When human detection is enabled and snapshots are uploaded via FTP, the images are only stored at 640 × 360, even though the camera sensor and main video stream support a much higher resolution.

I also noticed that the same low resolution is used even when using basic motion detection instead of human detection.

This makes the feature much less useful, especially for security purposes, because the uploaded image often lacks enough detail to identify a person or properly understand the event.

It would be very helpful if the camera could save FTP snapshots at the main stream resolution, or at least provide a setting to choose the snapshot resolution.

Is this a firmware limitation, a configuration issue, or a design choice?

If I have no solution I will send back this hardware as it is of no use to me.