protocol align volume: bugs and improvements

[albertmena]

• mask number is an integer number - alway bug comes up.
• exhaustive - never finishes
• local works but depends on the initial orientation
• exhaustive + local — nope
• fast fourier - works best
• my walk around - to make the smallest box size possible and add a mask. But then after Fourier - the density of mask appears on the corners
• also will be good if the program could recognise the handiness of 2 structures
• about accuracy - sometimes it is very hard to align 2 maps when they a) have big box size and b) oriented way differently in space