Small repeating oscillations on frequency response

Hi

Trying to debug this issue, not sure where it stems from (yet).

If you look at the two pictures you can see that the measurements has reoccurring oscillations, with a frequency separation of 2x/0.5x. It is easily identifiable in the impulse response, if these errors occur. Most, but not all measurements, will have these oscillations on the current setup. Note that their placement frequency wise, is not persistent, they move around for each measurement done.

I will try debugging this issue by switching measurement setup (entirely different OS and hardware, etc.) and go from there, I am just wondering if others have seen something similar, so that I can speed up the process.

Cant hear the deviations during the sweep, but I guess they are fairly minor, so I might not pick it up.

REW mdat dataset with multiple examples attached.

Found 'audmes' via Flathub, will try to do some simple measurements with that software tomorrow and compare.

John Mulcahy said:

They look like dropouts in either replay or capture to me. Why are you using a 2-year old REW version, by the way?

Click to expand...

Hi John

Will investigate dropouts.

I have several installs of REW on different systems for convenience, I think that one is from my laptop (I am just viewing the measurements there).
Edit: I am running 5.20.9 on main system, I think I downgraded to use the old GUI framework at some point, not sure. Will update them all and test again.

phofman said:

@Quickmcj : Dropouts are very easy to check in RTA with averaging. Every dropout ruins the averaged chart.

Click to expand...

This was done super quick, so might be error prone, but I do not see dropouts in the averaged plot from RTA (doing sweep loop from around 1kHz to 3kHz). I can only see the problem with a regular single sweep measurement and if I gate strongly to remove the room. Both plots in attached mdat file.