Dual Layer Title Set With Extra Directory

[Handler]

Hi,

 

I did search, but couldn't find what I was looking for. I've always included an EXTRAS directory with the VIDEO_TS and AUDIO_TS directories in the root of a single layer disc. Up until now I've not had a problem but that was single layer, and when I tried to do the same thing with my first dual layer disc I ran into a Layer Break problem.

 

As near as I can figure, the contents of the AUDIO_TS directory (empty) and the EXTRAS directory (100 MBytes) gets added alphabetically, followed by the VIDEO_TS directory and that seems to throw off the Layer Break by pushing it further into the disc. If I move the EXTRAS directory out, then add just the AUDIO_TS and VIDEO_TS directories then I'm presented with the Layer Break window, otherwise I get 'The End of The World' problem with the addition of the EXTRAS directory.

 

I tried to add just the AUDIO_TS and VIDEO_TS directories first, then after picking a Layer Break cell, add the EXTRAS directory. The hope there was to get the EXTRAS directory to be burned on the remaining free area of the second layer, after the Title Set. But that seems to run into directory structure issues with the Test.ISO I've created placing the VIDEO_TS in a child directory off root and that's no good.

 

Is it possible, and if so, could someone please explain how one goes about adding an additional directory to a Title Set with Dual Layer media (in this case Verbatim MKM-001-00 DVD+R DL) and the latest version of ImgBurn?

 

Thanks.

Edited September 23, 2008 by Handler

[LIGHTNING UK!]

Change 'Folder Content Location' to 'Start' and try again.

 

(Tools -> Settings -> Build -> DVD Video)

[Handler]

Change 'Folder Content Location' to 'Start' and try again.

 

(Tools -> Settings -> Build -> DVD Video)

 

Wow! That was exactly what was needed to get the Title Set to load at the beginning. Thank you.

 

I went ahead and created another test.ISO and while everything directory structure wise is correct, the CRC value of several, not all, of the IFO files are no longer identical to the originals. This is quite puzzling considering that the main purpose of the EXTRAS directory is to store PAR2 files should the disc become unreadable in the future. The files that do not match the originals do not show a different modification date or time. This occurrence has absolutely nothing what so ever to do with the EXTRAS directory. On the second test2.ISO I completely eliminated the EXTRAS directory and once again several of the IFO files had been internally modified.

 

I should point out that the VTS_01_0.IFO file in the Title Set already included a Layer Break. When creating the test ISO I load the Title Set and the Layer Break window opens. Not sure why it's necessary to make the same Layer Break selection twice, but each time that the Layer Break window opens I select the same 'gold' Layer Break position that is already present in the IFO file (VTS_01, PGC: 1, Chapter: 13, Cell: 14).

 

After the EXTRAS-less test2.ISO was created, I then use DT to mount the image. Using a binary comparing program, it shows that out of the eight IFO files, four of them are different (ie: VIDEO_TS.IFO, VTS_03_0.IFO, VTS_05_0.IFO and VTS_08_0.IFO). Loading the mounted test2.ISO VIDEO_TS.IFO file into IFOEdit results in 10 errors in the VMG_PPT_SRPT sector addresses table in 04, 05, 06, 07 and 08. Also IFOEdit tried to correct VMGI_MAT sector addresses in 03, 05, and 08. None of these errors are present in the original Title Set when loaded into IFOEdit.

 

I don't know what this all means, but from my stand point, I would think that since the Title Set is now being added to the start of the disc and the Title Set's own Layer Break is being used, that the test2.ISO files should be identical to the original files. Can someone please explain why there are so many differences and why these differences result in many errors in the Title Set's structure?

 

Thank you ever so much.

[LIGHTNING UK!]

ImgBurn has to correct certain pointers when it burns to disc due to the layout and positioning of the files.

 

The layout isn't the same when you look at the files as files... but then DVD players don't do that, they go by the pointers.

[Handler]

Hi,

 

I don't have any experience with Dual Layer, only Single Layer discs. Using programs such as CDCheck (which also works with DVDs), I've never seen a Single Layer disc that didn't match exactly byte for byte unless the burn wasn't 100 percent successful. A DVD for the most part is just a data burn. So having discs not identical to their source files is new for me and I suppose it must be something specific to Dual Layer discs. However, it throws a monkey wrench into using PAR2 files as a means of securing future data integrity. At least the VOB files are identical.

 

As a result of ImgBurn having to "correct certain pointers" "due to the layout and positioning of the files", IFOEdit is now reporting ~13 address errors. Is it possible that ImgBurn isn't doing a comprehensive job of correcting all the necessary pointer changes prior to the creation of the image?

 

Thanks.

[blutach]

They are not "address errors". All that is being done is that the pointers to the start of titlesets are being adjusted for their actual physical position on the disc, given padding, your selected layer break, etc. IfoEdit is trying to set them back to a situation where the disc is crowded together one sector after another (a very dangerous and incorrect practice in a DL burn, which Ner0 uses - or at least used to use).

 

Remember, a player is dumb - really dumb. It knows nothing of filenames etc. When a disc is inserted, it goes around looking for DVDVIDEO_VMG which indicates the start of VIDEO_TS.IFO. From various internal pointers in that file, it determines where the PTT_SRPT table starts. This table contains, amongst other things, the start sectors of the titlesets. From there, each titleset has various internal internal pointers. If a VTS is padded, (as it usually must be to honour the rules of DL recording), some of these VTS internal pointers get shifted a bit, too. So, "correcting" these things with IfoEdit is actually "breaking" them. You get it?

 

There is nothing to be worried about. If you construct a disc that needs no padding (i.e. the LB will always fall on an ECC boundary - sector number divisible by 16), then it is unlikely you will see these variances again. If the LB does not fall on an ECC boundary, you can see that "crowding the sectors together" will be incorrect. So, you pad a bit (i.e. skip over some sectors) and change the VTS start sector references.

 

Is it possible that ImgBurn isn't doing a comprehensive job of correcting all the necessary pointer changes prior to the creation of the image?

No. I have never had a DL disc not play properly. Stop worrying.

 

Regards

[Handler]

They are not "address errors".

Just to clarify, IFOEdit did say "Corrected sector addresses in table [VMG_PPT_SRPT ]" and there were ten of those. Followed by three "Corrected sector addresses in table [VMGI_MAT]. Whether it was right or wrong of IFOEdit, IFOEdit found those sector "addresses" to be in "error". I used a slightly, but already explained, shorting of the terminology that IFOEdit used so that others would know, if they are familar with IFOEdit, what I was referencing.

 

All that is being done is that the pointers to the start of titlesets are being adjusted for their actual physical position on the disc, given padding, your selected layer break, etc. IfoEdit is trying to set them back to a situation where the disc is crowded together one sector after another (a very dangerous and incorrect practice in a DL burn, which Ner0 uses - or at least used to use).

That makes sense.

 

Unfortunately, that also suggests that there is no way to add PAR2 files to an EXTRAS directory on a Dual Layer disc without them immediately finding a checksum error as soon as the disc is burned. #39;( <sniff> Is there anyway to add PAR2 files to a Dual Layer disc without that happening?

 

Remember, a player is dumb - really dumb. It knows nothing of filenames etc. When a disc is inserted, it goes around looking for DVDVIDEO_VMG which indicates the start of VIDEO_TS.IFO. From various internal pointers in that file, it determines where the PTT_SRPT table starts. This table contains, amongst other things, the start sectors of the titlesets. From there, each titleset has various internal internal pointers. If a VTS is padded, (as it usually must be to honour the rules of DL recording), some of these VTS internal pointers get shifted a bit, too. So, "correcting" these things with IfoEdit is actually "breaking" them. You get it?

Yes.

 

There is nothing to be worried about. If you construct a disc that needs no padding (i.e. the LB will always fall on an ECC boundary - sector number divisible by 16), then it is unlikely you will see these variances again. If the LB does not fall on an ECC boundary, you can see that "crowding the sectors together" will be incorrect. So, you pad a bit (i.e. skip over some sectors) and change the VTS start sector references.

Got it.

 

Is it possible that ImgBurn isn't doing a comprehensive job of correcting all the necessary pointer changes prior to the creation of the image?

No. I have never had a DL disc not play properly. Stop worrying.

Quid, Me Vexari?

 

To tell you the truth, money is very tight and these Dual Layer discs are not cheep. I'm use to working on Single Layer discs, because Dual Layer discs are so different, I just wanted to do everything I could to avoid burning my first Dual Layer coaster, on my first Dual Layer burn.

[LIGHTNING UK!]

The only way to do it is to forget the extras folder for a second, make an image, build par2 files based on the files inside the image (mount it in DAEMON Tools or something) and then add the extras folder.

 

You might find you're better off just par2'ing the image file and burning the par2's from several discs on a disc of their own (and burning 2 copies?!).

[blutach]

Yeah - if you wanna make checksums, make an ISO first.

 

Whether it was right or wrong of IFOEdit, IFOEdit found those sector "addresses" to be in "error".

It was both right and wrong. It was right if you wanna pack the disc closely together sector by sector. But, as I've explained, this is the wrong thing to do.

 

Remember IfoEdit has not been updated in years. Much better is PgcEdit, which at least allows for 32k buffers.

 

Good luck in your burning. Remember, for DL burns, ONLY use Verbatim, +R, 2.4x, MKM-001-00, Made in Singapore.

 

Quid, Me Vexari?

BWAHAHAHAHAHAA

 

Regards

[Handler]

Hi,

 

Yes, the creation of the ISO first, then creating the PAR2 files from the mounted image and then copying the newly created PAR2 files back to the source Title Set in the form of a EXTRAS directory would definitely be the way to go since no matter how many times the ISO is created, the IFO/BUP files should always be adjusted identically.

 

To give you an idea how long it's taken me to burn my first Dual Layer disc, the Singapore made Verbatim DVD+R DL (MKM-001-00) discs were purchased in I believe 2006 (box says 2005 on the back).

 

Big thanks to LIGHTNING UK and blutach for all your time on this subject. Hopefully this thread will continue to help others who are trying to do the same thing.

 

Thanks