UDF anchor points

[tartak]

First of, I'd like to thank LUK! for his outstanding program. Have used it to burn more than 1000 DVD-Video, about half - DVD-9, all without a glitch!

To spread the word, I've translated LUK!'s guide into Russian, added some additional info found in this forum, and posted the whole thing on some of the biggest russian forums, about half a year ago. It's been spreading around ru-net ever since.

 

Well, I had no choice but become a resident "expert" and now they hold me responsible for any problem - as long as ImgBurn is somehow involved. After some discussions about what makes a DVD conforming, I got an idea to run a few ImgBurn'ed disks through Philips DVD-Video Verifier. The check did not catch any problems I would consider of much importance. Yet one thing was curious, I thought I'd better ask here.

 

Here is from the log:

>>> [UDF] ERROR 5213  (ref. UDF 2) :
#anchor points found is: 1, two anchor points were expected
at sector/block 256, byte 16 bit 0

And the explanation in the manual:

============================

>>> [DVD] ERROR 5213 (ref. DVD-2 2.3, UDF 2 and ECMA 3/8.4.2.1) :

ERR_ANCHOR_POINTS_NOT_TWO

According to [uDF] 2, the number of recorded AVDP must be exactly two. They must be placed at two of the following three places: logical sector number 256, N-256, or N, where N is the last addressable sector of a volume.

According to [DVD] 2.3, the AVDP must be recorded at 256 and N, where N is again the last logical sector number.

============================

 

As I understand, the disk is no longer usable if there is a defect at the anchor point, so it is duplicated for protection. ImgBurn seems to produce only one anchor point - is it a problem? What's the deal with these anchors, how do they work?

Edited January 25, 2008 by tartak

[LIGHTNING UK!]

It writes 2, one at sector 256, the other in the last sector on the disc.

 

I've checked 2.4.0.0 with the philips verifer some time ago and it passed everything with flying colours (after a few modifications) - and that's using UDF 1.02 -> 2.60.

 

That's not to say 2.3.2.0 didn't write 2 anchor points, it did. All I can think that happened here is the 'end of the disc' changed between ImgBurn building the ISO and you testing the disc.

[tartak]

Well, I did not write an ISO, I wrote DVD Video files directly to a DVD-9 in the Build mode. It should not make any difference, though, right? Could it be because of double-layer media? And, of course, I don't have 2.4.0, I wish I did It's not that I'm unhappy with 2.3.2 in any way - it's just they ask me questions, a lot.

 

Just in case - I used Verbatim MKM and BenQ-1640. Today I repeated the test, different DVD, different computer, burner - Samsung SH-S203N, media +R DL Verbatim MKM. Exactly same errors. Here is the log from Verifier (1.5.0/4.6.0, released April 11, 2003) - I skipped all mpeg and DVD verification:

 

********************************************************************

*

* DVD stream input type = Disc(image)

*

* Decoding skipped for MPEG levels :

* TRS stream

* PES stream

* Video Sequence

* Macroblock

* Audio Sequence

* Audio multi-channel data

* Dolby AC-3 audio

* Padding Stream

* Private data

* DVD levels :

*

* Verification disabled for MPEG levels :

* PSI-tables

* Pack header

* System header

* DVD levels :

*

********************************************************************

 

>>> [iSO9660] ERROR 5518 (ref. ISO 8.4.26.1) :

Date Time format has an incorrect Year attribute:

The value must be in the range 1..9999, but the actual value is 0000

at sector/block 16, byte 847 bit 0

 

>>> [iSO9660] ERROR 5518 (ref. ISO 8.4.26.1) :

Date Time format has an incorrect Month attribute:

The value must be in the range 1..12, but the actual value is 00

at sector/block 16, byte 851 bit 0

 

>>> [iSO9660] ERROR 5518 (ref. ISO 8.4.26.1) :

Date Time format has an incorrect Day attribute:

The value must be in the range 1..31, but the actual value is 00

at sector/block 16, byte 853 bit 0

 

[skipped rest of ERRORs 5518]

 

>>> [uDF] ERROR 5213 (ref. UDF 2) :

#anchor points found is: 1, two anchor points were expected

at sector/block 256, byte 16 bit 0

 

>>> [uDF] ERROR 5064 (ref. ECMA 1/7.4, UDF 2.1.5) :

Regid identifier is '*ImgBurnInfo';

It must be as specified in UDF Appendix 6.2.

at sector/block 33, byte 21 bit 0

 

>>> [uDF] ERROR 5231 (ref. UDF 2.2.4, 2.2.7.2.2) :

LogicalVolumeIdentifier of IUVD (UGLY_SWANS) does not equal (null) (LVD)

at sector/block 33, byte 52 bit 0

 

>>> [uDF] ERROR 5232 (ref. UDF 2.2.7.2.2) :

No IUVD has Implementation Id of *UDF LV Info

at sector/block 33, byte 52 bit 0

 

>>> [uDF] ERROR 5064 (ref. ECMA 1/7.4, UDF 2.1.5) :

Regid identifier is '*ImgBurnInfo';

It must be *UDF LV Info.

at sector/block 33, byte 20 bit 0

 

[skipped rest of ERRORs 5064, 5231, 5232]

 

>>> [uDF] ERROR 5064 (ref. ECMA 1/7.4, UDF 2.1.5) :

Regid identifier is '*ImgBurnInfo';

It must be *UDF LV Info.

at sector/block 49, byte 20 bit 0

 

>>> [uDF] ERROR 5355 (ref. UDF 3.3.3.4) :

Unique Id 0x0000000000000002 is not 0 for root directory or between 1 and 15

at sector/block 264, byte 160 bit 0

 

[skipped rest of ERRORs 5355]

 

>>> [iSO9660] ERROR 5530 (ref. DVD A.14) :

The System Use field must contain the Copyright Management Information,

and no more information is allowed

at sector/block 259, byte 35 bit 0

 

[skipped rest of ERRORs 5530]

 

************************************************************

*

* ERROR SUMMARY :

*

* No Informations

* No Recommendation violations

* No Oddities

* No Warnings

* No Syntax errors

* 74 Errors

* No System errors

*

************************************************************

Edited January 26, 2008 by tartak

[LIGHTNING UK!]

Have you tried with a more recent verifier?

 

http://www.hitech-projects.com/udf/download.html

 

As I say though, I have tweaked some stuff but the two anchor points have ALWAYS been written.

[tartak]

I'm very happy to report that the same disk made it through Philips UDF Verifier with no problems. Well, it did detect the same "errors" with ImgBurn identifiers as DVD Verifier, but it's found both anchor points:

	Reading Volume Information
256	read block
AVDP at 256	(MVDS: 32, RVDS: 48)
First Tag Serial Number: 0
Note: Tag Serial Number 0, no disaster recovery support,
-		  ECMA 3/7.2.5, 4/7.2.5, UDF 2.3.1.1.
- ==>	Message printed once, ignored from now.
3454111	read block
No AVDP at N-256
3454367	read block
AVDP at N	(MVDS: 32, RVDS: 48)

Number of AVDPs: 2, AVDPs at  256,  N

I got the most recent version of udf verifier (1.5 r6, more recent than at hitech-projects.com) from Philips, UDF Verifier Software, one simply needs to register there. They do say in some of the release notes that they improved AVDP detection.

I'm shocked that Philips got it right in their free udf verifier and screwed up in dvd-video verifier, which is far from being free. For a second, I thought it might have something to do with DL media, so I got a dvd-5 through dvd-video verifier - same result, only one AVDP found.

 

Is there any more recent software than philips dvd-video verifier, that could still check the whole thing? Something more or less reasonable, of course, not the $10K+ Eclipse and MEI stuff the replicators use...