OK, Friends & Neighbors. First, I'm starting to think that the original tape is damaged. The cptape utility suggested by Jochen seems to be Doing The Right Thing, but gets errors. dd has no qualms at all as long as the TKZ50-GA is set to a variable blocksize, but errors out on file #7. If I set "conv=noerror" for dd to ignore read errors, it saws away forever at the same file. And, if I boot the uVAX from either the original or the one "probably good" copy [1], sometimes it sees the RD53 disks as RD51. Not consistently. When it does, it hangs solid on a custom format. I'll be putting the -GA on a VMS box this evening and try the VMSTPCE copy. It makes sense to make the tape in its native OS. [1] I did get a good set of files with the TUHS cptape. Problem is that AFAICT the partner program, maketape, only understands 512*N blocksizes. The diags tape has a lot of files in 80B blocks. The good copy was made with dd, setting blocksize file by file. 82 files, that is.... Anybody remember the Bash test statement for [ FILE <larger-than> FILE] ? It would come in REAL handy, and I'm having a brain-fart. Doc