Strange phase behavior

[Johan b-lund]

Hi All,

I have recently seen a strange behavior that i dont have seen before, not that i can remember anyway.
If i measure LFE the phase looks really strange but if i measure with a acoustic timing Ref it falls in to place.
I have tried on Win 7 and Win 10 - same behavior and the exact same phase but on Win 10 the timing is all over the place.

It started that i where helping a friend and his measurements looked the same.
Am i stupid and have missed something?

mdat file from win 7 REW 5.19
Includes:
LFE T = Timing Ref
Dirac Small + sub T = Timing Ref
LFE No T = No timing

 

[John Mulcahy]

The appearance of the phase plot depends on where time zero (t=0) is defined on the impulse response. When using an acoustic timing reference the arrival of the timing reference signifies the position of t=0. When not using a timing reference it can be placed at the peak of the impulse response (which is the case for the measurement you attached) or the start of the impulse response depending on the setting in the Analysis preferences. The start of the impulse response would give a similar phase plot to the acoustic reference for your plot since the response is starting pretty much where the timing signal arrives.

 

[Johan b-lund]

The appearance of the phase plot depends on where time zero (t=0) is defined on the impulse response. When using an acoustic timing reference the arrival of the timing reference signifies the position of t=0. When not using a timing reference it can be placed at the peak of the impulse response (which is the case for the measurement you attached) or the start of the impulse response depending on the setting in the Analysis preferences. The start of the impulse response would give a similar phase plot to the acoustic reference for your plot since the response is starting pretty much where the timing signal arrives.

Hi John and thank you for the replay!

So i should not check the "Set t=0 at IR Peak"
I unchecked it and tried again, the measurement is called "unchecked"?

Thank you for the help!

 

[John Mulcahy]

On that version an estimate of IR delay is used if not placing t=0 at the peak. Probably better updating to the latest V5.20 beta.

 

[Johan b-lund]

Will do