Discrepancy between DiscImageCreator and redumper for SecuROM 4 disc

[maxz]

Recently I began dumping my discs with both DiscImageCreator and redumper.

On the disc for Empire Earth, I discovered a discrepancy regarding the errors. More specifically regarding a single mode 1 sector among the mode 2 sectors. Both tools detect that correctly. But the redump.org entry and DiscImageCreator state it as an error and redumper does not.

So what is the case here? Is it an oversight in redumper? Or will the redump.org policy going forward be to not record these sectors as erroneous anymore?

Logs: https://archive.org/download/redump.org-logs/EEARTH.zip

DiscImageCreator's EEARTH.img_EccEdc.txt:

[ERROR] Number of sector(s) where mode is invalid: 1
        Sector: 290471, 
Total errors: 1
Total warnings: 0

Redumper's EEARTH.log:

CD-ROM [EEARTH.bin]:
  sectors count: 290475
  mode1 sectors: 1
  mode2 (form 1) sectors: 290474

  REDUMP.ORG errors: 0

[superg]

So that was an ongoing discussion that we had @redump.org. Historically, Errors count field @redump is the output of edccchk.exe util. That utility goes over all sectors linearly and checks for 2 things:

1. ECC/EDC errors
2. XA subheader mismatch

redumper does exactly the same.

Now, SecuROM discs have 1 sector with flipped mode close to an end of the data track. Otherwise that sector has matching ECC/EDC e.g. sector, if viewed independently, is correct but positionally we can say that it's non compliant. However I haven't seen that mixing sector modes in scope of the same data track is illegal in rainbow books but I didn't spend too much time on this. For some other systems (PCE if I remember correctly) we also have mixed modes in track gaps but we don't mark them as errors.

This one error was useful in early DIC development days where all SecuROM schemes weren't implemented and it was a hint that there might be an undiscovered scheme. Today this is a non issue and on redumper side when scanning for protection there still is a warning if mode is flipped but SecuROM scheme wasn't detected. My opinion is that this mode flip doesn't belong to errors.

[maxz]

Ok, it certainly makes sense to view it this way. My exact verdict would depend on the format's specification.

It still leaves me non the wiser what the correct error count for submissions would be in this case, especially since the redump submission form still states Errors count (edccchk v1.26) and therefor specifically refers to that point of view. I guess I will try using the redumper error count as basis for my submissions and then go by the moderator response.

Do you maybe have a link to the discussion on the forum? I tried finding it but could not. Did it come to any conclusion regarding which of the viewpoints to prefer for submissions?

[superg]

That was discussed privately and there is no 100% consensus at this point. Submit with either error count and hopefully moderator will bring this up again.