Will a 2-recording approach work to remove ambient sound to isolate insect activity noise?

Hi. I am recording insect noise - buzzing, rustling in flowers, munching, etc. with a lav mic on a long skinny stick. My new goal is to remove ambient sound like planes, distant cars, etc.

Poor results (tons of ambient) so far with omnidirectional Audio Technica AT803B, which I had lying around. So switching to cardioid, something like a DPA 4080.

I’d like to record the insect with 1 mic and then ambient with a 2nd mic. Can I then use software like Audacity to subtract the sound in the 2nd mic from the 1st (since 1st won’t have any insect sound), giving clean insect sound?

Yes, but maybe a little more serious than you think.

This is two Electro-Voice 635A microphones in noise cancelling configuration.

They are three-wire, entertainment-quality microphones.They are connected to the sound mixer with a “Y” cable and the bottom microphone goes through a phase reverser. Pin-2 to pin-3 and pin-3 to pin-2.

Electro-Voice microphnes are famous for their ability to match each other in sound characteristics.

The upshot is total silence in the recording until the performer jams the microphones against their chin and talks directly over only one of the microphones. No academy awards given for sound quality, but you can do a video presentation in an insanely noisy environment and be heard clearly.

The further you get away from that ideal, the worse the result. Separate the microphones and the effect starts to go away immediately. Use two different microphones and there is no effect. Just distortion.

You can’t use Time Offset. Both recordings must be made at the exact same time.

You can try making a stereo recording with two identical microphones on the stick with one pointed backwards. You’ll need a test bed and the ability to listen in real time so you can mess with spacing and hear it change. If you try to do it with actual recording, that’s a retirement project. You’ll still be adjusting it when you retire.

Given your performers, you might try the two identical microphones facing the same way about an inch apart in stereo and then use the Audacity phase reverser on one of them. Anything you do that affects one without the other will damage the effect.

The effect is delicate. How are you going to get two identical microphones into the computer? You can’t use one of those inexpensive, simple headset adapter cables. Those are mis-matched right out of the gate. One has computer battery on it and the other doesn’t.

Post back if you get something to work.

Koz

You’ll need a contact-mic to have a chance of hearing an insect moving / eating,
(some of the nature documentaries fake the sounds :astonished: ).


“remove centre” is spectral-subtraction … https://manual.audacityteam.org/man/vocal_reduction_and_isolation.html

you might try the two identical microphones facing the same way about an inch apart in stereo and then use the Audacity phase reverser on one of them.

Missed a step there. Get the insect close to one of the two microphones. The other should be physically close enough to do some environment cancellation.

You have the same problem that people who graduate their podcast from one to two microphones have. How do you plug in the second microphone?

You could use two identical computers.

It gets crazy in a hurry.

An airplane going over isn’t a thing. What is a thing is super tiny air vibrations and it’s your job to capture the vibrations twice. One track with the performer’s vibrations included and one without—at the same time.

Koz

That microphone is an RE-15 highly directional microphone. Those vent holes in the head and the line of little holes down the side are to capture the sideways sounds and cancel them out—all inside the microphone. It only picks up from the front.

That’s another way to do it. A very, very, very directional microphone. That’s a shotgun used for an interview in a noisy environment.

He’s holding about a thousand dollars usd, not counting the sound mixer and recorder.

Koz

One more. That shotgun has a wind sock on it to help with wind noises. That will be a problem for you being out in nature and all.

Koz

https://www.youtube.com/watch?v=6fp-A3CidAk&t=111s

No experience with contact microphones. You’re on your own there.

Please note theyr’e doing that in a completely, totally, 100% soundproofed room. That’s what those wedge-shaped things on the wall are.

Koz

Perhaps worth considering Rode NT5 MP (matched pair). They have extremely low self noise and a pair will cost considerably less than a single DPA 4080.

Whatever the mics, you will also need an extremely quiet pre-amp.

Laser mic ? … https://youtu.be/y7qMqAgCqME?t=105

Wow this is brilliant! I’m really busy getting ready for shoots tonight and tomorrow, but very excited to dig into everything on this post!

Some quick thoughts. Lots of subjects are flighty and so a small pair of mics is best. The Rode might be workable. 1/2" diameter is about what the current rig is. I see lots of ‘miniature’ mics, but none that are cardioid. Small mics on a skinny stick get through plants and shrubs easier to follow the insects as they move around.

Good on the preamps. Using Zoom F6, lovely thing!

Question. Can this cancellation be done after the fact? All recordings are timecode synced. Can record to 2 different mono channels and then apply the effect on a desktop PC?

Uh-oh, my early adopter problem is flaring up. Lazar mic??? :open_mouth:

All recordings are timecode synced.

That will get you down to the video frame enough for high quality separate theatrical sound, but not nearly good enough for cancellation. Record the two microphones as a stereo pair so exactly the same sampling signal is used for both.

Tape the two microphones together and speak into the combination from about a shaka away.

Make the stereo recording and export it as WAV into Audacity. Play it to make sure it’s a normal recording. Split stereo to mono with the drop-down menu on the left, select one track, and Effect > Invert.

Play it. The speaker or headphone sound should drop dead as the two channels cancel. Hopefully.

If they do, then that’s your basic microphone equipment list and it’s up to you to get the mechanical spacing, positioning, and aiming right. One mike picks up a bug and environment and the other the environment by itself.

Koz

If you go with the frontward and backward microphone trick, you shouldn’t need to Effect > Invert. They should cancel automatically…it says here.

Koz

Unlike the invert phase-cancellation method, with spectral-subtraction 0.05ms accuracy is not necessary,
+/- a video frame would be close enough. [ Spectral-subtraction does generate computery artefacts though.]

If Audcaity’s native isolation tool doesn’t do it for you, there is a free plugin called Kn0ck0ut …https://youtu.be/evsnple7k88?t=388

Only if you must record genuine insect sounds, (rather than Foley), & money is no object …
https://www.youtube.com/watch?v=yRxYIfZUxW8
https://en.wikipedia.org/wiki/Laser_microphone
https://en.wikipedia.org/wiki/Laser_Doppler_vibrometer

I’ve been recording all mono channels on the F6. It can record channels 1 and 2 as single stereo file. So is it enough to run a separate cable to input 1 and 2 this way? I don’t need the Y adapter then right?

Now on to a fun homework assignment… to learn how invert phase-cancellation and spectral-subtraction work. Yes the TV news sucks every day, but this is stuff seriously fun!!! :laughing:

If you make an in-sync* stereo pair of the insect mic and the ambient mic,
then try Audacity’s isolation tool on that stereo track, set to remove centre.
That will reduce the frequencies common to both mics, so, theoretically, isolating the insect from the ambient noise.

[* synced as best you can but, unlike phase-cancellation method, synchronization but does not have to be sample-perfect with spectral-subtraction method]
suggested settings __.png

okay beautiful! I’ll try spectral. Since Audacity splits the stereo signal, I can just record each mic as a mono channel, no?

Well maybe not. I jam sync timecode with Tentacle. Since Audacity doesn’t read that or care about frames, I need to record stereo. Splitting tracks will keep them lined up. Is that right?

You should get your recorder to produce a stereo track. That means exactly the same digital sampling signal is used for both. If you split them later in Audacity, the relationship should stay.

Do that speaking test first. If that fails to cancel, then none of the rest of these jobs are going to work. Turn off any H6 effects like auto volume control, auto limiter, rumble, or wind filters. Also, you have to figure some way to get the two recording volume controls exactly the same. My H1’s recording volume control is smooth. It doesn’t click. If you miss the match here, you’ll have to do it with the volume controls in Audacity after the split.

Cancellation tricks are not for the easily frightened.

That Y-Cable trick only works with broadcast and pro microphones. In general ones that use the 3-wire XLR cables.

They send the sound down the cable twice. Once right-side up and once upside down. That’s pins 2 and 3. Pin 1 is the cable protective shield and not part of the sound. If you reverse two and three before the recorder, then the microphone sound will reverse. Two-wire microphones or digital microphones can’t do this trick.

An accidental reversal can cause magic sound failures. In shops that do their own electronic servicing (raising hand), it’s super important that cable repairs put the colored wires back exactly the way they came out. You can accidentally get a 2-3 swap and that will give you one microphone and cable that mysteriously sounds “funny” when used with all your other microphones. It’s phase cancelling, exactly the effect you’re trying to achieve.

A cable checker will resolve this.

Koz

Thanks again. My system is XLR only and I am eyeballing a pair of Sennheiser MKE105s for this purpose.

Wondering again about which approach. Some of the noise I wish to eliminate is transient, like planes flying overhead. Does either approach have a preference for uniform noise vs variable and transient noise?

I guess insect noise is transient too, so is it even possible to remove temporary noise (that may overlap with the desired signal) if that noise is not the insect subject?

That seems optimistic: these BBC guys were having difficulty recording individual insects in an anechoic room.

Felix Blume made a portable bee-studio in the field to exclude ambient noise …
recording a bee.png
https://felixblume.com/essaim/

The two microphone cancellation method—no matter how you do it—depends on both microphones being recorded at the exact same time and in close proximity to each other. This will suddenly make a lot more sense to you when you make the first recording. Tie two microphone heads together and do a simple voice recording.

I’m concerned that only B&H PhotoVideo claims to carry the microphone, but they won’t sell me one because they insist it’s no longer available.

Is this something you found on eBay?

B&H does have tutorial videos on how to shoot video. As far as I watched, they do very well. They’ve done this before.

https://www.bhphotovideo.com/c/product/798270-REG/Sennheiser_MKE105S_4_MKE105_Lavalier_Microphone_Kit.html

I know nothing about the spectral cancellation technique, but pay attention to the caution that the sound can have odd sound errors and tonal distortion.

Koz