares /4) thus -6dB… again.
If we want to light up the walls of a large room, using a single light bulb, equally bright as the walls of a much smaller room we need vastly different power ratings of the light bulb OR light sources which have a much higher efficiency.
A 60W ‘normal’ light bulb will be perfectly able to ‘sufficiently’ light a small room.
For a large room to be lit equally bright you may need 500W OR you could get the same amount of light with 300W Halogen, 150W energy-saving incandescent bulbs or even a 30W LED light source.
Thinking the other way around the small room can be lit with a 3W LED source and the large room can be lit equally well with a 500W glowbulb.
The same is true for audio.
A lot of research has been done on the propagation of sound and this online calculator can show you how much dB (as it is a ratio) is lost with a certain distance.
Of course a room has reflecting walls and speakers are ‘spreading’ sound differently depending on the frequency and thus will be different from sound travelling in free air.
In free air the SPL will decrease 6dB per DOUBLING of distance.
To compensate for those 6dB you will need to quadruple the amount of power if you want the same SPL at double the distance.
Problem.… in general we do not listen at 1meter but perhaps anywhere between 3 meter (smaller room) to 6-10 meters in larger rooms.
To make it more tangeable … assume 2 speakers of 87dB/w/m effciency (so 90dB SPL).
On 1 meter distance we have 90dB. From the first meter to the second meter of distance we will ‘loose’ about 6dB so at 2meter we just have 84 dB left.
At 3meter 80.5dB, at 4 meter 78dB, at 6 meter 74.5dB.
Of course in a room this isn’t exactly the same as we also have reflections of sound energy.
Back to the light analogy again… When we have a ‘normal’ room with wall paper and furniture, curtains etc and we look at the poststamp at 4 meter a certain amount of light will fall on it. When a big part of the room is covered with mirrors a lot of the light energy that would otherwise have been absorbed/scattered by the things inside the room is now reflected and SOME of those reflected rays may also hit the poststamp increasing the amount of light falling on the poststamp.
For the amount of light it can easily add but for audio, however, it isn’t as simple as reflections differ per frequency and higher frequencies may well add to a measured SPL when using a meter but our brain can tell direct sound apart from reflections. Still we will perceive the same amount of SPL in a very echoing/reverberant room (think tiles on the floor, large walls/windows and no curtains/furniture) louder than the same SPL from the speaker in the same room but carpetted and with lots of furniture and curtians.
So in the end how loud we experience a speaker will also be determined by the room itself and our distance to the speakers so the attenuation of SPL of 6dB/doubling of distance will in practice be somewhat smaller depending on the room geometry and what’s in it.
Of course there are more factors that are important.
An obvious one that we won’t include is time of day (or should I say night) which also adds to the discussion but as our objective measurements do not have the ‘automatic gain control’ the ears have we will leave it out of the equasion.
So factors are the amount of SPL, room size, listening distance and room conditions.
The amount of SPL is determined by 2 factors. Efficiency of the speakers and applied power (assuming the speakers can handle the applied power levels)
The efficiency of speakers can vary between 80dB(@1W@1m) and 115dB(@1W@1m).
The ones with the super high efficiency are, without exception, horn loaded speakers.
(picture taken from ‘stereophile’ website)
Most ‘home speakers’ will vary between 85dB and 95dB (at 1W and 1m distance) though.
Home amplifiers exist ranging from a few Watt (generally these are all tube amps) to several hundreds of Watts.
Another aspect is how LOUD do you want the music to be. Obviously with neighbours and old houses you can’t play very loud, certainly not in the small hours without running into problems of a personal nature with people who do NOT want to listen to your music.
Those fortunate enough to live in big mansions without direct neighbours obviously may be able to play their music a LOT louder.
So how much peak SPL do we want at the listening position ?
That may vary between 70dB (background music that can be heard clearly) , 80dB (comfortable and decent levels) to 90dB when we want to hear things at a realsitic (is pretty loud !) levels.
To emulate live rock concerts or classical concerts 95 to 110dB SPL will be needed.
These are AVERAGE levels so you can take the readings from the DR tables to see what peaks can be present.
The fun of the DR scale is that it measures the difference between peaks and average SPL in dB’s. dB meters in ‘slow’ and ‘C weighting’ settings will show average values. A CD with a DR8 will thus have peaks that are 8dB above it. DR20 albums will have +20dB peaks etc.
So for your average livingroom with direct neighbours (that are at home) about 80dB (average) will be possible. Perhaps 90dB when they aren’t home.
For the bigger mansions/rooms and or no direct neigbours 90 to 100dB will be very rewarding.
Some calculations:
A speaker with a high efficiency (Say a Klipsch hornspeaker with 105dB efficiency) can fill a smal room with a high SPL with just a few Watts.
A 15W will produce 11.8dB more power than 1W so 1 speaker will be able to produce 116.8 dB… we have 2 of those (stereo) so 119.8dB at 1m.
That will be PEAK SPL so NOT average.
The same speaker but connected to a 2x 250W amplifier… 250W will produce 24dB higher SPL than 1W so 129dB SPL … we have stereo so 132dB @ 1 meter.
Of course we do not listen at 1 meter distance but say 3 meter (in a smaller living room) or say 8 meter in a much larger room.
With 2x 15W in the smaller room (so at 3 meter) the SPL that can be reached is around 112dB (peak) so average levels (DR10 recordings) of 102dB can be reached.
This is VERY loud.
With 2x250W in the smaller room (so at 3 meter) the SPL that can be reached is around 122dB (peak) so average levels (DR10 recordings) of 112dB can be reached.
I can assure you that you will be covering your ears IMMEDIATLY.
Now for the bigger room...
With 2x 15W in the big room (so at 8 meter) the SPL that can be reached is around 103dB (peak) so average levels (DR10 recordings) of 93dB can be reached.
This is pretty loud but still comfortable (remember the amp is on it’s toes pushing out maximum power peaks).
With 2x250W in the big room (so at 8 meter) the SPL that can be reached is around 115dB (peak) so average levels (DR10 recordings) of 105dB can be reached.
This is rock concert level and quite impressive.
So … advocates of ‘who needs more than 15W’ have a big point there. Even a big house can be filled with music … using these expensive and bulky horn speakers.
More power (you need 10x more power to double the perceived loudness) will get you more SPL and perhaps more enjoyment but also higher distortion in general.
Let’s get real and use something like the small B&W PM1:
A speaker with a low efficiency (Say B&W PM1 with 84dB efficiency) will need a LOT of power to fill the same room with an equally high SPL.
Same calculations, same amsp, same room conditions.
A 15W will produce 11.8dB more power than 1W so 1 speaker will be able to produce 95.8 dB… we have 2 of those (stereo) so 98.8dB at 1m.
That will be PEAK SPL so NOT average.
The same speaker but connected to a 2x 250W amplifier… 250W will produce 24dB higher SPL than 1W so 108dB SPL … we have stereo so 111dB at 1 meter.
I know the PM1 is rated for 100W (unclipped) so we would have to be carefull with the volume control.
With 2x 15W in the smaller room (so at 3 meter) the SPL that can be reached is around 89.5dB (peak) so average levels (DR10 recordings) of 79dB can be reached.
Good enough for listing to music in circumstances with direct neighbours.
With 2x250W in the smaller room (so at 3 meter) the SPL that can be reached is around 99dB (peak) so average levels (DR10 recordings) of 89dB can be reached.
This is a pretty loud listening level for ‘quite realisitic’ reproduction.
2x15W thus will only be enough for smaller housed people with direct neighbours which don’t need to listen at realistic levels.
Given the fact that 90dB SPL is the highest they can reach and assuming 20dB is the audible threshold in such a room a 70dB dynamic range is enough.
Now for the bigger room…
Using 2x 15W in the big room (so at 8 meter) the SPL that can be reached is around 82dB (peak) so average levels (DR10 recordings) of 72dB can be reached.
Just good enough for playing background music, as soon as you want it somewhat louder it will start to sound nasty, bass shy and compressed.
Using 2x 250W in the big room (so at 8 meter) the SPL that can be reached is around 93dB (peak) so average levels (DR10 recordings) of 83dB can be reached.
While sounding twice as loud as the background music it will still be impossible to reach ‘serious listening levels’ for audiophiles and given the speaker is rated for 100W and the amp can provide 250W it will be quite easy to blow up the speaker while trying to impress a visitor.
In the end, 2x 15W and even higher power amplifiers with inefficient speakers won’t lead to satisfactory results unless one only likes to listen at lower (background level) music.
Of course when most of your favorite music is well below DR10 (most modern rock and pop music is) average levels will be somewhat more impressive.
The short answer to whether or not your 2x 15W amplifier is enough thus depends on the efficiency of the speakers, the distance to the speakers, the maximum SPL that is ‘enough‘ for you (how loud you want to be able to play) and the DR rating of the recording.
For headphones the same is true but the efficiency ratings can be even further apart. IEM’s and some ear buds can have very high efficiency ratings (110dBm) and thus need little power to reach deafening levels where other headphones can have really poor efficiency ratings (77dBm).
Most on/over ear headphones will have an efficiency between 85dBm and 105dBm.
Assuming ‘normal’ background listening is done at 70dB average SPL, more ‘active’ listening at 80dB SPL and ‘lively realsitic’ (a single song or part of it) at 90dB SPL it would appear we don’t need that much power.
When we like to listen to higher quality recordings, which as a rule have a higher DR rating as well, you need to add 20dB for the needed headroom.An SPL of 110dB needs to be reached without clipping to listen to the better recordings at ‘real life’ levels. Maybe have a few dB extra to spare for pure sonic bliss ?
How loud your headphone can play and how much power and or voltage is needed to drive your headphones to these levels is shown in THIS HEAPHONE POWER table. Remember… the powers and SPL mentioned are PEAK levels, average levels will be much lower.
DIY-Audio-Heaven 