REW messurements in churchacoustics

[BenneZappe]

Hey folks!
I am currently studing Acoustical Engineering and as part of an project we have to messure different churches to make a database for our area. Our goal is to get data in form of RT30/RT20, EDT, C50 and C80. We´ve already messured a smaller one (Volume around 3200 cubicmeter) and had no problems using the system and REW. Our problem occured with larger volumes like 20000 and above. The Photos you i link show it. Everytime we got a dip at like 200-400hz. Sometimes on more frequences. We are currently thinking of an volumen issue. So our tech is not capable of getting enough loudness to stimulate the space. If you need the exact data of the tech wew using i can link it afterwards.
Thank you in advance!

 

[staticV3]

Hi @BenneZappe! Welcome to ASR.

I'm sure @JohnPM can help you with your issue

If you could save your measurements as .mdat, Zip them up and upload them here, that could be useful.

 

[BenneZappe]

Hey! Also the messurements from the smaller church? I´ll do it as soon as possible!

 

[staticV3]

Only the problematic ones should be fine.

 

[BenneZappe]

I had to make a wetransfer link, because its to large for this site it says.
Here you can download: https://we.tl/t-VZrMqbpX7i
(Link active for 7 days)

 

[DaveBoswell]

Very interesting. One thought is given the very long decay time, you may want to change the sweep length as the calcs for RT60 using the different techniques may run into errors if decay times are extremely long (and longer than the sweep duration). And alos may want to try raising the test volume to get as far above the noise floor as possible.

In the Measurement dialog, change the Settings for length to something like 1M or 2M (or higher) instead of the default 256k.

 

[BenneZappe]

Very interesting. One thought is given the very long decay time, you may want to change the sweep length as the calcs for RT60 using the different techniques may run into errors if decay times are extremely long (and longer than the sweep duration). And alos may want to try raising the test volume to get as far above the noise floor as possible.

In the Measurement dialog, change the Settings for length to something like 1M or 2M (or higher) instead of the default 256k.

We also had this thought and changed it up to 1M. Also the numbers of sweeps up to 8. Same results.

Your first idea unfortunately we cannot do due to the fact, that we cannot get the volume of our system louder then we already did. Noise floor was by 53 db and the testsignal was 103 db. If we wanted to boost the signal even higher we got a distorted output.

You think it has something to do with long decaytimes?

 

[olieb]

Your impulse responses are seem truncated. That is REW default. In preference you can change that. Maybe that helps.

 

[BenneZappe]

Your impulse responses are seem truncated. That is REW default. In preference you can change that. Maybe that helps.

Okay we’ll try this out next week. Maby i can burrow the measurementsystem before we make the next take.

 

[DaveBoswell]

Yup - this is it. You need impulse response long enough to see decay from -5dB to -35dB for T30. The 4.4s is (probably) good enough?