Discrepancy between calibrated UMM-6 and BAFX 3608 readings

[austinrafael]

Hi folks,
I am having a fundamental disconnect about SPL readings. I have seen similar posts here but nothing definitive - anyway, enough babbling:

Pioneer SC-91 connected via HDMI (only using one output connected to MiniDSP 2x4 HD)
REW 5.20 beta 60 with SPL meter up
Dayton UMM-6 with calibration file loaded
BAFX 3608 SPL meter on C/slow
(mains disconnected for sub SPL tests)

When trying to measure basic SPL, the dayton mic is picking up anywhere between 5 and 10dB lower than the BAFX. Could this be because REW is calibrated for lower frequency pickup, and the BAFX is not? That doesn't feel right. Or just that the Dayton mic is "more correct" due to it's cal file?

I am placing the Dayton/BAFX side by side, pointing upwards.

Many thanks,
Raf

 

[WaynePflughaupt]

What calibration file are you using for the UMM-6? Don’t how things are currently but their calibration files were known in years past to be worthless.

In order for the UMM to deliver an accurate SPL reading, the calibration file needs a sensitivity figure as the first entry. Is there one?

Regards,
Wayne A. Pflughaupt

 

[austinrafael]

What calibration file are you using for the UMM-6? Don’t how things are currently but their calibration files were known in years past to be worthless.

In order for the UMM to deliver an accurate SPL reading, the calibration file needs a sensitivity figure as the first entry. Is there one?

Regards,
Wayne A. Pflughaupt

Indeed there is a first line, it stars with "Sens Factor =-25.025dB"

 

[austinrafael]

Could be related to a bad sound card calibration, specifically loopback readings? I had a really hard time getting the "In" level to read close to the "Out" level in Preferences->Soundcard.

One other note is that although I am using the Java driver, I did grab the ASIO drivers if that's preferable.

I appreciate the time- I am probably missing several key noob things, so please send me beginner articles.

Thanks again!
Rafael

 

[WaynePflughaupt]

The BAFX meter is cheap, but even meters like that are typically +/- 2-3 dB at worse, so I’m thinking there’s an issue with your REW signal chain. Can’t tell you what or where though. Really, the only way to verify for sure would be with a third device.

Regards,
Wayne A. Pflughaupt

 

[austinrafael]

Ok, bit more info - I hooked up the mic and installed REW on a second laptop - and the readings were the same between the two computers using REW and the same mic. In fact, when sending test tones for L or R from the receiver, the readings between both laptops and the BAFX were within around 3dB for L/R/C

The dramatic change comes when I send the tone from the subwoofer channel. There, the two laptops match, but the BAFX reads over 10db higher. Bizarro.

I have confirmed that both the BAFX and both REW laptops are set to C weighted and oriented the same way, within two inches of each other. Could it be I am overlooking something in how REW is configured to measure the lower signal as compared to the "out of the box" BAFX? Or should it be sound pressure across the board and the issue is somewhere else?

 

[austinrafael]

Well, I installed Windows on my older laptop (was using REW on linux a few weeks ago) and retested the mic/sub level - and it was more consistent with the BAFX - so it looks like the problem followed the combination of subwoofer tone+new Windows laptop. It was def using the dayton mic but regardless, it looks like you are right

Thanks all,
Rafael