Glad to hear they are making improvements. Hoping these improvements trickle down to the VX5 and VX7 in the firmware updates. A lot of potential if they use these updates to learn from previous issues.
We are still waiting to see Minelab or Garrett take a leap into firmware that learns from users and environments. Given that there is always an end-user, this can help the machine learn faster. Recursive learning and retention of important signal data would allow metal detectors to become smarter over time, adapting to different terrains and filtering out noise more effectively (for example, my park soil in Washington may be different than your park soil in Florida). From my understanding, VX9, Manticore and XP rely on fixed set of algorithms. Imagine an approach that could personalize detection based on a user’s habits, locations and past targets. That would be a a game changer!