Luxo
Thank you for the detailed and thoughtful feedback.
As someone on the internal team, I can confirm that the core logic for GOTO centering correction (i.e., plate-solving-based refinement) has not changed between v2.3.3 and v2.4.4. From a code perspective, there should be no fundamental difference in behavior—especially not the significant degradation in pointing accuracy or meridian flip recovery you’ve described.
That said, we fully trust that you are observing this issue in your specific setup. When logic remains unchanged but behavior diverges, the cause is often subtle—such as interactions with particular hardware, firmware states, calibration artifacts, or environmental context.
To help us pinpoint the root cause quickly, could you please run a controlled side-by-side test?
Keep everything identical: same mount, camera, location, target, and time window.
First, test with v2.3.3 using your two scenarios (GOTO accuracy and post-meridian-flip positioning), then export the log (long-press the “ASIAIR” icon on the home screen).
Next, upgrade to v2.4.4, repeat the exact same steps under the same conditions, and export the second log.
Send both logs to us at luffy.liu@zwoastro.com.
By comparing these two logs from identical real-world conditions, we can isolate exactly where the discrepancy arises—this is by far the fastest path to a fix.
We truly value feedback from experienced users like you. Your help in reproducing this will make a real difference.
Best regards,
