Problems measuring Phase with UR22MKII

[SPo]

Hi John,

for measuring my speakers with time alignment i ordered a ECM8000 and i am using it on my Steinberg UR22MKII.
I made the following setup:
1. no loopbak in Steinberg`s ASIO driver with 48kHz, i switched phantom voltag ON
2. i plugged MIC into INPUT 1 and my amp in OUT 1
3. i made a loopback from OUT 2 to IN 2 (via cabling)
4. i think after leveling the input and outputs everything should have enough signal (no warnings for clipping or too silent signals)
5. I made a calibration and the diagram fits, i guess
6. i measured my speakers and i got an awful phase measurement

So i think, i am having a problem in my setup. Do you have any idea whats wrong ? I am using your newest Beta...

 

[John Mulcahy]

Please post the mdat file, it is a lot easier to interpret than some pictures.

 

[SPo]

Here are my two measurements with the same setup.

I already tried to deactivate USB sleep modus, but i get sometimes the jump in the phase anyways.

edit#1: and i tried java soundcard driver, too - no change

 

[SPo]

... and the calibration measurement

 

[John Mulcahy]

The measurements with a loopback include a time delay corresponding to the distance of the speaker from the mic, 2.635 ms in those measurements. To see the phase you should first remove that delay using the controls on the Impulse response graph (or enter it on the Measure dialog as a Timing offset for subsequent measurements).

Since your measurements are in-room, there are reflections in them. Those reflections can cause nulls deep enough to produce phase wraps (non minimum phase behaviour), that can vary from measurement to measurement.

Your system appears to be inverting, by the way.

 

[SPo]

Hi John,

thank you - adding the delay and putting some basotect panels around and in front below of the mic did the trick with jumping and the huge phase-span.

This problem is solved!

About the inversion: i thought that wont be a problem for measurements of only one chassis ? But how can you check this with REW ?the captured data wont tell me much.

SPo

 

[SPo]

ahh - never mind... look at the graph you added in the post...

The error came due to my 'fast' testing of the new equippment.