Grahame wrote:Helicon and Zerene have trials give one of them a spin, if your machine doesn't crash probably not your hardware.
Giving these stacking applications a trial is a fine idea, if the intent is to find a stacking application that best matches one's subjects, workflow, visual standards, and budget. But using these applications to test hardware stability is wildly inefficient. Prime95 is free, takes less than five minutes to setup and initiate a stress-test under, and is a well-understood standard for computer stress-testing among thousands of system administrators world-wide. It produces a lot of heat--quickly--and while the system is subject to this heat, instructs the CPU to perform fault-intolerant calculations. It keeps a record of any computational errors produced. This record is highly predictive for errors produced by computers under heavy load--many of which can produce crashes. This record is often more useful than attempting to induce crash events, which tend to occur far less often than calculation errors; this said, crashes are still useful data points, and in my experience, nothing precipitates them faster than Prime95, if hardware is at fault.
Granted, we could conceivably press focus-stacking software into service for computer stress-testing. But doing so--especially across multiple stacking applications--creates a vast array of things to be learned. Meanwhile, well-understood protocols already exist for computer stress testing, one of which is Prime95's stress test. Why re-invent the wheel?