>I am getting lazy and hoping someone will take pity on my ignorance.
>
>I have a data for a line graph that will plot the volumn (amplitude), in decibles of a fog horn at regular intervals (period frequency), say every 10 minutes.
>
>I want to resolve the average volumn (amplitude) for a segment of time. Is this an RMS (root means squared) solution - or is there another method that I should use!
>
>Thanks!

How can I not have pity on the lazy, since I am myself very familiar with this condition!

RMS seems appropriate for things like voltage or current, since the power is proportional to the square of the voltage, for instance.

Decibel is the logarithm of the power (times 10). For instance, 50 decibel is 10 times as much power than 40, and 60 decibel is 10 times as much power than 50.

Therefore, if it is average power you want, you would have to convert the individual readings from decibel to power units, take the average of that, and then convert back to decibel.

I am not sure whether a straight average of the decibels would have significance - perhaps it does.

Sample calculation (converting to power):

Sample readings:
30 decibel
35 decibel
40 decibel

Convert each one to an arbitrary power unit. For instance, divide 30 decibel by 10 (= 3), and take the antilog of that (10^3 = 1000).

This gives you:

30 decibel = 1000 power units
35 decibel = 3162 power units
40 decibel = 10000 power units

The average of that is 4721 power units, equivalent to 36.7 decibels. The straight average would give you 35 decibels.

HTH,

Hilmar.