The following is my understanding? When doing an imaged based backup, TI directly uses the file systems internal structures to directly read the clusters constituting each file and to get the atributes of the files, i.e., the file system is not actually used to OPEN any file. When doing a file-based backup, TI does operate thru the file system using the built-in API functions, e.g., FindFirstFile, FindNextFile, CreateFie ReadFile/BackupRead, GetFileAttributes, etc. . Or is the above a misunderstanding?