GD and Excess Group Delay

[Sonder]

Hi I am trying to understand what the Excess Group Delay EGD is telling me, I wonder if could you tell me if this is correct.

If there is a large spike vertical in the group delay window. It is telling me that there is a reflection of X amount milliseconds, if this corresponds to a null in the FFT graph is it because there is destructive interference and it is cancelling out?

But if there is a null in the FFT graph and there is no spike in the group delay, then what is this null caused by room mode pressure, or something else?

Seems there are many spikes in the excess group delay the further up the frequency I go so I'm presuming that these are reflections in an untreated room.

But there are some in the lower range too, so this is where I am confused, what is causing the differences?

Thanks

 

[John Mulcahy]

There's an explanation in the help. When sound from multiple sources or multiple paths is summed at a point there can be departures from minimum phase at some frequencies, which gives rise to excess phase.

 

[Sonder]

There's an explanation in the help. When sound from multiple sources or multiple paths is summed at a point there can be departures from minimum phase at some frequencies, which gives rise to excess phase.

Thanks, so I read all this, so room modes can also cause fluctuations in the excess group delay, creating a non-minimum phase. So how can I tell when its a room mode or a destructive summed reflection?

 

[John Mulcahy]

Modes are reflections along closed paths, their peaks and troughs are also the result of summed reflections.

 

[Sonder]

Modes are reflections along closed paths, their peaks and troughs are also the result of summed reflections.

I see, thanks, it's difficult to say which is which from looking at the GD and FFT graphs, so I am presuming all of the delays in the higher frequency range in the GD graph from reflections.