[EDIT: This is a repost; made this compliant with the rules, so far as I can read.]

Since there appears to be some demand for backing up StarForce protected titles but only bits and pieces of info scattered around and about in a ton of everything but the kitchen sink as it usually happens, below goes the attempt of a guide. Credits go to AlB for developing the method and the associated SFCopy utilities; I believe this person is known in these forums as eip.

The title which we will use as case study is Prince of Persia: The Two Thrones, which, upon looking at driver version infos, uses Starforce (further: SF) version 3.5 (some say 3.6).

First, about what does not work.

With earlier versions of SF, you could create Alcohol image with DPM, then burn with RMPS option on. Of course, you'd need to have Alcohol installed with RMPS emulation enabled to play from the backup copy. Today, SF accesses IDE drives using its own device drivers -- so if you have an IDE drive, this method will no longer work since SF will ignore Alcohol's emulation.

If you have multiple IDE and non-IDE (including emulated) CD/DVD drives in your system, you need to physically disconnect all IDE drives before SF will accept a disc in non-IDE or virtual drive, because SF prefers drives that it can access directly. SFNightmare, to my knowledge, no longer works, nor does disabling the corresponding IDE channel in BIOS.

DVD Versions and Emulation

If you have a DVD version, pretty much the only option that I know of is emulation. Rip an image with Alcohol with DPM option on. Typically the image is larger than 4G, so you need to have NTFS file system (FAT/FAT32 cannot store files larger than 4G). Then physically disconnect IDE drives and use Daemon Tools v4 to mount the image. Alcohol's virtual drive will not work because SF attempts to block Windows file system for the period it checks the disc in order to detect emulations. DT gets around that, Alcohol does not. If you want to use Alcohol, you probably need SFCure as well to prevent SF from blocking the file system -- but I haven't tried this. Verify that the image works, compress the image using some archiver (RAR, for instance), and burn the file on another DVD. Voila, a backup. You won't be able to restore a DVD from it if you accidentally put the retail disc in your microwave during your cleanup and then forget to take it out before you cook a few eggs there, but at least you'll still be able to play the game then.

CD Versions

You are a bit more in luck if you have a CD version of the title you are looking at.

Limitation: The method described should work if you have a non-recording (read only) CD/DVD drive in your system. On recorder drives, SF will check ATIP and reject the disc if it determines that it's not ROM. "Hide Media Type" options will not work on IDE drives, again because SF accesses them directly.

Just like with the DVD version, rip an image (if you don't already have it) of the play disc with DPM on (high precision/4x setting recommended, depending on the drive and the quality of the disc). Have two identical CD-R media (or one CD-RW) ready.

Now burn the image of the play disc onto one of the CD-R media (or CD-RW) using Alcohol with "Securom *NEW" setting (i.e. Write method: RAW DAO, Write: On, Simulation: Off, Don't close last session: Off, Rectify subchannel data: Off, Burn "RMPS": Off). Speed does not matter, but make sure you use exactly the same settings, including speed, later on; this is important. Once it's done, try to run the game with this disc in a non-recorder drive (IDE, if present). If it works, you are done. Most likely, it won't because the DPM info of the copy does not match that of the original.

Further, you'll need to use the utilities found here -- from now on referred to as SFCopy. These command line utilities, after a few manipluations on your part, will allow you to patch your .mdf image so that, when burned on identical media, the DPM info of the copy (so far as StarForce protection is concerned) matches that of the original disc. The archive contains multiple versions of the utilities in sub-archives. Use the latest, currently v2.0. Extract the contents of this subarchive into the folder where you have your .mdf image, then open the command prompt and change to that folder ("drive":, then cd "folder name"). A few notes here: SFCopy was written by a Russian member of this forum; the manual found with the archive is in Russian. People have since translated bits and pieces; for more information on SFCopy in English, refer to this thread in the Copy Protections forum. Last but not least, remember that you need to patch based on your particular media, for your particular recorder, for your particular recording settings. Tools that exist that dont't take all of these into account are likely not to work, because all of the above influence DPM.

SFCopy utilities use a config file. The first 4 lines of the file are the same for a given product. The last 2 lines you will have to figure out yourself, using the utilities. For Prince of Persia: The Two Thrones CD version, you will use the following 4 lines in the config file to start with. The first line contains volume label, the second line -- the CD key (which is the same for all CD copies of the game -- DVD version has a different key), and I won't go into technicalities about the 3rd and the 4th line -- we'll just say that they are based on the CD key and figuring them out is not trivial -- but again they are globally the same for this particular title/CD key combination. For more info, and on how to get the numbers for a particular title you are looking at, refer to the above thread and/or read/translate the SFCopy manuals; you will also need an additional tool.
Note: The CD key for this particular title, CD version starts with BZ7A9 (just so you can verify). SFCopy does not utilize it -- it's just there for reference; what's important are the 4 numbers that are derived from the CD key. Therefore you can put any key on that line, but make sure you use the real CD key when prompted by the game.

Save the file in the same folder as where your .mdf image is (I'll use the name popt2t.txt from now on).

Next, you need to measure the angular shifts between 6 sectors that SF tests, on the original disc (here inserted in drive H:). To that end, we use the following command in the command prompt:Note: to get more accurate measurements, it is preferred to run this when the drive is fully spun up and positioned, or run this a second time. When done, you will get something like this:
What we are interested here are the angles. Stringing them together, we add them to the config file, which now looks like this:
Note here that your numbers for this particular title can be slightly different (because of a different copy) but should be within the ballpark (+/- 45 degrees). If they are something wildly different, something is wrong. Make sure your drive is in CAV mode as you do this; "RPM changed by" above should read no more than 2% generally -- if it's something crazy like 15%, you are probably in CLV -- the utility, from what I understand, does not handle CLV properly. In this case, use a tool such as Nero DriveSpeed to set the drive to a higher speed setting (forcing it into CAV) or append the sX parameter to the command line, where X is the speed factor (16 or above should yield CAV mode).

These angle values are what StarForce expects from the disc. Now let's see how different the copy that you've just made is. Insert the copy into the drive of your choice (here I:) and run the same command:You should get something like this:
As expected, not even close. Next, notice the 6 numbers at the bottom. These are the sector shifts for the patcher; we need them. Grab these and append them to the config file, which now looks like this:
With this final config file, we are ready to run the patcher. The patcher will add twin blocks to the image file so that, while the disc remains (as seen by any software) an exact copy, the angle values measured above match that of the original disc. The disc thus created should be readable in most readers provided that there are no negative numbers on the last config line (negative means remove blocks, positive means replicate blocks). If there are small negatives, you may try replacing them with zeroes and seeing if the resulting copy works, or you may create separate copies for installation (the copy that you've just created will do fine) and playing -- most games nowadays only need CD in drive to authenticate; they don't normally read data from the CD.

Run the patcher giving it the name of the config file (popt2t.txt), original image (here popt2tc) and patched (output) image (here popt2tc_p), without .mdf/.mds extension:There will be a progress indicator; the messy explosion will let you know when eggs are done.

Now use Alcohol to burn the resulting image onto an identical CD-R media (or erase the CD-RW and burn on it again). Just like with Twinpeak, you will get a warning stating that the image is larger than the TOC says. Select "Keep TOC, burn to the end of the image" and continue.

When done, insert the new copy into the drive of your choice (here I:), remove the last line (sector shifts) from the config file and run sfcdvd again to check the angle values and see how close the copy is to the original:You should get something like this:
Ah. Much better. Now place the copy in the non-recorder IDE drive and try to run the game. It should work -- given that the values in the 3rd and 4th config lines were correct, that particular CD key was used, and the numbers on the copy are close enough to the original. For SF, "close enough" is within around 45 degrees I think. I believe SF will perform several attempts to authenticate given that the first one fails due to random variation. It will give up in about 2 minutes if it consistently can't get a match.


I've tried this particular title with a few media types and all of them worked. I have 2 drives: Plextor W8432T (CD-RW) and Sony DDU1621 (DVD-ROM). I used Plextor to obtain adjustments, but used them both to verify the resulting copies. The following media were tried:

Philips/TDK 1x-52x CD-R, 700M (CMC Magnetics 97m26s66f)
Memorex 1x-4x CD-RW, 650M (CMC Magnetics 97m26s65f)
Fujifilm 1x-48x CD-R, 700M (Prodisc 97m32s19f)

Below are the before (original copy) and after (patched copy) results:

Philips 1x-52x CD-R, 700M (CMC Magnetics 97m26s66f)

Before (Plextor drive)
Phase3: Step 1: Angle 106 degree = 287+179 (SPR 19.6 = 18.4 deg/sector)
Phase3: Step 2: Angle 122 degree = 297-175 (SPR 20.1 = 17.9 deg/sector)
Phase3: Step 3: Angle 136 degree = 317+179 (SPR 20.6 = 17.4 deg/sector)
Phase3: Step 4: Angle 351 degree = 206+145 (SPR 21.2 = 17.0 deg/sector)
Phase3: Step 5: Angle 2 degree = 216+146 (SPR 21.7 = 16.6 deg/sector)
Phase3: Step 6: Angle 88 degree = 320+128 (SPR 22.2 = 16.3 deg/sector)

Needless to say, the original copy does not work.

After (Plextor drive)
Phase3: Step 1: Angle 280 degree = 287-7 (SPR 19.6 = 18.4 deg/sector)
Phase3: Step 2: Angle 298 degree = 297+1 (SPR 20.1 = 17.9 deg/sector)
Phase3: Step 3: Angle 324 degree = 317+7 (SPR 20.7 = 17.4 deg/sector)
Phase3: Step 4: Angle 215 degree = 206+9 (SPR 21.2 = 17.0 deg/sector)
Phase3: Step 5: Angle 222 degree = 216+6 (SPR 21.7 = 16.6 deg/sector)
Phase3: Step 6: Angle 331 degree = 320+11 (SPR 22.2 = 16.2 deg/sector)

After (Sony drive)
Phase3: Step 1: Angle 251 degree = 287-36 (SPR 19.5 = 18.4 deg/sector)
Phase3: Step 2: Angle 279 degree = 297-18 (SPR 20.1 = 17.9 deg/sector)
Phase3: Step 3: Angle 307 degree = 317-10 (SPR 20.6 = 17.5 deg/sector)
Phase3: Step 4: Angle 198 degree = 206-8 (SPR 21.1 = 17.1 deg/sector)
Phase3: Step 5: Angle 206 degree = 216-10 (SPR 21.5 = 16.7 deg/sector)
Phase3: Step 6: Angle 326 degree = 320+6 (SPR 22.0 = 16.3 deg/sector)

The patched copy works.

Memorex 1x-4x CD-RW, 650M (CMC Magnetics 97m26s65f)

Before (Plextor drive)
Phase3: Step 1: Angle 15 degree = 287+88 (SPR 19.9 = 18.1 deg/sector)
Phase3: Step 2: Angle 31 degree = 297+94 (SPR 20.5 = 17.6 deg/sector)
Phase3: Step 3: Angle 55 degree = 317+98 (SPR 21.0 = 17.1 deg/sector)
Phase3: Step 4: Angle 284 degree = 206+78 (SPR 21.5 = 16.7 deg/sector)
Phase3: Step 5: Angle 299 degree = 216+83 (SPR 22.1 = 16.3 deg/sector)
Phase3: Step 6: Angle 61 degree = 320+101 (SPR 22.6 = 16.0 deg/sector)

Again the original copy does not work.

After (Plextor drive)
Phase3: Step 1: Angle 295 degree = 287+8 (SPR 19.9 = 18.1 deg/sector)
Phase3: Step 2: Angle 310 degree = 297+13 (SPR 20.5 = 17.6 deg/sector)
Phase3: Step 3: Angle 318 degree = 317+1 (SPR 21.0 = 17.1 deg/sector)
Phase3: Step 4: Angle 211 degree = 206+5 (SPR 21.5 = 16.7 deg/sector)
Phase3: Step 5: Angle 241 degree = 216+25 (SPR 22.0 = 16.3 deg/sector)
Phase3: Step 6: Angle 315 degree = 320-5 (SPR 22.5 = 16.0 deg/sector)

After (Sony drive)
Phase3: Step 1: Angle 278 degree = 287-9 (SPR 19.9 = 18.1 deg/sector)
Phase3: Step 2: Angle 294 degree = 297-3 (SPR 20.4 = 17.6 deg/sector)
Phase3: Step 3: Angle 302 degree = 317-15 (SPR 21.0 = 17.2 deg/sector)
Phase3: Step 4: Angle 202 degree = 206-4 (SPR 21.5 = 16.8 deg/sector)
Phase3: Step 5: Angle 232 degree = 216+16 (SPR 22.0 = 16.4 deg/sector)
Phase3: Step 6: Angle 324 degree = 320+4 (SPR 22.4 = 16.0 deg/sector)

Again the patched copy works.

Fujifilm 1x-48x CD-R, 700M (Prodisc 97m32s19f)

Before (Plextor drive)
Phase3: Step 1: Angle 261 degree = 287-26 (SPR 19.6 = 18.3 deg/sector)
Phase3: Step 2: Angle 291 degree = 297-6 (SPR 20.1 = 17.9 deg/sector)
Phase3: Step 3: Angle 323 degree = 317+6 (SPR 20.7 = 17.4 deg/sector)
Phase3: Step 4: Angle 194 degree = 206-12 (SPR 21.2 = 17.0 deg/sector)
Phase3: Step 5: Angle 201 degree = 216-15 (SPR 21.7 = 16.6 deg/sector)
Phase3: Step 6: Angle 297 degree = 320-23 (SPR 22.1 = 16.3 deg/sector)

Before (Sony drive)
Phase3: Step 1: Angle 243 degree = 287-44 (SPR 19.6 = 18.3 deg/sector)
Phase3: Step 2: Angle 268 degree = 297-29 (SPR 20.2 = 17.9 deg/sector)
Phase3: Step 3: Angle 295 degree = 317-22 (SPR 20.7 = 17.4 deg/sector)
Phase3: Step 4: Angle 169 degree = 206-37 (SPR 21.2 = 17.0 deg/sector)
Phase3: Step 5: Angle 186 degree = 216-30 (SPR 21.6 = 16.6 deg/sector)
Phase3: Step 6: Angle 292 degree = 320-28 (SPR 22.1 = 16.3 deg/sector)

As we can see, the numbers are within the ballpark as they are. So we try this original copy to see if it works, and indeed it does; no need to patch.

EDIT: Sorry, haven't seen your link (but why explain the whole procedure twice ?)

After how many CD's does topology of a certain brand change ?

Oh that's simply to put it in a visible place. Not everyone may have the patience of sifting through 400+ posts in a thread without any obvious keywords to go with (or those that would return little more than unanswered questions and shot-in-the-dark suggestions). Especially that the thread discusses a bunch of other things. Initially, I had to copy that particular title, and it took me quite a while to even find that topic, let alone study it. So I figured I'd share the results, so people with analogous problems at hand would have a quicker start. Thanks for adding the link referencing that particular post, as it's more or less a translation of the SFCopy manual by the author.

Not sure how many exactly, but >5. Also, I've tried two brands (Philips and TDK) which have the same media type (but of course from completely different batches). The differences in the angles were negligible -- no more than 8 degrees. I didn't do a t-test to see if this is attributable to chance, but I think it's safe to say that same media ID => same topology. For the same recorder and write settings, of course.

About 10-15 on cakes. It is not needed to say, different mill packs, different angles. Angles will be passed if SOME angles goes of a maximun of 30º (not in first angle of course). Critical angles are first and second one.

Thanks for this. I successfully backed up POPT2T using this method. I used generic CD-RW media with a Lite-On writer (SOHW-1633S) and reader (XJ-HD166S). I needed to run sfcdvd on my writer rather than the reader; the reader produced values that did not work.

Now all we need is a guide for coming up with those mysterious sector values on lines 3 & 4.

Pls post file .CFG for X3 Reunion version 6 cd

i have the dvd vercion of prince of percia . id did wath you say in dvd method
the game work perfectly in first timte but after when i tried to lunch game my computer restar automaticly .
some one can help me please.

Heard others say that happened; probably has to do with SF drivers trying to patch Windows FS drivers, or DT trying to patch them back so that emulation works. Instead of DT, try using Alcohol virtual drive with sfcure (there are usage instructions posted in multiple topics), maybe it works better.

Alternatively, try updating the StarForce drivers, for instance, by installing a game with later version of the protection (3.7), if you have such. You can uninstall such game later, this should still leave updated SF drivers in the system. Maybe there is a more straightforward way of doing this.

Was this determined simply by experimentation, or by actually disassembling the StarForce libraries? Because I had an instance where the first angle was almost always >30 deg off whenever I tried, but the check would pass and would not take long at all. Besides, 30 deg seems too small, given that I often got angles to differ by some 20 deg on the same CD in two different readers -- and the error was systematic more than random.

With CD-RW, you probably hit a CLV mode; I don't think sfcdvd can properly adjust for increased seek times in CLV so the angle measurements will not be correct.

do u know the sites where we could find cfg files to defeat SF?

or only solution is resolving the content of disc acceses log ?

Just to ask a wierd question, would two CDs of the same brand but with different maximum speeds(say one is a Sony 48x and the other one a Sony 52x) would be different Media Types?

Sorry i try to make starforce copy using sfcopy, but i really don't understand (i know it's for advanced user, but i want to learn it)


Originally Posted by Terramex
@cuiviemen, maximino72

28 00 00 02 b9 60 00 00 01 00 178528 (-18) -(-18)-1=17=spr1
28 00 00 02 ba 00 00 00 01 00 178688 (+160)
28 00 00 02 f8 00 00 00 01 00 194560 (+15872)
28 00 00 03 36 00 00 00 01 00 210432 (+15872)
28 00 00 03 74 00 00 00 01 00 226304 (+15872)
28 00 00 03 b2 00 00 00 01 00 242176 (+15872)
28 00 00 03 f0 00 00 00 01 00 258048 (+15872)
28 00 00 04 2e 00 00 00 01 00 273920 (+15872)
28 00 00 04 2d fe 00 00 01 00 273918 (-2)
28 00 00 04 2d fb 00 00 01 00 273915 (-3)
28 00 00 04 2d f6 00 00 01 00 273910 (-5)
28 00 00 04 2d f0 00 00 01 00 273904 (-6)
28 00 00 04 2d e9 00 00 01 00 273897 (-7)
28 00 00 04 2d e0 00 00 01 00 273888 (-9)
28 00 00 04 2d d6 00 00 01 00 273878 (-10)
28 00 00 04 2d cb 00 00 01 00 273867 (-11)
28 00 00 04 2d bf 00 00 01 00 273855 (-12)
28 00 00 04 2d b2 00 00 01 00 273842 (-13)
28 00 00 04 2d a3 00 00 01 00 273827 (-15)
28 00 00 04 2d 93 00 00 01 00 273811 (-16)
28 00 00 04 2d 82 00 00 01 00 273794 (-17)
28 00 00 04 2d 6f 00 00 01 00 273775 (-19)
28 00 00 04 2d 5b 00 00 01 00 273755 (-20)
28 00 00 04 2d 46 00 00 01 00 273734 (-21) -(-21)-1=20=spr2

->
178688 17
95234 20

I don't know how to get that blue number, and where 95234 come from? anyone who know about this, please help!

One more question, does every media has each config? how can i find it? thanks a lot!!!

So there is not a way to generate a cfg file direct from the CD-KEY ?
All info. must be rendered by the game itself ?

When we run the game to obtain the info., will it be better to use a full installed game? I mean if some of the normal data such as movies are read from CD, will it interferes the sfcfg.exe ?

Doest this work for the newer starforce protections?