So you’ve built the hardware, programmed and tested it! What now?

The programs

I’m sorry to say that the client program is very badly documented — it’s moving too quickly for the documentation to keep up. It does respond to --help or help depending on context. There are some common properties, described below.

Core concepts

FluxEngine fundamentally takes file system images and puts them on disk; or reads the disk and produces a file system image.

A file system image typically has the extension .img. It contains a sector-by-sector record of the decoded data on the disk. For example, on a disk with 512 byte sectors, one sector will occupy 512 bytes. These are typically what you want in everyday life.

FluxEngine can also record the raw magnetic data on the disk into a file, which we call a flux file. This contains all the low-level data which the drive produced as the disk rotated. These are continuous streams of samples from the disk and are completely useless in day-to-day life. FluxEngine uses its own format for this, .flux, although it’s capable of limited interchange with Kryoflux, Supercard Pro and Catweasel files. A flux file will typically contain from 80 to 150 kilobytes of data per track.

In general, FluxEngine can use either a real disk or a flux file interchangeably: you can specify either at (very nearly) any time. A very common workflow is to read a disk to a flux file, and then reread from the flux file while changing the decoder options, to save disk wear. It’s also much faster.

Source and destination specifiers

When reading from or writing flux (either from or to a real disk, or a flux file), use the --source (-s) and --dest (-d) options to tell FluxEngine which bits of the disk you want to access. These use a common syntax:

fluxengine read ibm -s fakedisk.flux:t=0-79:s=0
  • To access a real disk, leave out the filename (so :t=0-79:s=0).

  • To access only some tracks, use the t= modifier. To access only some sides, use the s= modifier.

  • Inside a modifier, you can use a comma separated list of ranges. So :t=0-3 and :t=0,1,2,3 are equivalent.

  • When specifying a range, you can also specify the step. For example, :t=0-79x2 would be used when accessing a 40-track disk with double stepping.

  • To read from drive 1 instead of drive 0, use :d=1.

  • To read from a set of KryoFlux stream files, specify the path to the directory containing the files with a trailing slash; so some/files/:t=0-10. There must be a files for a single disk only in the directory.

Source and destination specifiers work entirely in physical units. FluxEngine is intended to be connected to an 80 (or 82) track double sided drive, and these are the units used. If the format you’re trying to access lays out its tracks differently, then you’ll need a specifier which tells FluxEngine how to find those tracks. See the 40-track disk example above.

If you don’t specify a modifier, you’ll get the default, which should be sensible for the command you’re using.

Important note: FluxEngine always uses zero-based units (even if the disk format says otherwise).

Input and output specifiers

When reading or writing file system images, use the --input (-i) and --output (-o) options to specify the file and file format. These use a very similar syntax to the source and destination specifiers (because they’re based on the same microformat library!) but with different specifiers. Also, the exact format varies according to the extension:

  • .img or .adf: raw sector images in CHS order. Append :c=80:h=2:s=9:b=512 to set the geometry; that specifies 80 cylinders, 2 heads, 9 sectors, 512 bytes per sector. For output files (--output) the geometry will be autodetected if left unspecified. For input files you normally have to specify it.

  • .ldbs: John Elliott’s LDBS disk image format, which is consumable by the libdsk suite of tools. This allows things like variable numbers of sectors per track (e.g. Macintosh or Commodore 64) and also provides information about whether sectors were read correctly. You can use libdsk to convert this to other formats, using a command like this:

    $ dsktrans out.ldbs -otype tele out.td0

    …to convert to TeleDisk format. (Note you have to use dsktrans rather than dskconv due to a minor bug in the geometry hadnling.)

    FluxEngine’s LDBS support is currently limited to write only, and it doesn’t store a lot of the more esoteric LDBS features like format types, timings, and data rates.

  • .d64: the venerable Commodore 64 disk image format as used by the 1540, 1541, etc. This is a special-purpose format due to the weird layout of 1540 disks and while you can use this for non-Commodore disks the result will be gibberish. Use this to image Commodore 64 disks and load the result into an emulator.

    FluxEngine’s D64 support is currently limited to write only. It will work with up to 40 logical tracks.

High density disks

High density disks use a different magnetic medium to low and double density disks, and have different magnetic properties. 3.5” drives can usually autodetect what kind of medium is inserted into the drive based on the hole in the disk casing, but 5.25” drives can’t. As a result, you need to explicitly tell FluxEngine on the command line whether you’re using a high density disk or not with the --hd flag. If you don’t do this, your disks may not read correctly and will certainly fail to write correctly.

You can distinguish high density 5.25” floppies from the presence of a traction ring around the hole in the middle of the disk; if the ring is not present, the disk is probably high density. However, this isn’t always the case, and reading the disk label is much more reliable.

Lots more information on high density vs double density disks can be found here.

Other important flags

These flags apply to many operations and are useful for modifying the overall behaviour.

  • --revolutions=X: when reading, spin the disk X times. X can be a floating point number. The default is usually 1.25. Some formats default to 1. Increasing the number will sample more data, and can be useful on dubious disks to try and get a better read.

  • --sync-with-index=true|false: wait for an index pulse before starting to read the disk. (Ignored for write operations.) By default FluxEngine doesn’t, as it makes reads faster, but when diagnosing disk problems it’s helpful to have all your data start at the same place each time.

  • --index-source=X, --write-index-source=X: set the source of index pulses when reading or writing respectively. This is for use with drives which don’t produce index pulse data. Use 0 to get index pulses from the drive, 1 to fake 300RPM pulses, or 2 to fake 360RPM pulses. Note this has no effect on the drive, so it doesn’t help with flippy disks, but is useful for using very old drives with FluxEngine itself. If you use this option, then any index marks in the sampled flux are, of course, garbage.

The commands

The FluxEngine client software is a largely undocumented set of small tools. You’ll have to play with them. They all support --help. They’re not installed anywhere and after building you’ll find them in the .obj directory.

  • fluxengine erase: wipes (all or part of) a disk — erases it without writing a pulsetrain.

  • fluxengine inspect: dumps the raw pulsetrain / bitstream to stdout. Mainly useful for debugging.

  • fluxengine read *: reads various formats of disk. See the per-format documentation linked from the table in the index page. These all take an optional --write-flux option which will cause the raw flux to be written to the specified file as well as the normal decode. There are various --dump options for showing raw data during the decode process, and --write-csv will write a copious CSV report of the state of every sector in the file in machine-readable format.

  • fluxengine write *: writes various formats of disk. Again, see the per-format documentation in the index page.

  • fluxengine writeflux: writes raw flux files. This is much less useful than you might think: you can’t reliably write flux files read from a disk to another disk. (See the FAQ for more information.) It’s mainly useful for flux files synthesised by the other fluxengine write commands.

  • fluxengine writetestpattern: writes regular pulses (at a configurable interval) to the disk. Useful for testing drive jitter, erasing disks in a more secure fashion, or simply debugging. Goes well with fluxengine inspect.

  • fluxengine rpm: measures the RPM of the drive (requires a disk in the drive). Mainly useful for testing.

  • fluxengine seek: moves the head. Mainly useful for finding out whether your drive can seek to track 82. (Mine can’t.)

  • fluxengine test bandwidth: measures your USB throughput. You don’t need a disk in the drive for this one.

  • fluxengine test voltages: measures your FDD bus signal voltages, which is useful for testing for termination issues.

  • fluxengine upgradefluxfile: occasionally I need to upgrade the flux file format in a non-backwards-compatible way; this tool will upgrade flux files to the new format.

  • fluxengine convert: converts flux files from various formats to various other formats. You can use this to convert Catweasel flux files to FluxEngine’s native format, FluxEngine flux files to various other formats useful for debugging (including VCD which can be loaded into sigrok), and bidirectional conversion to and from Supercard Pro .scp format.

    Important SCP note: import (fluxengine convert scptoflux) should be fairly robust, but export (fluxengine convert fluxtoscp) should only be done with great caution as FluxEngine files contain features which can’t be represented very well in .scp format and they’re probably pretty dubious. As ever, please get in touch with any reports.

Commands which normally take --source or --dest get a sensible default if left unspecified. fluxengine read ibm on its own will read drive 0 and write an ibm.img file.


When doing a read (either from a real disk or from a flux file) you can use --write-svg=output.svg to write out a graphical visualisation of where the sectors are on the disk. Here’s a IBM PC 1232kB disk:

A disk visualisation

Blue represents data, light blue a header, and red is a bad sector. Side zero is on the left and side one is on the right.

The visualiser is extremely primitive and you have to explicitly tell it how big your disk is, in milliseconds. The default is 200ms (for a normal 3.5” disk). For a 5.25” disk, use --visualiser-period=166.

Extra programs

Supplied with FluxEngine, but not part of FluxEngine, are some little tools I wrote to do useful things. These are built alongside FluxEngine.

So you’ve just received, say, a huge pile of old Brother word processor disks containing valuable historical data, and you want to read them.

Typically I do this:

$ fluxengine read brother -s :d=0 -o brother.img --write-flux=brother.flux --overwrite --write-svg=brother.svg

This will read the disk in drive 0 and write out a filesystem image. It’ll also copy the flux to brother.flux (replacing any old one) and write out an SVG visualisation. If I then need to tweak the settings, I can rerun the decode without having to physically touch the disk like this:

$ fluxengine read brother -s brother.flux -o brother.img --write-svg=brother.svg

Apart from being drastically faster, this avoids touching the (potentially physically fragile) disk.

If the disk is particularly dodgy, you can force FluxEngine not to retry failed reads with --retries=0. This reduces head movement. This is not recommended. Floppy disks are inherently unreliable, and the occasional bit error is perfectly normal; FluxEngine will retry and the sector will read fine next time. If you prevent retries, then not only do you get bad sectors in the resulting image, but the flux file itself contains the bad read, so attempting a decode of it will just reproduce the same bad data.

See also the troubleshooting page for more information about reading dubious disks.

Previous page