Seemingly identical configuration, different results

[Sentient Mud]

Hi,

I'm trying to improve the acoustics of my drum practice space using REW, mostly by adding/moving rigid fiberglass panels and looking at the effect on the RT60 graph.

I just did an experiment thought that made me lose confidence in my tests: I simply ran the measurement process multiple times without changing anything -- just running the measurement process multiple times -- and got very different looking RT60 graphs, particular when looking at Topt.

I'm hoping someone here might be able to clarify what I'm seeing. Attached below are three of the measurement that I thought should give identical results. The only variable I can think of is I might have moved slightly between measurements (I'm in the room under test).

Thanks in advance!

Dan

 

[John Mulcahy]

Can you make the mdat file available somewhere? Don't think you'll be able to attach yet. Variations at the low end are not unusual as RT60 isn't really meaningful at low frequencies, it is based on an assumption that no longer holds in small spaces at low frequencies.

 

[Sentient Mud]

Hi John,

Thanks for the clarification re: RT60 at the low-end. (And of course for REW!).

Since I can't post links yet, I'll try to send you the mdat file via DM here. Hopefully I can add links there.

Thanks a ton for your help.

Dan

 

[John Mulcahy]

Nothing odd looking about the measurements, fairly typical variations from noise contributions. Using a longer sweep might reduce the variation a little.

 

[Sentient Mud]

Got it -- thanks.

Can you give me some sense of the frequency range at which RT60 is not meaningful?

Also, is the ~100 ms bump at 2 kHz in one of those traces likely just a noise artifact?

 

[John Mulcahy]

Yes, just a noise artefact at 2k. The room varies from modally dominant to a more diffuse sound field above the transition or Schroeder frequency. That frequency is approximately 2000*sqrt(RT60/V) Hz when room volume V is measured in cubic metres and reverberation time RT60 is measured in seconds, see room acoustics, for a very brief overview. RT60 estimation is based on an assumption of a diffuse sound field.

 

[Sentient Mud]

Ah, very helpful -- I now have a rough understanding of the of Schroeder frequency, etc.

FWIW, the Schroeder frequency of the room in question is about 180 Hz.

Thanks again John. Much appreciated.