We've been troubleshooting rampant issues on E5490, E7400, and E5400 with Realtek Audio. All of the common steps have been taken from rolling back, ripping out to test WHQL, etc. All of these users have Zoom/Skype. They seem to call when after attempting to join a call suddenly the onboard microphone array + headsets are not responding. A quick flip of the driver or device seems to bring it back again. It's not clear yet whether they're on a Skype call then transition to Zoom. Or, on a Zoom call, then possibly join one from a web-based plug-in while the Zoom Desktop is installed. All I know is that Zoom is taking a long time to respond to tickets.
We're trying to pursue this as an escalation using the MDS (Multiple Down Systems) form w/ Dell.
Event Log:
We'll update as the issue is investigated further. Please post any information you have about this problem here for others to correlate with.
We see a crash from Skype on one device in the Event Logs referencing 'DllHost.exe' and the C:\WINDOWS\System32\combase.dll file. The 'Zoom.exe' crash references this same file as well on the device in the event logs.
We are experiencing the same issue with Dell aio Optiplex 5270s.
Once you select to Join with computer audio, zoom crashes along with the windows audio service stops.
The issue seems to have started with the Realtek Audio driver 6.0.8940.1 - If you are able to downgrade the driver to the previous version, it seems to fix it but that is a manual process which is not viable for the 200 workstations across multiple sites that I am experiencing this issue with.
Have you had any joy with this issue?
I've tried 3x Dell Drivers and 2x WHQL drivers by removing all Realtek software and letting Windows take over managing the device. I was able to confirm it continued to happen on these versions. Users are on the latest Zoom using auto-update. The trouble seems to be it's hard to reproduce so we never know if it's fixed. You'll think that a new driver does the trick until it triggers again. Keep an eye out for users reporting the issue on that downgraded driver. I recommend opening a case with Dell so they can escalate to the Advanced Resolution team. You'll be asked to fill out an MDS (Multiple Systems Down) forms to provide details on the machines.
My next thought is that rolling back Zoom from 5.1.x to 5.0.x might help but I can't risk using any more users as canaries. We're going to swap out some of these new laptops with competitors' devices as we can see this is going to be a long crawl through support territory with Dell.
Thank you so much for your detailed troubleshooting and making this issue known.
We've had inexplicable issues with our Dell Latitude 7400 laptops and the folks who heavily use Zoom (execs). This exact same thing is happening and we've been at our wits end. We've tried literally everything. I have tickets in with Dell, Zoom, our MSP, and I just don't know what to do anymore!
Katie
@fixfinder you're right about it being hard to reproduce! I was able to fix it by simply disabling and enabling the realtek microphone array for one person, and the other person's mic actually worked when the mic was DISABLED! No idea why. These are probably very unstable, nonpermanent fixes, so also ordered external mics for those users. The other person's doesn't work no matter what I do!
I have a ticket in with Zoom that just got escalated to their engineering team so will report back there.
The problem is that once the issue occurs, the mic won't work for other programs either. How frustrating!
@kniemann, indeed! I've got users who have the onboard Realtek array go completely dead. Nothing comes through the microphone in the 'Sound Settings' until I reinstall the driver, or delete the audio device from compmgmt.msc. The folks with headsets seem to have the microphone go dead at the same as the onboard. I've been messing around with removing everything but the Realtek Audio driver like USB/Components and other tools associated w/ Realtek. I've got a machine right now E5490 on 6.0.8895.1, A23 that's having the problem repeatedly. This is the version that support.dell.com is offering as the latest for this model. It's odd because I have a ton of other people on these combinations of drivers/laptops. At this point, I'm convinced it's intermittent across many versions. The hard part is that all of the testing is happening with real people who need to get work done on these calls.
We've sent in an escalation w/ Zoom + Dell Advanced Resolution. Trapping these AUDIODG.EXE crashes with the RMM tools to proactively identify users with issues. The next best thing might be trying to capture the event with Procmon in an effort to further help the vendors isolate where this interaction is happening.
Guess what? After tickets open with everyone, Zoom engineering team provided this response - Microsoft is releasing a hotfix for this problem at the end of Sept! Big relief! Unfortunately the suggestion of rolling back drivers did not help all the users, but we have alternatives until this hotfix is released. Wanted to let everyone know. Thanks, @fixfinder for helping us all work together!
- Katie
Zoom's response:
Hi Katie,
Apologies for the inconvenience. We do have existing tickets having the same issue with Dell machines, the engineering team confirmed that it is an audio driver issue with Realtek. Our engineering team already notified Dell about this issue and in addition, Dell also reached out to Microsoft informing the said fault. Microsoft advised that they will release a hotfix for this issue around end of September.
While waiting for the hotfix, you may try to roll-back to the old driver or uninstall a Windows driver update dated June 11th, Thursday. This information came from a similar post from this link.
Let us know if you need further assistance.
Best,
Ren
I have the same issue on the OptiPlex 5270. My workaround is to switch to the Realtek driver or "High Definition Audio Device" made by Microsoft:
devcon update c:\windows\inf\hdaudio.inf "HDAUDIO\FUNC_01&VEN_10EC&DEV_0256&SUBSYS_10280936"
Strangely, the OptiPlex 5250 with the same driver doesn't have this issue.