Zombse

The Zombie Stack Exchanges That Just Won't Die

View the Project on GitHub anjackson/zombse

What disk image format for use vs preservation

When imaging 3.5" (both double sided double density, as well as high density) floppy disks for the Macintosh, using a Kryoflux…

​1) What image format should I be dumping if my intent is to use these images within emulators such as Sheepshaver, Basilisk II, or Mini vMac?

​2) Is this image format I will be dumping for use purposes also sufficient for long term preservation? Should I also be dumping a format that preserves lower level information?

Ben Fino-Radin

Comments

Answer by Michael Kjörling

I cannot comment on the specific products mentioned, but I do have some general thoughts on the issues raised in the question.

​1) What image format should I be dumping if my intent is to use these images within emulators such as Sheepshaver, Basilisk II, or Mini vMac?

Most emulators want simple byte reads of the media in question. On Unix and Unix-like systems, this is usually achieved using the dd tool which ships with the OS, although other options do exist; on Windows, separate imaging tools are normally needed.

Barring such difficulties as inter-sector copy protection measures (which I believe were used to some extent at one point, although I don't know if they were employed on the Macintosh), this gives you a copy of all the data on the media that the original operating system would be able to access. Boot sector, file system metadata, recorded-as-unused blocks, everything gets copied.

Depending on one's needs, knowledge of the file system used could then be used to extract files and folders from the media image if that is sufficient for preservation. If you are trying to preserve the payload data rather than the exact data and metadata structures on the disk, and if it can be done either losslessly or with acceptable losses, it makes sense to not just image the disks but also transform the data into a more accessible format. The raw disk images could still be retained, but used for secondary access should any specific need arise.

​2) Is this image format I will be dumping for use purposes also sufficient for long term preservation? Should I also be dumping a format that preserves lower level information?

In almost all cases (see caveat about copy protection above), I don't see any real use for any lower level information than what can be obtained by making a byte-by-byte read and copy of the media. You would then be looking at specialized tools to read (assuming magnetic media such as floppies; optical media such as CDs would be different, and solid-state media would be a whole different ballgame) the magnetic flux variations in the physical space between sectors. Such reads could, in principle and when properly analyzed, reveal traces of previously overwritten data, but they could just as well simply indicate drive read/write head misalignment. No two floppy disk drives have read/write heads which are perfectly aligned with each other other than by pure chance, so some variance in the physical location of the written data is to be expected. Normally the read head(s) simply detect whatever signal happens to be strongest, since the difference between the desired and old signals will be orders of magnitude.

The hard part about copying old Apple media in particular, such as old Macintosh or Apple II/III floppy disks, is that depending on the media size, some of them use GCR encoding rather than the MFM encoding which eventually replaced it. Sourcing a working floppy disk drive today which can read GCR and can interface with modern hardware might be difficult, so a multi-stage approach might be needed: first transfer a binary image of the media onto a system which can natively read GCR floppies, then transfer that image by some other means onto a more modern system for further use.

Comments