Nev Young wrote: > why not create an image that is a full track image from index > hole to index hole including all the inter sector guff. That > should cover just about everything.
It would indeed, if there was a reliable way to dump them! You'd really need all the sync marks as well as the raw track data, but even reading the raw track data is a problem... Unfortunately, the WD1772-02 controller leaves the address mark detector enabled during the diagnostic tracks reads, meaning certain patterns in the data can cause a false sync. If this happens the controller returns MFM bits instead of data bits from that point on in the track, making what you read essentially useless. I tried this method when writing the original MakeSDF scanner, before spotting the problem and working out what the heck it was doing. The PC controller allows you to read ID field headers and data, but doesn't give access to anything in the gap areas. There is a Disk2FDI program that can do it, but is a commercial program and requires 2 floppy drives. It spins both drives up and starts the read of a high-density disk in drive 1, then switches the drive-select to drive 2, with the controller then returning MFM _and_ data bits for the disk in drive 2. Very clever, but it's DOS-only and wasn't particularly reliable when I tried it a while ago. Another option is to use a Catweasel controller card, which can read just about any disk at very low level. They're not cheap though, and it's not really something most people will have access to for dumping. For now the header/sector-level scanning seems to be about the best solution, and will cope with all but one SAM disk I've seen so far :-) Si
