Thanks for bringing up that problem work-unit. As was the case before, this is an older image that predates the auto-centering software we developed for the imaging robot.

I just checked on the original image data for that batch -- some images are good, but many in that batch are badly off-centre, like yours. We won't get any usable data from that image. Fortunately, we have images earlier and later of the same experiments, so we can interpolate.

Thanks again,

Christian

Hello cubes,

you could create a small Distributed Human Project where volunteers could classify images as either "well centered" or "off center".Could help reduce or even eliminate these images from being processed and wasted on the grid.