to compare how similar both rasters are you can use the scatter plot tool
you can define a minimum and maximum by yourself in the color manipulation tab and apply it to both images. You can also apply the color of the first result to the second with this button 
(select ‘no’ when asked to search for min/max)
Both should acutually produce the same result. I guess the colors are just distributed differently because of different value ranges of both products.
Thank you for the suggestions. I will try it and get back to you 
.
Hi, I’ve been seeing the same problem when processing decomposition step. In fact it cost 36 hours to process a whole graph(containing preprocessing and decomposition). Can I ask how much faster it get when you divide the graph into two parts? Thanks a lot!
this depends on your system’s capabilities. But often large graphs are unnecessarily slow. So if you only have to apply this processing once I recommend avoiding graphs in general. They only make sense if you want to apply some workflow on multiple images.
Thanks! By the way in other blogs someone said when applying “Polarimetric Matrices” after “Multi-looking”, you’ll get the wrong answer. I notice that you didnt do the “Polarimetric Matrices” step? and directly apllied decomposition?
SNAP allows to perform polarimetric decomposition directly on the Intensities without converting to S/T/C matrix before. As many of them include a spatial window I would not multi-look before unless you want to change the pixel resolution for a certain reason.
Thanks!!! helped a lot 