AC2020bsi: FITS DATE-OBS is sometimes, START, or END, but never(?) MID time of exposure

On first image after cold start, image FITS header DATE-OBS corresponds to the start time of the image. Subsequent images show a DATE-OBS value corresponding to the end time of the image. All are wrong: documentation says it should be mid-time of image.

Details:

Freshly booted Win11 computer, fresh start Mxm DL7 (7.0.3, DLAPI 2.8.0.0). Camera USB cable was already connected. Powerup camera, wait 5 sec after fan starts, start Camera Cntl wondow. Wait 5 sec after shutter click, go to Expose tab. Tab shows Readout Mode HDR (from previous run, before restart).

Take a 20 sec image, dark, bin 2. Watching sweep second hand on iPad, start image at time hh:38:00 hh:mm:ss. Save this is as image 12. Image FITS header has DATE-OBS hh:38:00, i.e. the START TIME of the exposure.

Without any reconfig, or rebooting, or cycling of power etc etc., take another image with same parameters. Start image at hh:41:30. Save this as image 13. Image FITS header has DATE-OBS hh:41:51. i.e. END TIME of exposure.

Repeat: start image at hh:50:30. Save this as image 14. Image FITS header has DATE-OBS hh:50:51, i.e. END TIME of exposure.

Note this is a variant / refinement of the report I filed under MaxIm DL, Oct 12, 2023. I was running DL6 at that time.

These tests were done using HDR readout mode and dark frames. Problem was first seen using High Gain and High Gain StackPro modes on light frames.

Gary Billings