All general stuff about SPS, like philosophy and mission - discuss it here!
5 posts • Page 1 of 1
Hello, I've had my new Kryoflux for a week or so now and have dumped a handful of disks with it including using to to recover data from some old disks I have that were starting to flake out a bit. I also happen to have one copy protected DOS game which I'm sure many here are aware of...Lemmings. I was hoping I could just make a stream dump of the disk and then write it to a new floppy and have it work but that clearly didn't happen. I spent several days of research about it and found the game is using weak sector protection and found a few threads on here talking about it but nothing about how to possibly write an image that would replicate those weak sectors so they appear the same as the original disk. If anyone has any tools I can use that would be great as viewing the image I made in HxC shows a couple of sectors on the first 6 tracks of the disk (0-5) with BAD CRC which I'm guessing are the weak sectors that I would have to alter the data of to make the Kryoflux write back all 0 bits or whatever is needed in those areas to recreate the same effect as the original disk so I can make a clean copy and preserve my original disks.
That kind of protection is probably not possible to write back as a direct disk copy currently as the software has no knowledge of what portion of a track is supposed to be "weak" and thus needs to be "written" using different techniques unlike standard data.
Yeah I figured that though I am aware that the steam dumps contain multiple rotations (5 by default) so it should be possible to see the changes in data along those weak sectors as random 1 bits most likely show up in different rotations unlike the rest of the tracks which decode the same data and pass the CRC check. Those BAD CRC areas really stand out when using the Track Analyzer in HxC which is where at least I believe the weak sectors are and I did read that if you could break the normal MFM flux transitions and just write all 0 bits in those areas it should recreate the same effect as whatever they did on the original disk but I simply have no way of modifying the image to do that other them maybe a hex editor but I have no way to tell what parts of the data I'm looking at or what to replace them with if I did to make the Kryoflux write all 0 bits in those areas and then cross my fingers and hope it works. I even considered just using HxC to convert the Kryoflux stream dump I have to another flux level format and see about modifying those but I ran in to pretty much the same situation, not being able to tell what data to modify or what to modify it with so something that would work with another format I can then convert back to a Kryoflux stream dump could be a viable solution for me if it works and if it doesn't well it will at least be an interesting learning experience in how to manipulate flux transitions in the disk images to get different results. If anyone else wants to see what I'm talking about there is a Kryoflux dump on archive.org of Lemmings that appears to be pretty much identical to my copy though mine has a different label and my dump has less jitter based on what I saw in HxC. I also made two different steam dumps of the disk but sadly the metal hub on my original disk has broken off so it now doesn't work and I have two other copies that were modified at some point (one even had a nasty boot virus on it that replaced the boot sector but I used a hex editor to replace it with the boot sector I had from a normal IMG image I found a copy of and thankfully it didn't effect the copy protection on the disk) and this was the only good unmodified copy I managed to get my hands on so making a replacement copy of it that actually loads without cracking the game would be nice.
Thinking about this for a moment, would a potential way for the KryoFlux to write weak sectors be to write transitioning 1's and 0's at a very significantly higher frequency than whatever the typical shortest transition is? That is assuming both the drive electronics and the board would be capable of doing so. I wonder if that would generate enough noise for the analogue circuitry to read and interpret incorrectly for different revolutions.