I googled the error and found this thread. Thank you for posting it. There's a reddit post that links here, too.
I am experiencing this issue on a 2018 MacBook Pro (T2, Intel Core i7) running macOS 12.6 Monterey. The Mac needs to be running continuously for a LONG time before the issue starts to occur. My Mac's uptime is about 75 days right now. All applications using audio are affected when the default device is the external USB adapter. The issue goes away immediately if I switch to something else, like the internal speakers.
Examples of aberrant behavior:
Teams hangs for 30 seconds while starting a call and becomes unresponsive
Videos in Safari will freeze, then "catch up", then freeze, and repeat every 3-5 seconds.
Apple Music plays for about 3 seconds then goes silent.
I am using a CalDigit TS3+ docking station and the USB device in question is a simple headset/mic adapter branded UGREEN. I bought it from Amazon.
According to Audio MIDI Setup, my present audio configuration for the USB device is 16bit, 2 channel, 48kHz.
So far I have tried:
Killing the coreaudiod service
Quitting / relaunching the affected apps
Unplugging / reconnecting the USB device
Plugging the USB device directly into the Mac (bypassing the hub)
Changing the sample rate from 48kHz to 44kHz.
None of these actions improved the behavior.
I made an interesting discovery: every time I killed coreaudiod, my primary external monitor went blank for about two seconds. The primary monitor is a Dell 4K connected via the CalDigit TS3+ hub's DisplayPort output. My secondary external monitor, however, did not go blank when killing coreaudiod. That monitor is also DisplayPort, but connected via a Thunderbolt > DisplayPort adapter on the CalDigit TS3+. I can repeat this behavior with coreaudiod ad nauseam and have been gathering a sysdiagnose as I write this.
The USB audio device in question is an inexpensive UGREEN branded headset/mic adapter that I bought from Amazon. According to System Profiler, the vendor ID is Realtek Semiconductor (0x0bda) and the Product ID is 0x4809. What are your USB audio device's vendor and product IDs? Have we found any in common?
There's a slightly different yet similar error when recording or monitoring audio, even if you're just watching the Input tab in System Preferences:
2023-01-09 15:24:56.298594-0800 0x3edf369 Default 0x0 34738 0 coreaudiod: (CoreAudio) index: 49, start: 0x1768beb4151d25, duration: 0x5d6, fault address: 0x7fd07175c000, fault pc: 0x7ff80f17f28f, faulting TID: 0x3ee1bb9, fault type: 0x9, PID: 0x296ea4
2023-01-09 15:24:56.298625-0800 0x3edf369 Default 0x0 34738 0 coreaudiod: (CoreAudio) index: 50, start: 0x1768beb55a781e, duration: 0x178b, fault address: 0x7fd07173b000, fault pc: 0x7ff80f17f28f, faulting TID: 0x3ee1bb9, fault type: 0x9, PID: 0x296ea4
2023-01-09 15:24:56.298648-0800 0x3edf369 Default 0x0 34738 0 coreaudiod: (CoreAudio) index: 51, start: 0x1768beb55a9515, duration: 0x6d6, fault address: 0x7fd07174c000, fault pc: 0x7ff80f17f28f, faulting TID: 0x3ee1bb9, fault type: 0x9, PID: 0x296ea4
[...repeated several times...]
2023-01-09 15:24:56.300733-0800 0x3edf369 Default 0x0 34738 0 coreaudiod: (CoreAudio) Audio IO Overload thread: 3ee24c4 inputs: '<private>' outputs: '<private>' cause: 'Unknown' prewarming: no recovering: no
2023-01-09 15:24:56.301066-0800 0x3edf369 Default 0x0 34738 0 coreaudiod: (libAudioStatistics.dylib) [com.apple.coreaudio:carc] CAReportingClient.mm:508 message {
"HAL_client_IO_duration" = 100285;
HostApplicationDisplayID = "com.apple.systempreferences";
cause = PageFaultsOffIOThread;
deadline = 48268;
"input_device_source_list" = Unknown;
"input_device_transport_list" = USB;
"input_device_uid_list" = "AppleUSBAudioEngine:Generic:USB Audio:00:1";
"io_buffer_size" = 512;
"io_cycle" = 413;
"io_cycle_budget" = 5222463225846134107;
"io_page_faults" = 0;
"is_prewarming" = 0;
"is_recovering" = 0;
"issue_type" = overload;
lateness = 38718;
"other_active_clients" = "[ { HostApplicationDisplayID_other_client: com.apple.systempreferences, sample_rate_other_client: 0.000008, io_buffer_size_other_client: 512 } ]";
"other_page_faults" = 52;
"output_device_source_list" = "";
"output_device_transport_list" = "";
"output_device_uid_list" = "";
"safety_violation" = 0;
"sample_rate" = 48000;
"scheduler_latency" = 54208;
"smallest_buffer_frame_size": <decode: missing data>
Since this is my work laptop, I filed a case with AppleCare for Enterprise and included a sysdiagnose after the problem was occurring, and also another taken while reproducing the issue, so that the diagnostics will capture memory footprints, spindumps, etc., of the affected process.
I'm hoping we all get to the bottom of this soon.
What about Ventura? Has anyone with this issue seen an improvement there?