REW Beta Release Time alignment offset abnormally high

[Silverprout]

Hello,
I've measured a loudspeaker and the tweeter impulse is 500µS faster than the medium impulse, could you help me please ?
The physical offset (face plate) of the tweeter is 21mm (closer) and the measured 14mm offset difference seems very credible to me.

The red impulse (medium)
Delay -0,56±0,01 ms (-0,193 m, -0,63 ft)
using IR peak time relative to Acoustic reference on Right (medium)

The blue impulse (tweeter)
Delay -0,52±0,01 ms (-0,179 m, -0,59 ft)
using IR peak time relative to Acoustic reference on Right (medium)

 

[John Mulcahy]

When you say 'medium' I guess you mean the mid-range driver? Is your concern that the 14mm system delay estimate is too far from the 21mm physical difference you measure? A couple of things can affect that. One is the positions of the acoustic centres of the tweeter and mid-range, which may be a little offset from the physical points from which you are measuring. Another is the differing bandwidths of the drivers. To look at alignment between the drivers in a speaker you may find the Peak Energy Time trace of the wavelet spectrogram of a measurement incorporating both units to be more useful, try using 1`/6 octave resolution.

 

[Silverprout]

When you say 'medium' I guess you mean the mid-range driver? Is your concern that the 14mm system delay estimate is too far from the 21mm physical difference you measure? A couple of things can affect that. One is the positions of the acoustic centres of the tweeter and mid-range, which may be a little offset from the physical points from which you are measuring. Another is the differing bandwidths of the drivers. To look at alignment between the drivers in a speaker you may find the Peak Energy Time trace of the wavelet spectrogram of a measurement incorporating both units to be more useful, try using 1`/6 octave resolution.

Thanks for your answer.
My microphone position should be more precise, few centimeters of microphone offest is becoming few milimeters of emissive point offset when the loudspeakers are at one meter.
The measurement was at 6000Hz, i've performed another measurement at 1000Hz and the mid-range driver is slighly faster than the tweeter, it seems to me that the mid-range cone inertia is delaying gradually the high frequencies.

The red impulse (mid-range)
The green impulse (tweeter)