As a reminder that some projects are never quite finished, I recently received a message from somebody who had come across a different type of Disklavier disk. These were made for the Yamaha Disk Orchestra Module (DOM-30), a pretty slick looking synthesizer from the early 1990s. Unfortunately, my disklav.py tool only showed garbage when ran against an image made from one of these disks.

Luckily, our new friend is something of a data collector, and uploaded the disk images to the Internet Archive. After downloading the images and comparing them to the ones used to create the script, it was clear there were some major differences in the format. But there were also plenty of similarities, which meant adding support for them was fairly straightforward.

The biggest difference is that the Table of Contents that lists the tracks isn't at a set position at the beginning of the disk like it is on the other Disklavier floppies. Instead, it's tacked on at the end after all the files. Naturally this means its position is variable (since the tracks themselves aren't of a fixed size),  so now the script has to search for it instead of picking it up from a known location. Of course when you're talking about a 720 KB disk image, the searching process is pretty much instantaneous anyway, so it doesn't change the user experience any.

That said, there does seem to be something unusual going on. Between the two disk images I have access to, three of the tracks end with a different byte sequence than all of the others. This means the script can't automatically determine the start/end points, and ignores them. I was thinking it could be some kind of corruption (after all, these are 30+ year old floppies), but it's such a specific issue that I wonder if something else is going on.

I'm going to keep looking into the issue, but without more DOM-30 disks to look at, it may be tricky. If anybody out there in the Hackaday community actually has some of these disks laying around, I'd love to hear from you.