Safedisk 2 and fermwares



here is some info from the developer of ddump

First, let me explain why it is so difficult to make a fully working backup of a SafeDisc2-protected CD...

On CDs protected with "normal" SafeDisc, you have an area of about 10000 sectors where there are always unreadable sectors followed by some readable and so on. This grouping changes from CD to CD and is it's typical SafeDisc fingerprint that is checked by the protection-code inside the executable. This fingerprint can simply be re-produced by writing the image in RAW mode.
SafeDisc2 now also has such an area but also some groups of so-called "weak" sectors. Why are they called "weak"? Well, it seems that these sectors can the read without problems form the original CD, but when trying to read them again from a copy they suddenly seem to be damaged and sometimes can only be read with the second or third attempt - and sometimes not at all! How can this be??
To understand this, you must know that all data is "scrambled" before it is written. Scrambling in this case means that every byte is XOR'ed with a certain value that changes from byte to byte. This is done to prevent so-called "uniform bit-patterns". But if you scramble such a weak sector, exactly that case occurs that should be prevented: You get an uniform bit-pattern like this one: 0x58, 0x59, 0x58, 0x59, ...
So why are these uniform patterns a problem? For some technical reason which would take too long to explain here, all bytes are encoded into 14 bit each right before they are sent through the laser and burned on CD. This encoder is called "EFM encoder". This EFM-encoder may (although it shouldn't if it has been implemented error-free) have problems with uniform bit-patterns that may result in damaged data.
It seems that the writers of certain manufacturers have big problems with this. The burned weak sectors can only be read in the burner itself or not at all...

What's the conclusion of this?

First thought is, that this is a real hardware problem. Yep, sounds quite logical... but only quite!
If it really would be a hardware-problem, why can these sectors then be read in the burner? If the EFM-encoder would have messed the data, you won't be able to read this sector with any drive - including the writer itself.
Also there's something very strange regarding the firmware releases and the approach of SafeDisc2: It seems that this "bug" has been implemented with full intense in newer firmware-releases of some major CDR manufacturers! Affected by this are as far as I know Plextor, Sony and Teac. There seems to be an unholy "marriage" between Macrovision (the developers of SafeDisc2) and these manufacturers...
How I can prove this claim? Well, I did some tests here with my Plexwriter 1210TA: I downgraded the firmware to version 1.01 (which I luckily kept on my HD) and suddenly the writer was able to re-produce these weak sectors like a charm!
Not enough proof? Ok, here we go: I tested with various firmware releases for this writer and figured out, that firmware revision 1.04 seems to be the last "good" release. What a chance that Plextor has removed all revisions below 1.05 from their web page...
And another proof: Newer revisions of this writer (recognizable by a TLA number other than 0000) have been protected against such a firmware downgrade. You won't be able to flash a revision lower than 1.05 to the internal eeprom...
Somebody now discovered that the old firmware revisions still seem to exist on Plextor's servers. But somehow the file-size has changed. When a normal firmware-update package had about 530 kb, these packages only have about 440kb now... is this pure chance or did the guys from Plextor remove/change some stuff in these old revisions? Strange, strange...

So what can we do against this?

Try to get an "original" firmware-package rather from somebody who has stored all old revisions somewhere than from the manufacturers website and flash it to your drive. If you're lucky and the downgrade works you may be able to copy SafeDisc2 now without problems...
But also if this works this can only be an intermediate solution! It can't be the golden way that the user has to downgrade his firmware to a lower revision that may contain serious bugs and eventually has less functionality!
Let's show companies like Plextor that they can't treat us like that! Write them email about what you think about this and tell them that this would be the last drive you have purchased from them if they don't change their manners...

Where can I find more information on this topic?

Have a look at the SafeDisc 2 Support Forum and at CDRSoft.netoper of ddump.


Well, I’ve tested at least 10 different PX-W1210A(FROM TLA #0000 to TLA #0307) writers and none of them was able to produce these bits correctly with any firmware versions. The backup did sometimes work in very few drives(pioneer DVD-ROM was one of them) but not in many drives and in most cases not in any drives.

So to me this is just bullshit.


Yes, OC Freak, this is a recycled article from about a year ago if I remember correctly or at least very similar to one that was making the rounds. It has been refuted several times (at least the article that made the similar claims). There appears to be no substance to the author’s assertions. Several people have taken it on and there used to be a series of posts about trials done using the plex and different TLAs which clearly showed that the TLAs did not affect the plex’s ability or inability to do SD1 or SD2.

My own limited burns with different plexs and TLAs agree totally with your findings.