Source of WS problem

vbimport

#1

After all, weak sectors are ordinary data sectors, so WS-based protections do not exploit any deviations from the standards to make it work. The question I don’t fully understand is what is responsible for not being able to write such sectors correctly. I guess it could be one of the following:

  1. CD-recorders don’t implement “merging bits” algorithm correctly, as required by the standard, which leads to improper pit pattern.

  2. Ecma-130 itself has a flaw. Though it makes a best effort to keep DSV around zero, it still allows for such a sequence which drives DSV unacceptably high.

  3. CD-readers technology has a flaw. It must read correctly high-DSV patterns though zero-DSV is preferred.

As I understand, the third reason is not the case, since pressed disks do contain the same weak sectors and are read correctly by all readers.
If my first guess is correct, then it could have some interesting legal implications, since all recorders don’t implement correctly there main function ? burning files. For example, I can sue a recorder producer for not being able to burn efmgame.iso file.


#2

1 and 3.

The drives wouldn’t have a problem reading if they could handle high DSV sequences, and drives wouldn’t have a problem writing if they didn’t have improper meging bit algorithms.

If my first guess is correct, then it could have some interesting legal implications, since all recorders don’t implement correctly there main function ? burning files. For example, I can sue a recorder producer for not being able to burn efmgame.iso file.

I don’t think that it is required for them to burn all types of data.


#3

I disagree with the last point Pegasus, ECMA states that a regular pattern conforms to the 3-7-11 rule (no less than 3 of the same bit, an average of 7 and no more than 11).

Recorders must be able to cope with any patterns that fall within this rule. Legally though, there is no obligation for writers to be able to cope with non-standard patterns which do not conform to this rule, as they have to be produced deliberately and don’t happen by accident.

It is also technologically difficult for a product to support such non-standard patterns, and as they don’t normally occur outside of copy-protection systems, you don’t have much ground to stand on.


#4

Originally posted by AtomicX
[B]I disagree with the last point Pegasus, ECMA states that a regular pattern conforms to the 3-7-11 rule (no less than 3 of the same bit, an average of 7 and no more than 11).

Recorders must be able to cope with any patterns that fall within this rule. Legally though, there is no obligation for writers to be able to cope with non-standard patterns which do not conform to this rule[/B]

The problem of weak sectors is not 3-7-11 rule. Merging bits algorithm assures the fulfilling of this rule on any input sequence. 3-7-11 rule restricts the average number of transitions between pits and lands. The real problem is DSV rule, which regulates the relative proportion of pits and lands. There are legal input sequences which result in high DSV even if the algotithm is applied correctly. On the other hand there are no requirements in the standard which restrict the legal values of DSV. Just a recomendation to keep it low.

So regular patterns on input are perfectly legal from the standard’s view. And since readers are able to read these patterns from pressed disks I conclude that recorders just fail to implement correctly merging bits algorithms. Am I wrong?


#5

There are legal input sequences which result in high DSV even if the algotithm is applied correctly.
The DSV can be high for a sequence of 12 bytes even if the rules are followed (for example 96 for the patterns 04B9 and 659A).
But this is the longest sequence of this kind; then there would be (irregular, not-high-DSV-causing) CIRC data.

To get an overall DSV of 0, the encoder would have to assure that the next chain of 12 bytes causes a DSV of -96 instead of +96 by adding one transition.

If you have any suggestion on how to implement this with a runtime of O(n), then post it :stuck_out_tongue:

When using “shall” in ECMA standards, replace it with “is required to be” in your mind.


#6

1 and 2. And note that on pressed cds you can
freely choose the merging bits.