That's not going to help the efforts on trying to reduce the memory footprint of imshow... (#6952,#8143, etc.). Not a blocker, but should be kept in mind.

Yeah, understood. This is the dumbest solution.

I don't quite understand why this rescaling is necessary, despite the nice comments....

OK, I only used the bigger float64 if the amin-amax > 1e8.

I think(?) you probably care more about maxabs/minabs.

Hmmm, not at all sure, because I once understood all the low-level reps of floats etc, but not so good at it anymore. Happy if someone wants to pick this up and do it better.

But, I think the error comes in when we do:

and the resulting float doesn't preserve all the bits in theunit32.

This is a reasonable approach.

The scaling is to be able to track the over/under pixels and get around ringing in the interpolations.

Discussed this on the phone call. Will investigate

• delaying conversion to float as long as possible
• avoiding float128 due to platform-dependence
• add a test

Test added, and changes requested above made... Thanks!