Isolating test tones in soundproofing measurements

Hello

I am dabbling in soundproof construction (not professional). I built myself a workshop and I want to test its acoustic insulation performance, and the difference before and after double glazing (DIY acrylic).

So anyway, to do this test I used Audacity to generate a sequence of pure tones at the standard OITC (like STC) frequencies, ie
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000

  • each 0.7 seconds long.

To define a baseline in open air I played this through a bluetooth speaker and recorded it on a Rode VideoMic at a distance of 2m.

Then I selected each time period and measured the RMS volume:

Frequency Open air inside
80 -24.6
100 -16.7
125 -13.1
160 -12.1
200 -0.8
250 -10.2
315 -3.2
400 7
500 -4.4
630 12.1
800 7
1000 10
1250 0.8
1600 -8.9
2000 -5.7
2500 8
3150 2.6
4000 7.4

I also measured the background sound level at about -47 dB. This was inside a quiet room.

Fine. The problem is that when I record through a wall from outside, the sound is faint, and it gets mixed up with other noise.

You can see the signal of interest on the spectrogram. What I would like to do is to isolate or measure the volume of that specific tone in the recording, without the other noise.

This seems like it should be possible, but I couldn’t figure out how. I’m new to audio processing. A windowed minimum filter across the spectrogram would do it?

The only way I’ve found to do this is to

  • adjust spectrogram settings to gain = 0dB, range = 80dB
  • screenshot the spectrograms
  • bring in to GIMP,
  • extract the lightness values at each frequency of interest
  • convert back to dB

…assuming that the colour function is linear between -80 and 0 dB.

It broadly matches the numbers I got from “Estimate RMS”.

This topic was automatically closed after 30 days. New replies are no longer allowed.