Sweep length leading to differences in measurement

[tof_sten]

Hello,

I run REW v5.20 RC9 on an Apple MacBook Pro with M1 processor.
I output sound from an Audient ID44 and my mic is a UMIK-1.

While doing measurement of speakers in a room, I found out that results vary with different value for the sweep length. Moreover all the impulse shows strange oscillations.

Same measurement with different values for the sweep length:

sweep length = 128K

sweep length = 256K

sweep length = 512K

sweep length = 4M

I think it is strange, any advice?

 

[John Mulcahy]

It indicates the Audient and the UMIK have slightly different clock rates, which causes dispersion in the IR. The longer the sweep, the greater the dispersion. The UMIK's clock comes from the USB interface. If you measure with an acoustic timing reference REW can correct the clock rate difference if you select "Adjust clock with acoustic ref" in the Analysis preferences. After you make a measurement any clock rate adjustment will be shown in the Info panel for the measurement.

 

[tof_sten]

Thank you for this answer John.
I now have more questions:
1/ I checked "Adjust clock with acoustic ref" is it activated for future measurement only? Can I apply it to previous measurement?
2/ will "Adjust clock with acoustic ref" lead to good enough results or would it be better to to buy a xlr measurement mic connected to the preamp of my Audient interface?
3/ you said UMIK1 clock come from the USB interface, does it mean that I would get better results buy using UMIK + MacBook Pro headphone output?

 

[John Mulcahy]

There is an option for manual clock adjustment on existing measurements in the Impulse graph controls, you could apply a figure matching what you see in the Info panel when using the automatic clock adjustment.
Automatic clock adjustment should work well, no need for another mic.
Doesn't hurt to try the headphone output.

A clock rate difference affects the IR appearance and the phase response, but doesn't affect the SPL response.

 

[tof_sten]

Thank you again!
It looks like it affects the Spectrogram view too. Am I right?

 

[John Mulcahy]

Yes, it will have some effect on spectrograms, waterfalls and decays.

 

[tof_sten]

Thank you John!