Page 1 of 1

Corrupt .pmi file - Beware !

Posted: Fri Jul 21, 2006 19:33
by Old Sadie
I have just encountered a problem with a .pmi file generated from a project.

The project constisted of set of stereo clips (1 per CD track) on a single pair of tracks with no processing applied: very basic. This was used to generate a CD image file from which I then burnt a client listening copy and a DDP master.

I had fully auditioned the stereo clips before making the pmi file, which I always do to ensure I have missed nothing.

However, the client came back reporting tiny faults in two out of 12 CD tracks. These appear to have crept in as the pmi was generated. Basically Pyramix seems to have muddled some 18 frame blocks of audio. It is random, and only ever on 1 channel at a time. Given that it was a quite extreme metal band most of the problems were almost undectable unless you knew the music intimately, so it was like looking for a needle in a haystack! This was a close call as the DDP had gone to the plant.

Has anyone else had this happen?

Merging say not, and that ultimately the system can be trusted not to occasionally throw up a problem.

The only way I can figure to be sure there is no problem is to generate a pmi, then generate the DDP, then import the DDP back into a new project then listen to it, or create another 'post DDP' pmi file for the final check CDR. Complicated and time consuming for a professional piece of equipment!

Having made another pmi file from the same project the problem has not recurred and it doesn't appear to affact any other masters we have done. We have never experienced any type of DDP problem before in 10 years (9 on SaDIE).

Any suggestions?

Jon

Re: Corrupt .pmi file - Beware !

Posted: Sat Jul 22, 2006 02:47
by Graemme
Hi Jon,


I had fully auditioned the stereo clips before making the pmi file, which I always do to ensure I have missed nothing.


I'd strongly suggest that you test the .pmi file itself. It can be dragged into the timeline (It's a kind of .WAV file) and auditioned just like any other file format. Given that it takes a couple of minutes for a 44k1/no processing master to be turned into a .PMI fiile, it's a small effort to ensure that your master is good.

You could also drag in the .PMI file, flip it's polarity and line it up against the existing audio- sort of a manual 'null' test.

I also used Sadie (for 5 years) prior to moving to Pyramix and always made a CD-R from the DDP - I'd *never* want to send a master that I hadn't listened to all the way through.

In the not-too-distant future, you'll be able to generate CDs directly from DDP images (in DiscWrite). This will definitely be helpful!

Has anyone else had this happen?


In 'Release' versions of software, only once that I recall and that was version 4.0.x. So, that's literally hundreds of masters without issue.



The only way I can figure to be sure there is no problem is to generate a pmi, then generate the DDP, then import the DDP back into a new project then listen to it, or create another 'post DDP' pmi file for the final check CDR. Complicated and time consuming for a professional piece of equipment!


I don't mean this to be rude or condescending - aren't we paid to do this sort of thing? I don't care who makes the workstation I'm using - a live human being needs to 'sign off' on the master.

There were certainly occasional bugs in Sadie's and Sonic's DDP/CD writing as well and I've just always made sure that I listen to anything going out for replication (as well as a media test).

Best Regards,

Graemme

Posted: Sat Jul 22, 2006 22:35
by Bernhard Guettler
Hi Jon,

I don't know, what caused your problem. But out of curiosity: did you do SRC during the pmi creation?

Best Bernhard

Posted: Wed Aug 02, 2006 13:23
by Old Sadie
Hi Bernhard,

No, no SRC was involved.

Yours,


Jon

Re: Corrupt .pmi file - Beware !

Posted: Wed Aug 02, 2006 13:31
by Old Sadie
Hi Graemme,

Thanks for the wise advice.

"I don't mean this to be rude or condescending - aren't we paid to do this sort of thing? I don't care who makes the workstation I'm using - a live human being needs to 'sign off' on the master."


No offence taken. This then begs the question about whether the audio should also be checked in glass mastering (1630 - those were the days!

"You could also drag in the .PMI file, flip it's polarity and line it up against the existing audio- sort of a manual 'null' test."


This is how I managed to isolate the problem in the first place. It is interesting that the problem 'blocks' were all the same length: on audiofile this would correspond to a block on the HD.

In the not-too-distant future, you'll be able to generate CDs directly from DDP images (in DiscWrite). This will definitely be helpful!

Thanks,


Jon

Posted: Fri Aug 04, 2006 16:36
by sonicjones
Greetings,

You might also consider that this could be a storage issue. Funny things can happen when volumes get near capacity.

Best,

Posted: Fri Aug 04, 2006 19:34
by Graemme
Yup, There's the oft-quoted "Windows needs to 'see' at least 10% free space on a volume for reliable disk operations."

Anybody have any hard facts about whether or not this is accurate?

Empirically, it 'fits' from my experiences.

Although, on large volumes, this may translate to an actual amount of space, rather than a percentage. (e.g. On a Terabyte RAID, you may be able to get away with having less than 100GB available...)

Graemme

sonicjones wrote:Greetings,

You might also consider that this could be a storage issue. Funny things can happen when volumes get near capacity.

Best,

Posted: Mon Aug 07, 2006 07:45
by Bernhard Guettler
sonicjones wrote:Greetings,

You might also consider that this could be a storage issue. Funny things can happen when volumes get near capacity.

Best,


Wouldn't it be possible for Pyramix to detect this, just like a playback stall? I'm sure anyone would prefer an error message in Pyramix over a call from the client ;-)
Best
Bernhard