What happens at the CD factory?

[Waltz Mastering]

Max C1 error rate for red book standard I'm pretty sure is 220/sec.

[Zownd]

Originally Posted by I.R.Baboon

Or to put it another way: If Samplitude finds no errors after it does a null test (comparing the data on the hard disk to the data read from the burnt CD), then it doesn't matter how many C1 errors show up in Plextools, right?

Let the fun commence!

Well, importing a CD will use the CD-roms error correction and your null test will most likely be perfect.
Anyway if they are under 20 at any given time, your fine. anything above I will bin it. But usually it does not go over 12.

Most (modern) CD players have great error correction so there is some room for error.
However the CD-plant might beg to differ.

these are my standards,.. I am quite curious how other ME's handle this.

Cheers,
Mark

[Cellotron]

This was the procedure when I was working at Europadisk -
* CD-R pre-master received was placed into a CD-ROM drive and verified by customer service rep entering the order into database to have same amount of tracks and total length as indicated on provided PQ sheet, and whether it contained audio, data, or multi-session (i.e. CD Extra/Enhanced CD).
* marked top surface of CD-R master disc with CD marker with work order # (so it could be kept straight as to what is what during manufacturing process) along with internal work order forms (also indicating total tracks, total length, type of disc, amount of stampers ordered, requested turn time, etc.), and forwarded it to mastering dept. for pre-testing
* Eclipse pre-testing is run - Welcome to Eclipse Data Technologies - where CD is read in a CD-ROM drive (usually Plextor) by Windows based workstation running Eclipse software. Test checks BLER (aka E11/E21 - C1's - the first level of error correction) C2's (E12/E22 - second level of error correction), CU's (aka E32's - which are UNcorrectable errors), as well as correctness of subcodes, TOC, index placement, timing sync, etc.
* CD-R master given a "pass", "fail" or "warning" grade. In terms of specs for this: PMCD spec allows for a BLER rate (C1 errors/sec) up to 220 - anything over would generate a "fail". A certain amount of C2's (can't remember what the threshold was - it depended on amount of burst within a single frame though) detected would generate a "fail" as well. The presence of even a single CU would generate a "fail".
* Discs with "fail" or "warning" kicked back to the customer service rep to see if the client wanted to either provide new disc - or have the pre-mastering department fix the master disc for them (by using a copying the disc's image with "secure" DAE routine, listening at any point where an uncorrectable error is noted, applying any fix of dropout or clicks at these points needed, and then re-burning a new master disc).
* Discs with "pass" grade are loaded into one of the Eclipse workstations hard drive via Plextor CD-ROM drives - using a highly "secure" Digital Audio Extraction (DAE) routine which re-reads sectors where an error flag is triggered so that any correctable errors are in fact corrected for in the copied image. These images are formated as DDP.
* DDP image is then streamed via EFM encoding to a Laser Beam Recorder which exposes what will become the "pits and lands" on a photo-sensitive glass disc. At this point (actually for about the past 20 years) all of this occurs inside a self contained machine meaning there is no need for "clean rooms" or guys in the white surgeon suits to do this anymore. Generally transfer to the LBR is done at 4x - and most LBR's are optimized for this speed (or faster) at this point - but for those who wanted to pay an upcharge a "real time mastering" transfer at 1x was offered as well.
* glass master is developed, and then plated with nickel. Nickel is pulled off, cut and sanded to size - this becomes the stamper that goes on the press.

There were lots of further quality controls at the pressing stage - but I figured I provided more detail you wanted already. Good info at Recording to CD: More than Meets the Ear

In answer to one of your questions: the replicated discs will have the pit geometry based on what is on the stamper (with potential variations due to the manufacturing process) and as such it is most likely they will trigger error flags differently from what the provided CD-R master disc did.

Hope that helps.

Best regards,
Steve Berson

[jamesp]

It might be worth warning people at this point that Plextools' C1 error readings aren't the same as the BLER since Plextools counts the number of error correction frames with errors - not the actual number of errors. If you test a disc on a Clover or Eclipse disc tester, you will usually find the number of C1 errors to be higher than the number Plextools reports.

James.

[tpad]

...When one sends a master CD (not a DDP) to the factory, do they produce a manufactured CD with the exact same pit layout as on your master CD....

The EXACT pit layout on your master CD contains errors, so you wouldn't want them replicating your master CD errors, would you?

Not very many involved with CD seem to appreciate that the medium statistically is always generating errors on playback. If it weren't for the robust error correction system, you probably wouldn't find the medium acceptable. One of the features of DDP is that it guarantees the data at glass mastering stage is correct and error free. If you've got a good quality master CD and error correction is working properly, you probably will also have good data at glass, but it isn't guaranteed.

[kingtoad]

Originally Posted by Cellotron

Loads of stuff

Steve, i just wanted to say thanks for that, it was very informative.

[I.R.Baboon]

Thanks all, 'specially Steve!

[aleatoric]

Originally Posted by kingtoad

Steve, i just wanted to say thanks for that, it was very informative.

+1!

[Petrus]

Originally Posted by Cellotron

* Discs with "pass" grade are loaded into one of the Eclipse workstations hard drive via Plextor CD-ROM drives - using a highly "secure" Digital Audio Extraction (DAE) routine which re-reads sectors where an error flag is triggered so that any correctable errors are in fact corrected for in the copied image. These images are formated as DDP.

This basically shows (combined with all the fuss about checking and re-cheking the master CD first) how idiotic it is to use CD-ROM as a carrier medium between the mastering house and CD plant. Make a DDP and send it over the net or on a USB thumb drive instead. Cheaper, safer and faster for all.

[Cellotron]

Originally Posted by Petrus

This basically shows (combined with all the fuss about checking and re-cheking the master CD first) how idiotic it is to use CD-ROM as a carrier medium between the mastering house and CD plant. Make a DDP and send it over the net or on a USB thumb drive instead. Cheaper, safer and faster for all.

The thing is that a good percentage of CD pressing plants keep their businesses firmly rooted in the 20th Century and still will not accept DDP.

For some of these places it's because they just don't have broadband FTP setup - but for others it is in fact because they only have CD-ROM (and possibly Exabyte and/or DLT) as the only way to get data into their older Eclipse workstations. That's why the USB pen drive you send to these kinds of places will likely just sit on the customer service reps desk instead of getting processed.

Some places also like to receive CD-R master discs because it allows them to format the DDP in a way they know will be correct and compatible with their system. Spending time and materials making un-useable or incorrect stampers or discs is a seriously an expensive mistake for a CD replicator.

It's also important to note that a majority of CD-R duplication facilities doing short runs can not accept DDP either.

Best regards,
Steve Berson

[White Elephant]

Sorry.......I couldn't resist.

[huejahfink]

Originally Posted by White Elephant

stays at the CD factory

whaa, whaa, whaaaaaaaa

[Trakworx]

Originally Posted by Cellotron

The thing is that a good percentage of CD pressing plants keep their businesses firmly rooted in the 20th Century and still will not accept DDP.

For some of these places it's because they just don't have broadband FTP setup - but for others it is in fact because they only have CD-ROM (and possibly Exabyte and/or DLT) as the only way to get data into their older Eclipse workstations. That's why the USB pen drive you send to these kinds of places will likely just sit on the customer service reps desk instead of getting processed.

Some places also like to receive CD-R master discs because it allows them to format the DDP in a way they know will be correct and compatible with their system. Spending time and materials making un-useable or incorrect stampers or discs is a seriously an expensive mistake for a CD replicator.

It's also important to note that a majority of CD-R duplication facilities doing short runs can not accept DDP either.

Best regards,
Steve Berson

Also, the #1 reason I've been given by plants who do not accept DDPs is self protection; they want a master disc that has been listened to and approved by the client prior to submission so there will be no surprises. There is some wisdom in this approach IMO, especially when dealing with independents.

Thanks for all the great info!

.

[kosmokrator]

Originally Posted by Cellotron

The thing is that a good percentage of CD pressing plants keep their businesses firmly rooted in the 20th Century and still will not accept DDP.

For some of these places it's because they just don't have broadband FTP setup - but for others it is in fact because they only have CD-ROM (and possibly Exabyte and/or DLT) as the only way to get data into their older Eclipse workstations. That's why the USB pen drive you send to these kinds of places will likely just sit on the customer service reps desk instead of getting processed.

Some places also like to receive CD-R master discs because it allows them to format the DDP in a way they know will be correct and compatible with their system. Spending time and materials making un-useable or incorrect stampers or discs is a seriously an expensive mistake for a CD replicator.

It's also important to note that a majority of CD-R duplication facilities doing short runs can not accept DDP either.

Best regards,
Steve Berson

Hmm.
Didn't burn any master CDR in about 3 years now. All DDP.

[acorneau]

Originally Posted by kosmokrator

Hmm.
Didn't burn any master CDR in about 3 years now. All DDP.

Lucky duck.

DDP has gone way up for me, thankfully, but I'm still making 30-40% of masters for my clients as CD-DA (audio CD). Still plenty of reference discs, though.

[Bob Olhsson]

DDP use seems much more universal in Europe.

A plant only needs to get burned a few times with "My 1000 CDs don't sound like my CD-R" to start refusing DDPs.