10

Usually it's pretty quiet near where we record, but it's summer time and cicadas are loud as hell.

It doesn't matter what kind of direction the mics are pointed, the sound of the cicadas always gets picked up!

Without sound proofing (to a ridiculous extent) is there anyway to block out the sound of humming cicadas?

  • 2
    While this question is IMHO on-topic here, I believe you can find the right audience at Sound Design.SE. – yo' Aug 4 '15 at 6:40
  • 3
    You might be able to put a mic behind your recording subjects to try to pick up just the cicadas. Record this to a separate track, then subtract it out. It won't be perfect and it probably won't be clean, but if you don't overdo the subtracting and you get lucky it might be an improvement? – lc. Aug 4 '15 at 7:44
  • Why don't you incorporate that comment into an answer Lc? Seems to answer the question pretty nicely. – Neil Meyer Aug 4 '15 at 8:41
  • You could try to scare off the cicadas with fireworks. When i was a kid, we'd do it for fun, and it actually worked fairly well. – MeesterTeem Aug 4 '15 at 15:20
  • Try recording at night when the cicada's are asleep. Then you will get tree frogs instead but they may blend better with your recording. – Rockin Cowboy Aug 5 '15 at 4:18
9

Humming would be doable but we are more likely talking about chirping. Which has a lot of harmonics. When you are filtering out those, you are missing out on the useful audio information in that range. And we are not talking about a single cicada here but a whole bunch of them so you have no real chance at modeling the sound and subtracting a predicted value.

So no, you are out of luck. Noise reduction techniques will work for intelligence purposes, uncovering intelligible speech. They will not work for musical purposes as they'll add warbling and waffling components ("musical noise") to the sound. You'll be able to improve things somewhat through appropriate microphoning (employing mics with a close-captioning characteristics) but what can be done in that regard is in the "somewhat" realm.

3

The only way to mitigate that much noise is to A) use hypercardioid mics, (no omnis!) B) very close micing of instruments; C)close all windows/doors and tape/seal the edges.

another possibility if nothing else works is to set up a noise-cancelling mike array. Two identical mics, set about 1 1/2 feet apart, sing/play directly into one. Take the signal of the 2nd mic and invert the polarity, and mix it in with the signal from the first mic. Any sound that hit the two mics simultaneously (i.e. distant cicadas) will cancel out. This won't work if you're recording multiple instruments in the room, or at least it's a very very long shot!

  • I think that this may work, especially if you take care to mute the area with blankets or felt to reduce reverb which scatters the sound and makes it more chaotic to cancel. For that matter you may have luck muting most paths by which the sound can enter your recording space. Cicadae's song is pretty high and, consequently, directional, such that you may be able to block it via occlusion much as you would light. – Epanoui Aug 13 '15 at 1:50
2

Regarding the suggestions about noise cancellation: they won't work for cicadas. There are basically two approaches when working with multiple microphones: phase independent filtering, and phase-specific filtering.

For the phase independent filtering, you get the characteristics of the noise and dampen the frequencies according to the expected signal-to-noise ratio in all of the related frequencies. Since cicadas are wide-spectrum noise, you'll be hitting all of the good content as well. So it's only workable when generally you are either quite louder than the cicadas or silent. This sort of frequency-dependent noise-gating cannot help with content at quiet levels, only with full silence in a particular band, and cicadas are wide-band. So either you are above cicada level in some frequency and the gating does nothing, or you aren't, and the gating puts you down along with the cicadas.

For phase dependent filtering, you basically do the same as "echo cancellation" (which subtracts the phase-corrected microphone input expected to result from the current speaker output from the actual microphone input before sending it on, so that as little as possible from the speaker output is sent back on the line) in a phone communication does: you try subtracting a phase-corrected version of the noise from your main microphone signal. This can work with noise levels above the signal level but it requires correlated versions of noise on the mics. A single cicada in the room (or even a single responsible opening) would be a good use case.

However, cicadas distributed around the house and with their noise coming in at multiple places are producing ambient sound where you don't really have reliable phase relations. Also you need some separation of "noise mics" and "signal mics" so that the noise mics don't get too much of the useful information. The annoying part of the chirps is at a wavelength quite below one foot, so echo cancellation techniques will be quite sensitive to exact direction/correlation. Anything but direct sound will not cancel well.

So your best bet, if the neighborhood is suitable, is likely recording sessions at late night. The next best bet is passive dampening as much as you can. If you cannot bring the cicadas down to a level where they are no problem, at least you might have a chance to bring them down to a level where noise-gating will be effective, namely where you need to only rescue actual silence from the cicadas.

1

Sure cure. Buy a few furniture blankets and surround the instruments - like a four wall little fort. That will deaden any noise. If you are in a small room, just cover the windows.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.