#1 Burning Software

I'm having an intermittent problem when writing DVD-R media. Specs:

Laptop
WinXP
LG GSA-4167B writer, firmware version DL11 (Not DL13, the newest, but I cannot currently flash it. Looks like DL11 solved some problems with unreliable writing.)
The writer is in a USB external enclosure, USB2.
Starburn Version 6.2.24
Delphi 7
I use wnaspi32.dll, although I think I read somewhere it's not needed with the newest sdk version (not sure)

My project writes about 430 meg in trackatonce, then closes the session and destroys the burner object. After the next 400 meg is prepared, I create the burner object, import the track and burn the extra 430 meg. This is repeated until the disk is full. ( 10 sessions per DVD-R).

I reduced the speed on this drive from the rated 16-speed to 8-speed after getting the error. This seemed to improve the problem, but not remove it. The last time I got the problem, I was browsing the web and downloading while the writing was going on.

In the post below I will post the high-level logfile I generated. This is similar to the trackatoncefromtree sample output, since I adapted a lot of that code.

Below that, I will post the Starburn debug output, as captured by the Delphi event log.

Note that I did remove some repetitive getmediatraystatus events from the debug output (I poll this repeatedly while waiting for a disc to be inserted into the drive at the beginning of the sequence.)

Just before this error occurred, I burned 2 discs without problems.

I think the speed might still be too high, but I have no idea which bit in the debug log will show me that for sure.

Please help,

Johan

1) Get latest version of the StarBurn. We don't support anything earlier then we publish on the site. Quite a lot changes between releases and it's a pain to backtrack and fix the same issues again and again in the old code.

2) I've deleted your logs as they were both mangled and truncated. Send complete ones to the support@rocketdivision.com and we'd be happy to help you

Thanks!

Oops, sorry, didn't know you had a newer version out. I'll get it running...

OK, waiting for your feedback.

The version that I'm using is actually version 6.4.10. I had it mixed up because of the following lines in the errorlog:

StarBurn.pas : StarBurn.dll version : 0x20060410

And then a bit later:

StarBurn CD/DVD Burning, Grabbing and Mastering Toolkit for Windows 95/98/Me/NT/2000/XP/2003 Version 6.2.24 ( 0x20060410 ) Copyright (c) Rocket Division Software 2001-2006. All rights reserved.

and a few lines further:

StarWave Media Streaming Toolkit for Windows 95/98/Me/NT/2000/XP/2003 Version 6.2.24 ( 0x20060224 ) Copyright (c) Rocket Division Software 2001-2006. All rights reserved.

So it seems your debug ID lines are incorrect in some way.

I'll send you the non-truncated output of the error logs now. Note that the 'mangling' that you see is the Delphi debugger that adds an identifier like 'Module load' or 'ODS' at the beginning of the line and the process number at the end. If you prefer, I'lll try to rather use dbgview or something like that.

I don't know how it happened but you're using old DLLs. Just grab latest StarBurn from the download section of our site.

I don't need Delphi mangled logs - it's pain working with them b/s of the broken formatting and extra info Delphi puts into it (and I don't need it). Please send me StarBurn.log our library generates itself.

Thanks!

Sorry, the library is up to date I was out of office when the guys released latest StarBurn so it came GOLD with old version Id strings. Sorry once again for blaming you - it's our fault

So I'm waiting for StarBurn logs (no Delphi debug console please!) from you to help with your issue.

Thanks!

No problems with the version mixup, I thought it might be something like that.

Okay, I'm trying to recreate the problem. I've set the logs to write directly to a file (and the previous file is deleted at program start), so if this happens again, I will have the correct info for you.

OK. A couple of ideas so far - slowing down the burn to avoid problems, running actove foreground process to increase the possibility to get burning problem means in 99% of the cases you've got buffer underrun problem. Since modern hardware has buffer underrun prevention I think you've just did not enable buffer underrun protection with SetBUP( TRUE ) before performing the burn operation... Please check your code!

Okay, I've managed to recreate the problem and sent you the log files. A couple of notes:

1. I Do activate BUP, as you will see from the starburn.log.

2. From the logs you will see that the error occured during the ninth track of the sixth disc that I was burning. Intermittent errors are the hardest to catch

3. I was browsing on Google Earth using Firefox while this error happened, so it could very well be that the machine was a bit overloaded. I did not record CPU or memory status.

4. The error seemed to happen during session closing. It might be that the actual error occured during the track burning session, but got picked up on closing the session.

5. In my own high-level log, I recorded all the write packet and bufer status callbacks, once per second, for the writing process. It looks like the buffer never dropped below 80%.

Hope this helps.

I'm starting to think that this is purely a 'trying to do too much at once' problem on my machine. If it was a bug in the SDK or my programming, it would be more consistent, as the process is the same for all the sessions.

Johan

I just thought of something:

Do I need to provide more space on the last session for the session closing? It looks like it was actually session 10 ( the last session on the disc) where the error ocurred.

I size the files so that they should fit on the disc with about 2 meg to spare at least. The line from the log for the last session:

File chain size in LBs 208764 should be OK for free LBs 209376

So I assumed it would be OK.

Let me check the logs, but it could happen we've seen this problem before. Under some OSes (Windows 2000) and some burners (Panasonic DVD-RAMs). For some reason reported disc space was too small for image. For all of the burning apps, not only StarBurn.

OK. Please give me a couple of hours

Johan,

I've verified the logs you've sent me and it seems like you're correct - reported number of free logical blocks is OK for actual user data but the drive cannot burn
final lead-out for the last session.

For a workaround skip burning the discs in the multi session mode "till the very end" as I don't see what can we do here...

Thanks!