Dear all,
I had problems with the reapeted use of graphs in Snap.
I have built a graph (actually a series of graphs, but I have the same problem for all graphs, so for simplicity, I will only talk about one) to prepare a stack for time series with StaMPS.
The processing of this graph has proceeded without problems and the results were correct, so I managed to process the time series without problems and also I got reasonable results.
In my test, I needed to process the same stack, but with different master: so I used the graph credated before, changing only the master image. Also in this case, Snap processing is performed (apparently) correctly.
I noticed that there were problems when I started processing with StaMPS. In the StaMPS procedure there is a routine that checks that the pixels relating to the interferogram phase are not distorted: in this step StaMPS identifies as distorting more than 90% of the pixels of the interferogram images and the processing is blocked. I tried to check the files processed by Snap: the text files were correct and the interferograms seemed reasonable.
I thought it was a master image intensity problem, so I always used the original graph using another master (no other changes): correct Snap processing, but StaMPS again encountered the same problem.
I have built a new graph, identical to the original in the contents and the name (not to change the whole procedure), but I’ve always had the same problem.
After a tenth of attempts, I thought I would use the same starting graph, changing the master and changing the name of the graph this time, and the problem disappeared, that is, processing with Snap proceeds well and StaMPS processing is executed correctly.
To do a cross check, I compared the results obtained by Snap with the original and renamed graphs: the text files are identical, while the interferograms in the case of the original graph (graph not renamed) appear to have been smoothed.
Is it possible that repeated use of a graph can produce incorrect processing in Snap?
Thank you to everyone.