The method providing q (in AI) should derive directly from the one providing 2Theta to avoid errors

Test 2D regrouping after alignment of images (with reference image from fit2d)

Saxs cannot use split pixels yet 

dark & flat correction even for calibration (L. Erra id11)

Write tools to create a mask (or recycle code from Armando) 

in AI, have single method to write xrpd2 images
 
inlcude Raster Orientation of images

have geometries as pluggable classes (!!!)