The decibel works on a logarithmic scale, not a linear scale.
That is to say, 80 dBs is 10x louder than 70 dBs. But that also means that 0 dBs is just 10x quieter than 10 dBs, **not** that it is “0 sound” the way that 0 km/hr is 0 speed.
So negative decibels are just some multiple quieter than that 0 dB point, which has a nonzero amount of sound.
Edit: it’s bothering me a bit that I didn’t define “quieter” better. The precise measure is that of energy intensity. So the pressure waves of a 0 dB sound have 1/10th the energy of a 10 dB sound. Our ears measure this difference in energy as loudness or quietness.
Sound is basically vibrating air. Vibrations in the air means the air is pressurised which can be measured to know the intensity of a sound. Since using regular units of pressure isn’t very clear, we use the decibel which makes it more clear by taking the human perception of the sound into account so 0dB doesn’t mean that the air isn’t vibrating, just that we can’t feel it vibrating. Thus a negative level of decibel doesn’t mean there isn’t any sound, more like we can’t perceive it with our ears.
Decibels is a log scale, and it doesn’t start at *zero energy*.
The first part of that means that it’s not that going from 10dB to 20dB is 2x more energy, it’s actually 10 times more energy! And 30dB is therefore 100 times more than 10dB.
The scale starts at the lowest possible energy for the average person to be able to hear. That doesn’t mean there’s no sound – no energy – it just means that it’s too low for most people to hear most of the time. -1dB is, therefore, 10x less energy and is even more quiet. You already probably couldn’t hear anything, but now you *really* aren’t likely to hear anything.
It’s a bit like how you can have negative temperatures when using Celcius and Fahrenheit – the scale doesn’t start at absolute zero, it starts at a “zero” that is meaningful to us.
Decibel is a logarithmic scale not a linear scale. A change by 10 dB is the power of a 10x, but that is true for all changes of 10dB. So a change of 20 dB it two changes of 10 so the power is 10x 10 = 100 times higher.
It is also true if you go from 10dB to 0dB it is a power change from 10 to 1 not to zero. You define the scale at 0dB =1 power with the unit you use.
So -10dB is 0.1 , -20dB is 0.01 you need to reach – infinity dB to reach 0 powered.
That is for power if you look for amplitude it is a factor 20 dB which is a change of 10x except for that everything works the same.
The result is -23 dB is around 1/100 of the reference in the power scale or 1/10 the reference in the amplitude scale.
A decibel is a relative measure of the power of something. Relative means that it is compared to some reference. In the case of decibels, it’s expressed in a logarithmic scale — you don’t really need to know what a logartihmic scale is, but it means that 0 is where the power is the same as the reference (**not zero power**). 10 decibels is 10x as much power as the reference, 20 decibels is 100x as much power, 30 decibels is 1000x as much power, and so on. It also means -10 decibels is 1/10th as much power, -20 decibels is 1/100th as much power, -30 is 1/1000th as much power.
When we talk about sounds, the reference unit is 20 micropascals in air and 1 micropascal in water (it varies depending on the fluid). A “pascal” is a unit of pressure, and those values were selected as the the average amount of pressure change that a person with good hearing can detect.
-23 decibels, therefore, is an amount of pressure (specifically 10^((-23/10)) x 20 mpa = 0.1 micropascals, in air); anything below 0 dB, a typical human can’t hear.
Decibels are a unit describing the ratio of two values. When used for amplifiers or attenuators this ratio is output to input. When used to describe a single value, like loudness, the ratio is between the measured value and a fixed reference value.
When describing sound levels that reference value is a sound pressure of 20 microPascals in air (because sound is just air pressure)
Decibels are a logarithmic representation of the measured air pressure divided by this reference air pressure. Specifically 10*log(measured pressure/reference pressure).
The way logarithms work , if the measured pressure is bigger than the reference you get a positive result. If it’s smaller than the reference you get a negative result, and if it’s exactly the same, you get zero decibels.
Decibels tell you how much louder a sound is than some reference level of sound pressure. The reference that is nearly always used is a pressure of 20 micropascal, which is considered to be the threshold of human hearing. So 0 dB is the point at which a sound is exactly as loud as this value, meaning you might just about be able to hear it. Anything below 0 dB you most likely won’t be able to hear at all.
The way decibels work is that they are on a log scale. You start with a ratio: the ratio of your measurement to the reference level. For instance, if we take 20 micropascal as the reference, then a sound pressure of 2 millipascal is 100 times louder than that. To convert this to decibels, we calculate the logarithm of 100 in base-10, and then multiply this by 10. The base-10 log of 100 is 2 (10^(2)=100), and so a sound pressure of 2 millipascal corresponds to 2*10=20 dB.
In other words, every 10 dB increase means the sound gets 10 times louder. Every 3 dB corresponds (approximately) to a doubling of the loudness.
If our measurement is equal to the reference, i.e. also 20 micropascal, then the ratio between them is 1, and the logarithm of 1 (in any base) is 0, which is how we get to 0 dB.
How about negative decibel values? Well, they arise when your measured sound is *quieter* than your reference. For instance, a sound pressure of 0.02 micropascal is 1000 times quieter than our standard reference, and so that corresponds to -30 dB. In the case of a -23 decibel room, that corresponds to a sound pressure that is 10^(2.3) ≈ 200 times quieter than the threshold of human hearing. The way you measure that is obviously not using human ears, but with a more sensitive sound level meter.
Latest Answers