For guiding, I'm using the cheapest IMX462 module I could find with the indi-libcamera driver, and Ekos's built-in guiding. My camera module has a strip about 200px wide on the right edge that behaves strangely -- it's usually brighter than the rest of the image, and in the boundary between the two areas, one of three things happens:

1. there's just a discontinuity in brightness
2. there's a black line between the two areas
3. some pixels along the edge are significantly brighter than either area

The autoguider with SEP-Multistar occasionally gets confused and detects pixels along the boundary as stars, occasionally locking onto them. About 70% of the time (like in the screenshot above) when guiding starts, I get lucky and it locks onto real stars from the rest of the frame, and will work until guiding restarts due to slewing, etc., at which point it's another roll of the dice.

I can work around it by manually selecting a single star, but that won't work for imaging sessions where the mount needs to slew (multiple targets, meridian flips).

I'd like to configure the guide algorithm to only consider a subframe, such as X: 0-1720px, Y: 0-1080px. This would crop out the bad part of the sensor. I've attempted to set the subfield in the indi control panel, which may work for one frame but gets reset the next time the guider takes an image.

I've also tried making a defect map, but the automatically generated defect map didn't have full coverage of the problem area. Is there a way to manually paint a defect map? (Presumably it's just a bitmap image of some sort.)

Any other ideas?

Read More...