As Koz says you will have to obtain a rumble-free comparison to confirm the 19Hz peak is the mystery sound and not an artifact.
I’ve often seen peaks on recordings around 20Hz which are artefacts, i.e. they are generated in the apparatus and appear on all recordings.
BTW I can tell when my neighbour’s central heating pump is on : it produces a very low frequency hum around 20Hz.
It is of such a low frequency that, although audible, and slightly annoying, it may not be recordable using a regular microphone, or reproducible on typical headphones/loudspeakers, (although if recordable it would show up in an Audacity spectrogram).
If this is the origin of your LFN then the clue would be it only occurs in autumn and winter when the central heating is used.
If you speed up the recording (which increases the pitch) it does sound like a motor speeding up …
The low frequency component can be isolated and amplified using the Equalization effect. This file is inaudible on my laptop speakers (they are too small to reproduce such low frequencies) but very clear on my big speakers or headphones. Is this the sound that you hear? As others have said, you need a recording made with the same equipment with everything set up exactly the same as a comparison to be sure that it is the environmental sound that is in the recording and not just noise created by the equipment.
Probably the only way to determine the source of the noise is to catch the noise on a recording exactly as it starts, then match the time with an event such as central heating switching on or a generator starting up.
Thanks very much for the replies, I’m afraid I have been doping myself up to try and get some sleep (I’m like a different person when I do get some), so my response has been late, please accept my apologies; I was not ignoring anyone.
So here is a clip with No LFN (recorded sometime late one morning), to be honest I cannot find anything in it which resembles the sound I am hearing so that is good news, but if you listen to that YouTube clip (at 1min 45 seconds onwards) the LFN that is disturbing me is not that far removed from that (perhaps a very slightly faster beat/ tempo).
It cannot be a coincidence that just about every night I turn off the lights, I can give it about ten or fifteen minutes and then the LFN begins (there are nights where absolute silence reigns so I know it is something that is manually turned on then left running) (fairly certain it is not central heating, as we have had some bitter days and nights this Winter and if it were central heating I am sure it would have been turned on during the day, but of course I cannot rule it out completely).
The recordings are using exactly the same equipment and taken from the same place. However my competence at using the equipment may well be at fault.
So my guess is that nothing can be found on that recording.
However here are two more (one recording with the microphone in the centre of the room, the other with it pressed up against the party wall), but this time the LFN was loud and I mean LOUD
I’ve been playing around with the Low Pass Filter after I’ve Amplified the file (a la Koz from his first reply) and I reckon I have actually/ finally/ hopefully captured something (what I can only describe as a rhythmic pulsing) so if anyone is interested maybe they can take a listen and share their thoughts and/ or expertise.
(I used Koz’s Amplify OK then LPF 36db 60Hz, then 40Hz, but 40Hz gets a bit lost on my iMac speakers)
…Edit: Well just reviewed some of the older recordings (same settings as above) and to my ear it looks as if I’ve captured nothing even in this most recent set of recordings: the LFN was so disturbing I felt sure I had got it … This noise is busting my ba$ll5.
If anyone knows of a way to capture LFN then drop me a line cause I have now run out of ideas.
I think we’re out of ideas.
The EXM8000 microphone is good down to about 15 Hz, so if there is any loud sound above 15 Hz then it should be picked up quite clearly by that microphone without the need for any processing. Sorry but I can see no physical explanation for why your repeated attempts to record the noise should fail.
A portable seismograph may better suit your needs. At least if the phenomenon is vibration-based.
There is sure a way to make those pulses visible.
You could for example observe how a detached laser beam (laser pointer hanging from the roof) - reflected by a looking glass on the shared wall - moves.
(picture with long exposure or similar procedure).
Or some classic pendulum method, although the recording method had to be adapted. It is not likely that you have a high precision plotter at home that doesn’t produce vibrations by itself. rather use your microphone with something that slightly scratches on it or some inductive recording method (magnet over electric circuit).
Hmph, I am at the end of my tether for the moment.
One of the problems with doing these sorts of measurements is calibrating the sound chain. For one example, RS is making conclusions about the sound capture using the built-in speakers of an iMac. The iMacs I’ve used have terrible speakers for just music and are completely out of their league for low frequencies. It’s a given that they won’t deliver enough volume to simulate the house moving.
Although the microphone would have no trouble delivering perfect sound down to 15Hz and lower, nowhere is it written that the rest of the microphone chain will do that. And past moving the whole system to the airport to record a jet taking off, there’s no way to test or calibrate it.
I can think of several ways to test and measure this kind of thing, but I’m sitting in a house with high-end audio test equipment trying to talk a non-techie down from many multiple time zones away.
One fluffy thought I had was to put a good signal generator into a not particularly powerful speaker amplifier and then on to a pair of “open air” headphones. During an episode of noise, move the frequency and volume of the signal generator until it starts beating with the noise – according to your ears. No microphone or fancy sound system needed. Read the frequency from the generator dial. If you have enough time and a good cup of coffee, you should be able to derive all the LFN characteristics.
Audacity will not move its generator in real time and trying to do it in post production would be a nightmare. Nobody but me has an actual physical sound generator which will do that.
So as we peel off elves one by one and fail test after test, we’re left with the original two suggestions: move or contract with the company whose job it is to do this.
If you lived in Texas, you could get your Smith and Wesson out of the cupboard and have a nice discussion with your neighbor about the wisdom of continuing the noise.
I’m extremely grateful to you for all your suggestions, and the time an effort you have put in to help me, you all went above and beyond so you have my thanks, a thousand times over.
This last episode was so loud that it had me call out the ol’ Bill at 0440, but they decided to come two days later which was unhelpful. However, even the ol’ Bill had to admit that given the location of the house and the surrounding area the only source of the noise can be from the neighbour’s house.
Given the fella next door is what I can only describe as an attention seeker that sits all alone in his sh1t-h073 submerged in paranoid delusions; I suppose it is through the creation of this disturbance that he gets his kicks.
I think I will take your advise and pay someone to come and set something up to monitor this (given the lottery when it comes to people claiming to be experts over here, I’ll just have to take my chances).
