Strange error on burnt Disc

[mitsubishi]

Hey guys, wondered if you know what happened here. I just burnt some of my avi encodes off to a DVD, but got this error on verificationSectors from 2116068 to 2116095 on disc are different to source
And I've never seen it before. So I analysed the original to the burnt file and got this.
http://img176.imageshack.us/img176/8168/startmi2.png (http://imageshack.us)

http://img84.imageshack.us/img84/7898/endbh8.png (http://imageshack.us)
As you can see the burnt file dropped 2 'FF' bytes, it is then 2 bytes offset right until the end of a 64k boundary when it adds '11 46' and is then back in sync.

Not sure what to make of this..

Who's fault do you think this is, RAM, windows, HD, Nero or something else? If it was a pure RAM error I don't get why it would offset like this, unless somehow the wrong memory address was retrieved by 2 bytes.

[Sharktooth]

it could even be a media problem or the laser didnt align properly...

[mitsubishi]

it could even be a media problem or the laser didnt align properly...

I don't think its the media as it reads OK and passes a disc scan, can't do a full PI scan as my only drive which does this is a lite-on (really must get myself a Benq) and thats not in. I actually thought I was using my LG drive, but thats apparently still in my old box, so I'm using my Pioneer 110D which has been caned in it's life time. Forgot to add the drive as a possible cause. But wouldn't the corruption have to occur before reaching the drive or it would fail checksum and I'd get read errors?

[Sharktooth]

CRC32 is not infallible... there are rare cases the CRC value stay the same while data is corrupted.

[setarip_old]

@mitsubishi

You might want to try copying the hard drive version of the .AVI to a different hard drive, or a different partition, or a different area on the same hard drive (without deleting the original) - and then burn it.

You might also want to try using blank media that's not from the same package as the problematic burn...

[mitsubishi]

Well I burnt it again and was without problem. This and the first go were the last 2 discs in that particular pack. I think it must have been some kind of random error, but I will do a memcheck tomorrow to be on the safe side. My games don't crash though, so I suspect software more. My HD though, yeah I've ad problems with it, it likes to shutdown every now and then which requires a full power cycle to get it to come back, particularly annoying as when it BSODs and I just reset my MB sees only my other drive and tries to boot off it, meaning I have to go and put the order right in the BIOS. Apparently I'm not the only one with this issue with Maxtor Satas.

[setarip_old]

My HD though, yeah I've ad problems with itThen this is the likely source of this specific anomaly...

[mitsubishi]

Well the problem I've encountered thus far is only that as described, I havn't noticed any read errors of this nature. The fact that it is offset, but then returns at a 64k point suggests that it is some kind of seek/address error which was contained in a fragmentation point. Coming back in sync I would have thought it was just in one operation, read buffer, disc sector I don't know. When I finally get this HD cleared of data it needs RMAing anyway though.