Wildly varying REQ results

[tensor9]

Here is an overlay of 6 sweeps took right after one another. I have an UMIK-1 mic.
Also, as an aside, MacOS keeps "forgetting" my USB dac. I have to click on another output source, then the DAC again and it starts working.

 

[John Mulcahy]

Can you attach the mdat file?

 

[tensor9]

Here you go

 

[John Mulcahy]

You are measuring at 44.1 kHz but the UMIK-1 is a 48k-only device. Best make sure Audio Midi Setup is configured to use a 48 kHz format for the UMIK and also configured to use 48 kHz for the output device and set REW to 48 kHz also.

 

[tensor9]

Thanks, but I actually changed that to see if it would help. Same thing was happening at 48

 

[tensor9]

Ok, here are the results with both the mic and output (Audio Midi) set to 48 k.

 

[sam_adams]

First off you need to update to the latest RC. Second those measurements are showing wildly varying acoustic timimg differences and huge amounts of 2nd order distortion.

What external DAC are you using? Do you have any DAW software running while trying to use REW?

 

[tensor9]

I’m using the USB input of a Matrix Element X with an Apollon Audio NC1200 (dual hypex NC1200). Magnepan 3.7i.
No DAW software or anything else should be in the signal path.

 

[John Mulcahy]

Pretty bizarre results. There are sometimes 2, sometimes 3 distorted copies of the IR. Does the sweep sound normal? A smoothly increasing-frequency tone? There doesn't appear to be any output device/output selection in the REW preferences for those so the system default will be used, but shouldn't particularly matter. There may be dropouts in either the sweep (which would be audible) or the capture (which wouldn't, but that can be checked for by looking at the "Captured" graph immediately after making a measurement).

 

[tensor9]

The sweep sounds pretty typical. The timing pulse does sound a little funny at times.
My DAC has an airplay input as well, and when that is used there is alll sorts of dropout. USB sounds on though.
Is there any continuous all frequencies at the same time test I could do?

 

[John Mulcahy]

Sure, you can play the full range Pink PN test signal and see the result on the RTA.

 

[sam_adams]

This is what John is referring to with regards to the impulse response:

It almost appears as though the sweep is being sent through the Element X multiple times.

Do you have any software like Soundflower, LineIn, Loopback or any other type of sound routing/mixing software installed?