Discrepancy between REW measurements and SPL measurements

[audiothings]

Writing in for a friend...

Four loudspeakers of identical make and model were tested under identical circumstances... The disparities in their response was only to be expected (+/- 4 dB or so < 8 KHz?), given that they are not high end drivers meant for any audiophile purpose... Please see REW graph below...

However the SPL readings when running program material and pink or white noise, show a 12 dB discrepancy between the drivers.

How can the over all output show a 12 dB discrepancy when the frequency response seems to show a +/- 4 dB discrepancy?

 

[jan.didden]

A question: when measuring freq response, did you equalize the drive levels to get the curves to overlap?

The other thing is the location. Did you move each one into the same position for the SPL measurements, not changing the mic position, while in the next test they were in a different location?

Jan

 

[markbakk]

Likely the conditions of your measurements aren’t controlled well. Same amp output? Same measuring position as Jan already asked? Same preamp setting? And so on…

 

[audiothings]

when measuring freq response, did you equalize the drive levels to get the curves to overlap?

identical gain on input and output and no processing

Did you move each one into the same position for the SPL measurements, not changing the mic position, while in the next test they were in a different location?

they were side by side. Mic was moved, but distances were maintained

 

[jan.didden]

That could give some changes, but 12 dB is a lot. Factor of four.
I can only suggest to carefully look again at the test setup, things like ref levels in REW between the various conditions.

Jan