Bob,
All excellent updates to 3dDeconvolve! Thanks!
One of them scares me though.
"If two identical (nonzero) regressors are input, the program will complain but continue the analysis. In this case, each one would get half the coefficient, which only seems fair. However, your interpretation of such cases should be made with caution"
Your last sentence is really, really important as you are well-aware. Can the complaint actually stop processing and ask for user input? I'm afraid that if run in batch mode / in the background, this could go by without being seen and make for some very bad data analysis.
Craig