# Spectogramm Plot analysis

Hello everyone,

unfortunately, I have little knowledge about the topic and therefore would like to turn to you experts for help. My own research hasn’t really led me much further either.

I have made several recordings using microphones with the aim of understanding how well insulation works. Unfortunately, I’m having trouble interpreting the frequency analyses. I was able to determine that the dB unit on the y-axis represents the difference from the reference sound pressure level (which I believe is supposed to be a sine wave at 0dB).

Here are the two spectrograms:

I am sorry for the bad quality of the picture but I am only allowed to upload one picture. Briefly, the peaks are at about 6000Hz at -37dB with insulation and -52db with insulation.
Does this mean that the insulation shields around 15dB at, for example, 6000Hz, while it shields only about 2-3dB at 1000Hz? (Which would be expected)

How can I determine the sound pressure level at 6000Hz?

In other words, how can I make statements about the volume of the sounds at specific frequencies from such plot spectrograms?

Thank you in advance for your support, sorry, I’m really a complete beginner at this!

Best regards,

Franz

Can only determine relative changes, (e.g. the 15dB reduction), not the absolute values.
Need a calibrated device to measure absolute values of loudness …
https://www.amazon.com/s?k=loudness+meter&s=price-asc-rank
(any old mic will not do)

I’m curious - Is that insulation in the wall, and are you measuring through the wall? I would have expected to see more attenuation at 1kHz. (But insulation isn’t really “soundproofing”.) If it’s on the wall and you are absorbing reflections in a room, that’s more like what I’d expect to see.

Correct!

With an SPL meter.

There are SPL apps for smart phones, but the microphone sensitivity on phones is not calibrated so you don’t know if they are accurate or not. (Probably “better than nothing”.)

But… Things get tricky…

1st the “easy part”. Digital dB levels are dBFS (decibels full-scale) where 0dB is the “digital maximum” and digital levels are normally negative. The 0dB reference for SPL is approximately the quietest sound humans can hear and SPL levels are positive.

But a dB difference will be reflected in both measurements… If the digital level drops by 15 (15dB more negative) the SPL also drops by 15dB (that’s assuming nothing else is changed and everything is linear so an amplifier isn’t overloaded and clipping or anything like that).

SPL measurements are usually A-Weighted to better-match our hearing which is most-sensitive to mid-frequencies. A meter with a Z-weighting is technically more accurate over the frequency range but not necessarily “flat” over the full-standard 20-20khz audio range.

Microphones aren’t flat either.

None of that matters if you’re just trying to measure a difference, with and without the insulation.

Measurement microphones are “calibrated”. They are individually measured and they come with a calibration file so you (or the software) can correct over the frequency range.

USB calibration mics are calibrated for SPL, but with analog measurement mics, the electrical & digital levels depend on preamp gain and the analog-to-digital converter so if you want SPL (in addition to frequency information) you have to do your own calibration with an SPL meter.

A frequency sweep or discrete measurements every so-many Hz (or fractions of an octave, etc.) is usually better than a spectrum.

There is some FREE software called Room Equalizer Wizard which is intended for a different purpose (for measuring/adjusting frequency response in a room) and it’s intended for use with a measurement mic, but it MIGHT help with what you’re doing.

Hey everybody,

thank you very much for your help and sorry for my late reply.

Then I will simply stick to a SPL meter to measure the volume.

To answer the question of DVDdoug. It is not an insulation in the wall. It is more or less foam that should isolate and seperate a box into two rooms which are then next to each other. One room should be very quite in the end and in the other room music should play.

Thank you once again for your help.

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