Published on March 4th, 2013 | by peterb2 Comments
Microphone specs 101: signal-to-noise ratio
Look at your microphone spec sheet. Does it have a signal-to-noise ratio? Do you know what that means? If you’re pursuing a career in audio engineering, you should.
The signal-to-noise ratio of your microphone describes how much noise your microphone is making (also known as self-noise) compared to a reference signal, usually a 1 kHz, 94 dBSPL tone. Let’s use Alchemea’s AKG C414 condenser microphone as an example. As you’re recording, the electronics in the AKG C414 are making some noise – it’s difficult, if not impossible, to find any piece of electronic equipment that doesn’t generate some noise on its own. Also, be aware that some self-noise isn’t actually being generated by the microphone’s electronics, but by random molecules hitting its diaphragm.
Based on the mic specs, we know that the AKG C414’s signal-to-noise ratio is 88dB. That means that when we used that 1 kHz 94dB reference signal, there was also 6dB of noise produced and output by the microphone. The signal (94 dB) minus the noise (6dB) gives us a difference of 88dB. Incidentally, 6dB of noise is very good! The higher the signal-to-noise ratio, the better. If your microphone doesn’t give you a signal-to-noise ratio as a spec, you can generally figure it out by subtracting its self-noise from 94.
“But wait!” you say: “On my microphone spec sheet the signal-to-noise ratio is ‘A-weighted!’ What does that mean?!”
Welcome to the wonderful world of how we perceive sound.
‘A-weighting’ can be thought of as an adjustment applied to sound level measurements. In 1933, two researchers – Harvey Fletcher and Wilden A. Munson – discovered that we are more sensitive to certain frequencies than others (which we cover in full detail in our Studio and Live Sound Diplomas )
A-weighting takes this research into consideration, and “adjusts” a sound level measurement according to how we would actually perceive it. Let’s use a typical problem in vocal recordings: the rumbling of a big lorry or train going by. Rumble can contain very low frequency sound, which we’re not particularly sensitive to. If we used a mechanical sound level measuring device, it might tell us that the rumble is 22 dBSPL. But we are not machines, and in our minds, the rumble is quieter. There is no sense in having technical information about a microphone (or any other piece of audio gear) that has been interpreted by a machine. Until such time as robots can mix an album, technical information has to come as close to our perception as possible. So we use some fancy arithmetic to take that difference into account, and arrive at a new number that corresponds to what we humans actually perceive when we hear the rumble. In this case, it’s 12dB. Written correctly, we would say 12 dB(A), for A-weighted.
A-weighting isn’t the only ‘adjustment’ available to sound pressure levels, but it is probably the one you will run into the most.
By now you will hopefully feel slightly more confident about using or purchasing a microphone, but there are plenty more numbers on there to go through.
Help us choose which one to decipher for you next by leaving us a comment below.