Abe, I just had a few suggestions that might alleviate a few of the issues that users are experiencing.
Allow the user to select Auto, Negative, or Positive for the type of film they are scanning, and default to auto. That way, an inexperienced user would still be able to make use of the automatic detection but could override if the detection isn’t working as expected. This would also allow experienced users to simply say what type of image they are attempting to digitize
Allow the user to select Color or Black and white. Black and white negatives get some funny color casts during the detection process now and have to be corrected in an edit, making the workflow take longer.
Use the UI to indicate what frame you are scanning. A box around the scan area would be fantastic. I know in iOS they have event triggers for TouchDown and TouchUpInside, touch down could show the box around the frame you are scanning and touch up inside could execute the capture. That way, if I put my finger down I know what I’m getting, and I get a cancel option by dragging my finger off the frame. I don’t know what kind of challenge the frame detection would add if you are doing it live though, might be a tough nut to crack performance wise.
Looks like a great first beta though, excited to see the update!