Some datasets with visualised results in blocking artefacts on edges

[cheesema]

Due to the way the APR, and the viewer works, you can result in non smooth edges of images on 2 sides.

This is because the particle cells used to represent lower level images, don't align in many cases when the image size is not a power of 2.

This could be resolved by doing a final clipping mask over the actual imaging domain? I think I quickly looked into this and couldn't immediately find a solution, but I'm sure there must be one.