Several failures occurred on an active production machine.
Several failures occurred on an active production machine.
Hello everyone, I’m working as a technician for a nightclub during weekends, handling live shows with several LED walls. About six months ago, our production PC began experiencing strange crashes. We mainly use two programs: DATATON WATCHOUT 6 and RESOLUME ARENA 6. All our video effects in Arena are .mov files encoded with Apple PhotoJPEG, while Watchout uses MP4 H.264. We tried switching to HAP and ProRes 422, but the crashes continue despite lower CPU usage compared to MP4. Most crashes happen when we open Watchout while Arena is already running, but they also occur when launching either program alone.
Our computer is a Toshiba built around 2016, equipped with an i7-7700k processor and 16GB RAM. I’ve replaced the RAM, yet nothing changed. The server PC functions normally; only the control unit crashes. Video from Arena is sent via NDI to both the production and server PC, which has always worked.
I’m unsure if the same issue always occurs, as I receive different error messages each time the system crashes. I’ll share all the minidumps I’ve saved recently; maybe someone can help interpret them. We’re planning a PC upgrade soon, but I want to understand the root cause to avoid repeating the problem. Thanks in advance for your help! Crash.rar
Random BSoDs often indicate unstable RAM. Or it used to. Nowadays...XMP might be the cause. I'll review the crashdumps and share my thoughts, though it could take some time to analyze them.
I've already swapped both DIMMs and ran MemTest86 several times, getting zero errors. I'm fairly confident XMP isn't the issue. Using the DIMMs from another machine in the office worked perfectly—no problems at all. It might be the slot itself. I also have another identical Toshiba built with an i5-7400, but I haven't swapped the motherboard yet. Thanks!
I’ll review the details step by step. The logs show a failure when trying to execute non-executable memory, pointing to a driver issue on the stack trace. The specific driver involved is identified, and its name appears on the debug screen. Several arguments and addresses are recorded, indicating where the crash originated. The system also reports a failure in a service routine, suggesting deeper problems beyond just memory access. It seems the audio stack drivers might be outdated or corrupted, possibly due to driver issues or hardware mismatches. Updating the drivers could help, but checking the BIOS/firmware version is also important. This situation highlights the need to verify both software and hardware compatibility.
A brief review of the other crash reports reveals largely the same findings as those I've shared. To add some spice, it's also conceivable a Windows 10 update might be the culprit, though this remains a guess unless all other possibilities are eliminated. I've encountered similar situations before within my agency.
Most drivers are current, some colleagues have already tested this. Various tools like CCleaner and DriverBooster are used. The audio interface is a Focusrite Scarlet 18i20, an older model. BIOS hasn’t been updated, yet the PC has functioned well for more than five years in this configuration. Changes haven’t been made recently, except when launching Watchout while Arena is running, and in our case Arena isn’t producing any audio. These two files are likely related to that issue. I’m fairly certain Windows updates were turned off on these machines to prevent forced reboots during performances. Would it be better to schedule updates for the week and then disable them again?
If no changes have been made and the system is offline, the issue likely stems from faulty hardware. Consider how challenging it is to reinstall or reconfigure the software; try booting with a spare system to observe performance. (Extracted from online sources) The file audiodg.exe manages the audio engine, handling all sound processing and digital signal processing. Manufacturers can add their own DSP and audio effects into the pipeline, which are then managed by audiodg.exe. portcls.sys - Port/Miniport Device system driver I don’t identify as an audiophile, so you might want to wait for a better option. If I grasp the hardware, it seems the audio card is external and connected to the PC, with all processing occurring there. If this aligns, the problem probably lies in the audio chipset—possibly outdated for the system’s age. Would a new PCIe audio card fit your budget?
The system is active yet updates remain turned off. Most components have been refreshed separately, such as GPU drivers and software. It's manageable but somewhat challenging, particularly with the upcoming PC purchase. We attempted to create a new user account, but it didn't replicate successfully—possibly due to chance, as the issue doesn't consistently occur even with the primary account. This involves a USB audio interface featuring several inputs and outputs. I'm uncertain whether this setup is standard for external interfaces or the internal machine. A PCIe audio card seems unsuitable, but a USB alternative might work if the existing one is defective. We require an external card to support multiple inputs, like main floor audio, microphones, etc., enabling individual tracks during editing and proper equalization. Multiple outputs are also essential, such as coordinating opening intros with lighting and video cues. In my Watchout timeline, main floor audio is assigned to channels 1 and 2, timecode signals sync the Compulite Vector console for lighting on channels 3 and 4. Previously, we used channels 5 through 8 for surround effects in the room. I also overlooked a detail: a few years ago, a fire damaged our AC ducts, leaving the building covered in ash. After professional cleaning and thorough testing, most equipment survived. While it seems unlikely to be connected to the incident, it's possible that everything functioned normally for nearly four years afterward—though we can't be certain.
When building a brand-new PC, it makes more sense to wait until the new one is ready rather than getting caught up in troubleshooting. Unpredictable crashes are the most challenging to identify. It seems the system is quietly deteriorating, so a fresh setup might be the best option.