Topic: Fireface 802 bug (very low input levels) after switching to DK?
Hi,
tl;dr: After updating fireface to most recent driverkit I find input levels in totalmix and DAW to be suspiciously low. Switching back-and-forth samplerate fixes this and levels are "normal" afterwards. I believe this indicates a bug in the driverkit software.
I'm using a fireface 802 on a macbook air m2 with OSX Sequoia 15.01.
Recently I switched from the kernel extension to the most recent driverkit. I'm using the 4.16 driver and v20 firmware.
During the update i encountered the below error in terminal:
sudo /usr/bin/kmutil rebuild
Password:
Checking the auxiliary kernel collection...
Error Domain=KMErrorDomain Code=31 "Error occurred while building a collection: Cannot build collection without binaries as only 1 codeless kexts provided" UserInfo={NSLocalizedDescription=Error occurred while building a collection: Cannot build collection without binaries as only 1 codeless kexts provided}
And after completing the driver update and reboot, the 'concurrent kernel extension detected' warning, which i seem to have fixed by disabling and re-enabling SIP in safe mode.
While testing post-update, I noticed that input levels coming into my DAW are suspiciously low in comparison to before the driver update.
Example; signal at the front mic/line input is coming in almost clipping, but is hardly reaching a peak level of -36 DB in DAW. Prior to the update I believe to have seen a much more healthy level (let's say 'the full dynamic range') in my DAW session on the same signal. I quadruple-checked total mix settings, input gains etc., but could not find anything strange. Also a total reset on totalmix did not help.
After a lot of fiddling around, it turns out that after switching back-and-forth the sample rate (from 44.1k tot 48k and back to 44.1k) immediately solves the issue and my levels are "back to normal" again.
So, I believe i've stumbled upon a bug. If you require anything from me to solve this, please let me know.
Regards,
Bas
Addition: It seems that after switching sample-rates as described above, this 'fix' persists between reboots as well. I've tested this once (1 reboot), so it may be a shallow observation.