I am evaluating pinpoint.

I have found that my raw image will get plate solved easily.

However, after I have calibrated it and aligned it with other images, and it has been
saved in 32 bit floating point format, it will no longer plate solve.

I think the complete calibration process requires a larger than 16 bit file format.

How can we handle this? Will pin point handle 32 bit signed integer?

Arthur Sweeney