It was really demoralizing to realize that so close to the program's end that my loop selection routine is straight up missing loops. I thought it was a local maximum issue, and I tried adjusting the background selection but that yeilded no fruit. I did some whiteboard tracing throught the program, and I thought it was a custom color table issue but couldn't find a solution.
I did some debugging with DMcK and emailed Anya about it. Thus we tried to do loop selection semi-manually, by orienting the loop selection window hotdog (instead of hamburger style) in order to see loops as streads instead of blobs.
Additionally, I ran into memory issues in working with the second data set, because using the homemade cutout was a memory hog, especially when working with full-sun images.
I reprocessed that first data set, started working on transition documentation, and began working on my final presentation.