Wrong "vector average" calculations

[BenF]

Hi,

Sometimes i see really weird "vector average" calculations, like in the screenshot below:

How can the average be 60dB, when all 5 sources are between 66.7dB and 72.7dB? The math looks wrong...

Here are 5 similar measurements, where the average seems to be calculated correctly:

The average line seems to be between the source lines, as it should.

The problem is the same in v5.19 and v5.20 beta 55.

I have attached measurements for both of the above cases.

 

[srossi]

Vector averaging uses both magnitude and phase.

 

[BenF]

Vector averaging uses both magnitude and phase.

Thank you!
This can explain the phenomenon - but if this is the case, "vector average" doesn't really calculate the average frequency response of headphones, it's closer to summing up the vectors as if they are all active at the same time.
How can I create an average vector taking into account only magnitude?

In the 5 measurements the phase was 105, 106, -126, 26 and -179 degrees - how can the same headphone driver on the same rig produce such different results?

 

[srossi]

You can compute an arithmetic average in the All SPL window - there is a button on the lower left of the graph "Averages the Responses" - it will arithmetically average all responses that are checked.

 

[BenF]

You can compute an arithmetic average in the All SPL window - there is a button on the lower left of the graph "Averages the Responses" - it will arithmetically average all responses that are checked.

Thank you so much! Don't know how I missed it before - I guess I expected it to be in the "Controls" menu. Looks like in v5.20 it is in the "Controls" menu, under the name "RMS Average".

I'm still baffled by the drastically different phase measurements - can anyone explain this?

 

[srossi]

It's difficult to speculate without knowing the difference physically or electrically between the two sets of measurements

 

[BenF]

It's difficult to speculate without knowing the difference physically or electrically between the two sets of measurements

I meant the differences between the measurements in the same set. 5 measurements taken within minutes - same headphone driver, on the same side of EARS rig.