UMik-2 cal file - not a recognised USB mic

[dlr]

I load the UMik-2 cal file using the measurement dialog "Calibration files..." that brings up Calibration data window. After entering the file the window pops up a message "The cal file has a Sens Factor entry but the input is not a recognized USB mic. The Sens Factor will not be used. " The file is a raw txt file with serial number for filename, but nothing to indicate the make of the mic. Why is it not accepted?

"Sens Factor =-10.77dB, AGain =18dB, SERNO: 8111506"

 

[John Mulcahy]

The UMIK has not been selected as the input, as far as REW can see. That might be because the input device and input have been left as default, or it might be because the input is oddly named (can happen on Linux, for example). What input device and input are shown on the soundcard preferences?

 

[dlr]

I'm using FlexASIO. I just saw that it had "Microphone" for input device, I changed that to "Line (Umik-2)". The Windows setting had been and is "Line (Umik-2)". I set FlexASIO Input to Umik-2. It seems to load now, no message popup. Just ran a new measurement. The db level _seems_ to be what I would expect, but I have no way to verify calibration.

Is there not some way to use the sound card ASIO driver, yet use the microphone direct input without a wrapper such as FlexASIO?

 

[John Mulcahy]

Is there not some way to use the sound card ASIO driver, yet use the microphone direct input without a wrapper such as FlexASIO?

No, because ASIO is structured as a single device protocol. Wrappers like FlexASIO create a pseudo ASIO device that under the hood is separate devices for input and output. For measurement purposes there's no benefit to ASIO over the WASAPI Exclusive Java drivers (device names starting EXCL).

 

[dlr]

I switched to the Java drivers. Works fine that way. Better in fact. I've been having occasional glitches during STHD and multi-measurement averages. Possibly due to the old laptop with FlexASIO I've been using. So far the Java driver has not exhibited that.

 

[phofman]

I've been having occasional glitches during STHD and multi-measurement averages. Possibly due to the old laptop with FlexASIO I've been using. So far the Java driver has not exhibited that.

It could be caused by FlexASIO merging the two devices with independent clocks. As John said that ASIO device merge is a hack for which ASIO was not designed. The java driver has separate paths for capture and playback devices and individual clocking of each does not matter. It's the correct way to go.

 

[dlr]

Back again. I'm using my relatively new Windows 11 laptop now. FlexASIO worked fine with the Umik-2, but I realized that the same issue occurred, not recognized as USB mic. So I tried switching to Java in REW again. The Umik-2 is not even seen by the Java driver. The ASIO works without issue through FlexASIO, but I need the senstivity. When I started this thread I was using an old laptop from 2012 Windows 10. I had planned to dedicate it for that, but it's just too slow. The problem appears to be Windows 11 somehow. The Device manager sees it as Umik-2 with miniDSP as manufacturer, device type "Sound, video and game controllers". But in FlexASIO it's seen as Line (Umik-2).

(Later): I spent a lot of time trying to trace this down. The last thing I did was open the miniDSP UACS Control Panel. It has little control actually. It was set for 2 channel, 24 bits. I changed it to 32 bits not expecting any change. However, REW then found the Umik-2 in Java, accepted the calibration file and is now working with the senstivity set. Strange. The control doesn't even show the real serial number, just 000000. But it's working now.