HVX200 : Data Corruption....

L

ltf3

Member
Hi

After a load of tests we finally felt confident to put the 200 on a real shoot.

Shot at 720 24PN with 2 x 4 gig P2 cards. When full, the cards were dumped to a Lacie d2 Hard drive via FW.

Looking at the footage afterwards showed that mostly all was well.... except for one 'dump' ( we think it's isolated to one transfer at this stage) to the drive where the clips in FCP show random data drop outs.

The drop outs appear as large (ie not single pixel) squares of random position, color etc within a single frame. There's usually 10-20 across each affected frame, and within an average shot perhaps 10 frames have the abberation.

The corruption seems never to continue for more than a single frame, ie the frames before and after are always clean.

The clips play fine in FCP and QT.

I think I can fix them by blending the missing data with a copy of the clip offset by a frame.

But my confidence is thrown.......

Anyone else seen this?

When we sort it all out I'll post some samples....

Thanks

Lee
 
I've seen this before and it seems to be a FCP problem.
Have you tried to reimport the clip from the original P2 files?
 
If I have time I try to look at playback from the card before off-loading. After I copy the cards I try to take 5 min and import the clips to FCP. I do a random check again before formatting the cards.
I know this is not possible on every shoot but I would try. Plus what was your power source?? Its possible you may have had a power spike during that transfer?? If plugging into an unfamiliar source maybe some sort of power conditioner would help?
 
We only use batteries on location shoots due to the "ya never know" factor. Plus it frees you up from an extra cable.
Keep us posted on your findings.
 
I wonder if this is an HD error. I've had two HDs fail in the past two months and my company backs up all it's important data, but recovery is a tedious pain in the behind, and slow at best. It seems that with so many factors involved and although HDs are fairly reliable, the only way to be sure is to view the dumped footage to make sure it's there. That kinda puts a damper on the workflow.

My question is; exactly how many different files are transfered from the P2 card to be able to import the DVCProHD file into FCP?
 
I think a good idea is to have two drives and offload clips to both. May take twice as long but at least you will have less chance of losing footage if a drive fails.

And big drives are pretty cheap now too.
 
No doubt booth.

IMO, I'd go for a mirrored RAID array, but you still need to make sure that there was no file corruption anywhere along the chain. I'm sure it's not something that will show it's ugly face often, but checking the footage is probably still a good idea. Better safe than sorry.
 
Checking footage is a right pain in the arse. Once you've established that you're camera isn't poked (Doing a small test before starting to shoot, and maybe between breaks) then you shouldn't need to check. What if you're just shot 16 mins of footage. Gonna watch the whole lot again before you do another take, or more to the next subject?
 
Aza said:
Checking footage is a right pain in the arse. Once you've established that you're camera isn't poked (Doing a small test before starting to shoot, and maybe between breaks) then you shouldn't need to check. What if you're just shot 16 mins of footage. Gonna watch the whole lot again before you do another take, or more to the next subject?
Click to expand...

That's my question, what does one do? I guess it's a matter of "picking your poison." Deal with tape or use kid's gloves with the HDs.
 
Well it's the same with tape stock, there's about as much chance of getting a dodgy tape and having loads of dropout during takes.
 
That's what I meant to say.

Film... mag scratches and gunk in the gate.

Tape... drop-outs.

P2... corrupted data files.

Pick your poison.
 
Yeah but come on, you shouldn't be getting data corruption with P2 unless you've got damaged P2 cards or other hardware. If you're getting dropouts with P2, something VERY serious is wrong with your P2 cards or your other hardware. Do you just put up with the fact that when you copy a file on your PC, to a USB keyring or another drive that it just might have corrupted the data. I doubt it, and we shouldn't accept that. Also, it very rarely, if ever, happens.




tim_brown said:
That's what I meant to say.

Film... mag scratches and gunk in the gate.

Tape... drop-outs.

P2... corrupted data files.

Pick your poison.
Click to expand...
 
Aza perhaps you can say rarely, but over the past two months I've had two HDs fail and I've had seven HDs fail in total since I've been working with computers (All Maxtors too but that's another thread--they couldn't pay me to use those things), so maybe I'm coming from a different background and in my world it happens a little more often than I would like.

I have no qualms or concerns with the durability of P2 media, my beef is with the flakeyness of HDs.
 
Haha, yes Maxtor is another thread and you wouldn't pay me to use them either. I had one that died, maybe, dunno 10 years ago, not sure when, and haven't used them since. Since then I've had 2 drives die on me, both work ones, a SCSI drive and a 5200 rpm 2.5" drive.

Seagate all the way ;)
 
tim_brown said:
I have no qualms or concerns with the durability of P2 media, my beef is with the flakeyness of HDs.
Click to expand...
That'd be my take on it as well. I'd wager that it was during the copy to the HD that the issue came about.

Redundancy is good, and that's why I've taken to doing duplicate copies from the card when offloading. Once to the internal, and simultaneously to the external drive. Doesn't take any longer (well, maybe a little bit longer, but not really noticeably) and it gives you two copies in case anything goes wrong (and I mean anything, like the client having the hard drive stolen out of his car, or a data error during copying, or someone dropping the drive or whatever). Two copies is better than one.
 
Thanks for all the replys....

The drop out issue is definitely an FCP import problem.

I isolated it to the fact that if you attempt to import more than 5 clips into FCP (from an ext FW HD where the 200 footage was dumped on location) then pixel drop outs will crop up in the FCP clips.

If you import them in groups of 4 or 5 though, everything is fine.

At least the 200 gets a clean bill of health!

Lee
 
Barry, how do you dump to two drives at the same time? I'm a little paranoid about my clients stuff. Esecially when you have crew and cast of 10 or more on a full day shoot.
Thanks.
 
Just copy to one drive, and while it's copying, start copying to another. The card is fast enough to support that, and because it's solid state there's no issue with seeking or head positioning. I've been doing it for a while with no ill effects; it seems to finish copying to two drives in barely any more time than it takes to copy to just one.
 
Lee,

I rented the P2 to shoot two :30 spots this weekend and in doing a radom review of the clips have noticed the same issue. the camera was new out of the box (helped bench test it myself)

Mostly only one frame at a time, and very random. I am a little sketched about it as well and am anxious to get the spots into post and see the extent of the problem.

We downloaded the 4GB cards to a PB 1.67 Ghz 10.4.4 1GB ram FCP 5.0.4. PB and 3 back up HDD's all running on AC power through a power conditioner.

I don't think this anything to do with ingest processes. It looks like a bug. Unfortunatly we were unaware of the issue so we didn't keep track of whether or not all the corrupt clips happened on the same P2 card. We were rotating 3 cards during shoots.

Please post info as you get it, as will I. I'm throwing this out to the Creative Cow P2 post as well.

Dan
 
Hi Dan

Sorry to hear you experienced the same issue.

As I mentioned, I just reimported the affected clips into FCP from the original Contents folder and they came in fine.... as long as it was just a few clips at a time.

Does anyone have either Apple or Panasonic contacts to escalate this issue to? It's kind of fundamental to find out where the issue is!

Thanks

Lee
 
You must log in or register to reply here.
Back
Top