Plex faulty cuesheet write

vbimport

#1

Hello folks.

As I realised there wasn’t a perfect program to burn CD+G cuesheets
(case here), I discovered a couple of faults while burning audio/data
cuesheets too, but this time the mistakes come in hardware’s design. :frowning:

I performed the tests with four Plextor drives and a couple of Pioneers,
they consist in imaging audio/data subchannels of a pressed disc and doing
an img+cue pair image of it aswell, then this last image is written in CD-RW
and finally a new subchannel extraction is made from this rewritable disc.

Results from comparation between first and second subchannel extractions
and a quick first pause and leadout comparation in CDTool reveal the next:

-Plextor PlexWriter Premium I and II:
Sony style start, incomplete leadout start flag (150th subcode block with
P = FFs is missing), Sony style pauses preceeded by a P = FFs subcode
block when length is equal or higher than 2 s (when gap is present).

-Plextor PX-716A and PX-760A:
Same as Premiums, but aditionally all P subcoding gets shifted one
block back (it’s in phase with Q subcoding, which is against the
specification since it must be delayed by one subcoding block).

-Pioneer DVR-104VA:
Sony pauses and… not a single subcoding block with P information! :eek:

-Pioneer DVR-A06U/Z:
Sony pauses and just like the Plextor DVD burners, all P activity gets
shifted one block back.


#2

I just want to clarify the disc’s start flavor
(Sony or Philips) depends on the software,
this time I used ImgBurn which writes Sony.


#3

[QUOTE=IpseDixit;2121030]I just want to clarify the disc’s start flavor
(Sony or Philips) depends on the software,
this time I used ImgBurn which writes Sony.[/QUOTE]
It depends also on the MMC write mode/block type/data form (for SAO) being used.

MMC has 4 write modes for CD writables:

  1. TAO
  2. SAO
  3. RAW
  4. Packet

TAO: this is cooked mode burning only, where drive’s hardware generates all subchannels, and leadin, 1st pregap, and leadout are fully generated by the drive, so those parts are not written by software.

SAO: most software use the cooked mode for this so it is like in TAO, but it has a data form byte that allows for subs to be sent by software and if supported by the drive the software has to send subchannel data and write that section. To complicate matters, this mode uses a MMC cuesheet and each section has a selectable data form flag (see page 278 in MMC2 draft). So it depends on that setting. If set for leadin and leadout then software will write them.

RAW: Software has to send subchannel data, and write leadin, gaps, leadout - no choice.

Packet: Only used for the floppy like format, and is cooked mode only like with TAO.

ImgBurn uses DAO/SAO which is SAO write mode with 1 session. But I have no idea what data form byte value is being used for the leadin and leadout. A query for LUK. :slight_smile:

*Cooked mode means only data part of main channel is sent to the drive.


#4

Burning an audio disc (12 tracks) with CD-TEXT…


SEND_CUE_SHEET

01 00 00 41  00 61 22 19
01 01 00 00  00 00 00 00
01 01 01 00  00 00 02 00
01 02 01 00  00 02 34 18
01 03 01 00  00 05 23 27
01 04 01 00  00 07 33 33
01 05 01 00  00 0a 28 48
01 06 01 00  00 0d 26 29
01 07 01 00  00 10 29 1b
01 08 01 00  00 14 08 33
01 09 01 00  00 17 12 1f
01 0a 01 00  00 1b 34 33
01 0b 01 00  00 1e 0f 47
01 0c 01 00  00 21 1b 11
01 aa 01 01  00 25 28 42

Same disc without CD-TEXT…


SEND_CUE_SHEET

01 00 00 01  00 00 00 00
01 01 00 00  00 00 00 00
01 01 01 00  00 00 02 00
01 02 01 00  00 02 34 18
01 03 01 00  00 05 23 27
01 04 01 00  00 07 33 33
01 05 01 00  00 0a 28 48
01 06 01 00  00 0d 26 29
01 07 01 00  00 10 29 1b
01 08 01 00  00 14 08 33
01 09 01 00  00 17 12 1f
01 0a 01 00  00 1b 34 33
01 0b 01 00  00 1e 0f 47
01 0c 01 00  00 21 1b 11
01 aa 01 01  00 25 28 42


#5

Thanks for reply LUK. :slight_smile:

Just to add some description for some noobs…

Ah, those are MMC cue sheet being used (sent) when burning with the wonderful ImgBurn.

For people interested please see page 278 in MMC2 draft for details.

I’ll mention a few things for the many lazy people. The structure of the columns are:
Q Ctrl/ADR, TNO, Index, Data form, SCMS, Amin, Asec, Aframe

The “Data form” byte defines the block size and whether subchannels are to be sent by software (or be generated by the drive). The upper 2 bits controls the selection of subchannel data to be sent by software.

I see only 41h (4th column) was used for the case of CD-TEXT and only for leadin. So I guess ImgBurn writes leadin fully. The data form is 01h for both cases and so leadout (TNO=aa) is generated by the drive. I see it’s 00h for 1st gap, TNO=01, Index 00, hmm I guess drive generates that part? Or perhaps I have that wrong.

Ops, forgot to say seems 41h is a special case for leadin and leadout, MMC2 draft says that if you use that then software will write that part.


#6

I see it’s 00h for 1st gap, TNO=01, Index 00, hmm I guess drive generates that part?

00h would be the same as the other Index 01 entries…and I’m pretty sure I write those :wink:

Yeah 41h is where I write the leadin myself - which I understand is required if you want cdtext on the disc.


#7

[QUOTE=LIGHTNING UK!;2121658]00h would be the same as the other Index 01 entries…and I’m pretty sure I write those ;)[/QUOTE]

Ops, I forgot to mention and you’re sending your own subchannel data with that? :o

Because gap style (Sony or Philips subchannel relative time formats) is to do with subchannels and I see data form is 00h, meaning no sending of subchannel data.

Not for you LUK or IpseDixit…

Just to let everyone else knows (except us experts) there are 2 different styles or (subchannel formats) for gaps, they differ by the r-time that counts down.

Sony counts down to 00:00.00, while Philips one counts down to 00:00.01 for the last gap sector. If your drive can reading sector -1 with subs then you can see this. :slight_smile:


#8

I never send any subchannel stuff, just the 2352 bytes of audio data (where it’s ‘digital silence’ (all 0’s) for the gaps).

I hope that’s right, lol!