How to clean up waveform (NOT the audio but the visual representation of the waveform)

[unknownsailor]

My GoPro audio waveform looks nothing like my other waveforms, despite being recorded on an external mic. When imported it looks like this:

If I slice this at any point, it results in clicking, unlike my waveform from my GH5, for example. But if I apply a simple filter like a high pass, the waveform now looks like this:

This results in a much 'cleaner' waveform, visually speaking. It's like it balances the audio along the 0db line, making it easier to edit, and now looks like waveforms from my other sources.

How do I achieve this without applying any filters to the waveform? And why does it do this? Why does this waveform look so 'chaotic' on import? I'm sure this is more than just audio quality or noise. The actual audio from the GoPro with the external mic is pretty clean, it's just the visual representation of the waveform that's different, or how it is interpreted by Audition upon import. Possibly. I'd like to understand why this happens.

Any pointers are appreciated, thanks.

[SteveG_AudioMasters_]

Could you post a snippet of actual audio, just as it comes from the GoPro?

[unknownsailor]

Hi Steve, sure. I'm away from my computer right now but I'll post up a sample and more info shortly.

[unknownsailor]

Hi again, Steve,

Here are two audio clips:

Original clip 

With high pass 

The first is the original, and the second is with the high-pass filter, both saved as .wav.

I've also taken a screenshot of the whole audio file. This is without the high pass:

And this is with the high pass:

[Richard M Knight]

Are you zoomed in the same amount in both pictures? The top one looks to be more of a 'sample' level

[unknownsailor]

Hi Richard,

Yes, they both show the same waveform at the same scale.

[SteveG_AudioMasters_]

Okay, I can tell you what's happening - and yes I needed the first sample. It's slightly ironic considering that the first sentence was 'absolutely no wind' - because that's exactly what is causing this to happen. Even a gentle breeze blowing across a microphone will cause the diaphragm to flutter, and that's what is causing the visual disturbance. GoPros tend to have fairly good bass responses (nearly an omni capsule) and of course no wind shield. And that's all it takes. The LP filter is the correct thing to do, as I don't think it's going to be easy to put a wind filter on the GoPro, somehow - and that's what you'd have to do to avoid this happening at the recording stage.

[unknownsailor]

Hmmm, I'm not so sure. For starters this is an external mic (Rode Micro) plugged in via the mod cage for the GoPro. The mic has a deadcat on it. And this occurs irrespective of location, inside or out, wind or no wind, handheld or mounted on a tripod.

[SteveG_AudioMasters_ · Accepted solution]

It doesn't have to be breeze - any shift of air across a diaphragm that responds nearly down to DC can do it - even air currents in a room. If you zoom in and look carefully, you can see the straight parts of the waveform where the diaphragm has been deflected. They don't look the same as audio, but the audio is modulated onto them:

If you want some proof that it's this then all you have to do is set the scientific filter to a really low value - much lower than the High Pass filter rolls off - like say, 10Hz - and apply that. This will shift anything below 10Hz (which you won't hear, but can see) back to the baseline. If you want a more sensible value to use, then try 25Hz - still below any audible range you'd be interested in.

I should point out also that this is why decent mics have HP filter switches built into them... it really does take more than a wind gag to stop this.

[unknownsailor]

Ok, I'll have a play with those alternative frequencies and set it as a favourite. I just need to find one that doesn't have the effect of a high pass, so any that isn't discernable to the ear will work. Thank you for your assistance, Steve.

[unknownsailor]

Here are some waveform comparisons, showing one second of audio with a Scientific Filter applied.  This was recorded outside with some background noise.

The 50Hz cut is close to tidying up the waveform so I can slice without clicks.

Thanks once again, Steve.