"Weak Disk" protections in TransCopy format

[NewRisingSun]

I am planning to try adding support for copy-protected PC disks to a certain popular DOS emulator box, and have begun researching the TransCopy format, because it should have all information about the disk necessary to emulate all but "laser hole" protections, and is much simpler and more convenient to use than Kryoflux stream files, and unlike IPF, can be produced by everyone who has an Option Board or a Kryoflux device, the latter using PCE's "pfi" utility to convert Kryoflux stream files to TransCopy format. It's also very simple to decode, making it easy to add support. So far, I use the PCE-PC emulator to try whether TransCopy images converted from Kryoflux stream files work or not.

Converted Kyroflux images work for all "normal" protections, including H.L.S. Duplication, Softguard SUPERLoK, OSI-1, Electronic Arts Interlock, unnamed Sierra On-Line method for Radio Shack releases, and of course anything that just uses non-standard sector sizes. The one exception is Formaster Copylock, and that one clearly is PCE's fault because it does not implement the Floppy Disk Controller's "Read Track" command correctly. The "non-normal" type of protection that fails to work at least with TransCopy images converted from Kyroflux streams are complex "weak bit" protections. The games affected are "Boulder Dash II" and "Arac" in their Prism Leisure Software releases, which use the "Minder" protection scheme (see page 15 in the PDF). The keydisk checking code reads a sector several times, but instead of just checking whether the sector is "somehow" different (as a simple weak bits scheme would), it wants precisely only certain parts of the sector to change.

While I have several TransCopy images from a real Option Board, none use a weak bits scheme. I therefore do not know how the format stores such information. From looking at the software, I understand you have to tell it to specifically look for weak bits, otherwise they are not reproduced properly either.

So, my request is the following: I would like to receive a Transcopy image, made with a real Option Board, of any game or application disk that uses Weak Bits protection. Ideally, there would be two images of the same disk: one produced without the weak bits option set in the imaging software and one with the weak bits option enabled, so I can see how the image differs.

 

[fs5500]

I can send TC image of Gauntlet (Original release) by Mindscape.
This game includes Weak bit copy protection on Track 0.

 

[NewRisingSun]

Looking at these images, I don't see anything special about them, apart from the fact that the protected sector is different between the several files coming from separate imaging attempts after a series of zero flux reversals, which is expected. I suppose then that the Transcopy Software's "copy weak bits" setting then does nothing more than to just accept read errors instead of trying to correct them.

 

[fs5500]

This is a useful tip to run on weak bit copy protected image on pce.

1. At least, dump the copy protected disks more than twice by using Teledisk or TCOB / KF.
2. Then convert dumped image to .psi format by using pfi.exe pri.exe psi.exe.
3. Merge the .psi image as follows.

psi disk1a.psi -m disk1b.psi disk1.psi

disk1.psi is now active with alternate sector.

This is only way to run the weak bit copy protected image on PCE.

 

[NewRisingSun]

I have started compiling PCE myself. Now PCE properly emulates the "Weak Bit" protection using TransCopy images. Gauntlet now works properly. The "Minder" protection still doesn't, because in addition to weak bits, it also reads a (different) sector across the index hole. The latter is a problem with TransCopy images because the track image ends at the index hole, and continuing after the end using the data from the beginning of the track involves a loss of MFM framing.

 

[fs5500]

I have started compiling PCE myself. Now PCE properly emulates the "Weak Bit" protection using TransCopy images. Gauntlet now works properly. The "Minder" protection still doesn't, because in addition to weak bits, it also reads a (different) sector across the index hole. The latter is a problem with TransCopy images because the track image ends at the index hole, and continuing after the end using the data from the beginning of the track involves a loss of MFM framing.

If TC Image can be mounted on PCE, it will be awesome news.

 

[NewRisingSun]

I just noticed that the sole reason for my problems is that PRI does not create proper TransCopy images. They're good enough for most types of protection, including weak bits. But for sectors that span across the index signal, PFI fails to include the necessary information, so that proper MFM framing is lost at the index signal, whereas .TC images created with the original Option Board software include such information. I will see whether the .PRI format, which is necessary as an intermediary format when converting from from Kryoflux stream files to TransCopy format, can be used for that purpose, either by adding support for it, or writing a better .PRI->.TC converter.

 

[anormal]

Don't you think is much better idea to support KF formats (stream or ctraw) and SCP dumps?
Both these formats contains much more information that TC, also how many working TCOB has you seen lately , at 300$ in eBay...

And there is source code out there (Hampa PCE and, some code in mess and also disk-utilities) for reading and supporting these new formats

 

[NewRisingSun]

Don't you think is much better idea to support KF formats (stream or ctraw) and SCP dumps?

No, working on Kryoflux streams directly is too slow even on current systems, and having eighty files per disk is awkward for the user as well. But adding raw MFM support for Transcopy will be something others can build upon if Kryoflux stream file support is desired after all.

Both these formats contains much more information that TC,

It supports variable bit rates, which cannot be used in IBM formats. Otherwise, a PRI/Transcopy image is just as much a simple record of flux changes.

also how many working TCOB has you seen lately , at 300$ in eBay..

You don't need an actual Option Board. You just need to convert Kryoflux stream files to PRI or Transcopy format (using an updated version of PRI).

and there is source code out there (Hampa PCE and, some code in mess and also disk-utilities) for reading and supporting these new formats

Of course I know that, otherwise I could not be compiling PCE. MESS code cannot be used because it is not GPL-licensed.

 

[anormal]

Nice, you are not using simple sector dumps, but MFM dumps or PRI image !!

Just for curiosity, other people could explain, for supporting advanced copy protections in a PC emu, is not necessary to implement cycle-exact FDC emulation and maybe cycle exact (or precise timing, or the level necesary) DMA / CPU / etc... ?

 

[NewRisingSun]

No, because there were so many different IBM PC compatibles even in the 1980s with vastly different speeds, cycle-exact timing was not something any software author would have even bothered to try. In fact, I have not seen a single protection type that verifiably does not work with an emulated floppy disk controller sending all data immediately, although I admit to not having seen all of them.