Phase wrapping with loopback

[Omholt]

Why does phase wrapping occur when loopback is used for measurements?

The phase here was linear when measuring without loopback. Using a USB sound card with internal loopbac function.

 

[dcibel]

If you're talking about using a loopback cable as a timing reference, what you're observing is excess phase due to time of flight for audio to reach the microphone, which could be several milliseconds. Have a look at the impulse, you should notice is no longer resides at t=0.

There's a timing offset option when measuring which you can use to move the impulse closer to T=0 which will make the phase response easier to view, and still maintain relative timing relationships between multiple measurements.

 

[Omholt]

Yes, I'm talking about using loopback as a timing reference. Why is the excess phase shown when I haven't generated what's minimum phase?

 

[dcibel]

You measured excess phase with your loopback timing reference. It takes time, as in milliseconds, between loopback impulse and actual measured impulse. This is reflected in the phase response.

In the impulse response window, you can select "estimate IR delay", which will calculate delay between t=0 and the impulse. You may then select "shift and update timing offset" so all subsequent measurements include the same delay. impulse will appear near t=0 and subsequent measurement will retain relative timing differences.