Looking for a cropping work-around

Hi, without going too deep into detail, I have maliciously noisy neighbors and I’m trying to build a case to bring to the authorities. I’ve finally opted to start audio recording all the wall banging that goes on from 5am to 2am (yes, just shy of all day, daily). I did a bit of quick research and decided to pick up audacity for my needs. The software does an excellent job for my recording needs, now I’m just trying to find a way to trim away all the ‘dead air’ between bangs.

I have two schools of thought for how to approach this, but since I’ve only been using the program less than 2 days, I can’t seem to figure out how–I’m hoping one of you knows a work around. Approach 1) Sound activated recording – this option is almost exactly what I need, but it cuts away (off) too quickly, is there an option somewhere to set the delta delay for how quickly the mic shuts off, after dropping below decibel threshold? I.E. cut mic off (5 seconds) after falling below sound activated threshold instead of immediately. Approach 2) Trimming tool – I don’t know if audacity has a tool like this but it’s not too hard to imagine, is there a way I can select an option to crop out all parts of a track that are below a certain decibel, then consolidate the remaining chunks back together? Essentially a retro-active noise-activated tool.

The second method might be preferable since I can capture an entire day of audio, then go back in and clean it up without fear of missing evidence because the bangs weren’t loud enough to turn on sound activated, but it also seems like it would be harder to implement/involve more work on my part.

I’m using audacity as a noisy neighbor audio “security camera”. Need to find a way to take a 24 hour track recording and isolate out intermittent wall banging. Trying to find the noise spikes and export them manually is too much work, so I’m seeking a work around. Contemplating sound activated recording with an extra buffer (maybe 5 seconds) for deactivating microphone, or using a tool that can scan through my track and delete out parts below a set decibel threshold.

Also open to any other suggestions; while coming to this forum the word macros caught my eye, it seems prudent to mention I write my own macro scripts in AutoHotKey.

I’m trying to build a case

You need an actual security company or service.

I know Audacity’s price was attractive, $0, but Audacity can’t be used for surveillance, security, law enforcement, or conflict resolution.


Perhaps worth getting in touch with the authorities and making a complaint.

A “home recording” is not evidence - your equipment is not calibrated so there is no evidence of how loud the noise is. Monitoring environmental noise requires calibrated equipment (expensive) which the authorities should have access to (or can be hired). Hire costs for such equipment are usually several hundred dollars per week.

Hi guys, thanks for the quick replies, but I think the responses got a bit de-railed. I’m looking to take an audio clip that is 24 hours long and cut out the 90% dead air. Up until now the best method I’ve found is to go through the ginormous clip meticulously, highlight the loud noises, do a CTRL-B label, and then do a multi-export. I reckon there is a less labor-intensive way to do what I need, suggestions?

BTW, Running windows 10
Audacity V2.4.2

You could try the “Truncate Silence” effect; https://manual.audacityteam.org/man/truncate_silence.html

A spectrogram could give an overview of when sounds are registering during a day.

So a record of the timing of the bangs and their relative volume, but not their absolute volume.