Problem "CF card unreadable"

[Chris Braga]

Hello all,

I've got a problem with our card reader in the post production house, I'm the Data Wrangler working a major Australian drama and we are shooting it on two Sony F23's with Nanoflash boxes as our recorders.

And as I have said we have a problem with our Nanoflash box in the post production house. Sometimes maybe 2 out of 10 cards we sent of to The Lab for rushes come up as "CF card unreadable" and that is causing a lot of issues due the post production house having to put the cards into the AVID to get the files off, but when they do it that way they cannot pull the timecode off as well, meaning they told they production if this keeps on happening they will have to charge more meaning and unhappy production office who keep on asking us why this problem keeps on happening.

Now we have rented our cameras and boxes of Panavision and they think the problem is in the metadata which I think it was as well, but the problem is we have done all they have said to avoid having corruption in the format such as don't unplug the battery straight after recording wait at least 10-15 secs, the same when I have to change the card. And all the boxes are the same firmware.

So we have done that, but still we have the same problem with some cards becoming unreadable.

Does anyone know why this keeps on happening or at least what else we can try to avoid this from continuing to happen?

[Andy Mangrum]

Hello Chris,

Are you on the Current Firmware with your nanoFlashes? 1.6.248, If not I would recommend upgrading both units to the latest firmware, As for the card's not being recognized is this an entire Compact Flash Card, or simply the last file recorded?

http://www.convergent-design.com/Por...sh_1.6.248.zip

Feel Free to email us at CDSupport "at" Convergent-design.com So we can help you resolve this Issue

Best Regards

[Chris Braga]

Hi Andy,

No we do not have the lasted firmware on the boxes and the reason is also due to post, the latest update cannot playing t/c out without a something like 20-25 sec delay and post cannot use that version so we are using version 1.6.29 because that is the last firmware update that could play out the vision and t/c synced together.

And also its the entire card, as soon as they put it into the box it shows up as CF Card unreadable even though if you put it into a computer you can view the files and the metadata but the actual box cannot recognise the card.

Like I said, it doesn't happen to all the cards just some, but when it does it usually happens to one or more card at the same time, like maybe 3 weeks ago we got a call from the lab and they told us 5 out of the 10 cards were unreadable. And we thought maybe it was just one box but it happens on both cameras on both boxes, so there has to be a common link somewhere in the chain that is causing all this problem, I guess i'll give you an example of the workflow;

Cards formatted in box ----> Cards recorded onto ------> Cards checked onset with Macbook pro and data dump for backups -------> Cards sent to post ------> Post put cards in their nanobox and record vision onto tape from their nanobox (and thats apparently the faster way with the timecode and thats also the reason why we can't use the lastest firmware).

So something is causing the issue and we have tried many things to and avoid the problem but it still happens, and post are not liking it, which is giving us grief.

[Dan Keaton]

Dear Chris,

Please send me an email with the Brand Name / Type of cards that you are using.

Where were these cards purchased?

Please note that a leading brand name card is frequently counterfeited.

How old are these cards? How much use have these cards had?

If certain cards are giving you problems, please consider replacing them.
Almost all reputable cards have lifetime warranties.

My email address is:

CDSupport (at) convergent-design (dot) com

Or you can use the Private Mail system of this forming by clicking on my name to the left.

[Ron Aerts]

Originally Posted by Chris Braga

Cards formatted in box ----> Cards recorded onto ------> Cards checked onset with Macbook pro and data dump for backups -------> Cards sent to post ------> Post put cards in their nanobox and record vision onto tape from their nanobox (and thats apparently the faster way with the timecode and thats also the reason why we can't use the lastest firmware).

I dont want to interfere, but I'm very curious why the data isn't copied direct to the NLE, so without the writing to tape. Are they still editing with taperecorders? Isn't the timecode embedded in the file or is this the exactly the problem for the NLE?
Thanks

[Dave Sperling]

Hello Chris,

I reacall reading some threads about the files that a MBP may (or may not) put onto the CF cards when they are ejected from the computer. I'm wondering if this could be part of the problem.

There were also some threads about issues with certain MBP's damaging cards if they were ejected using the 'trash bin.' The recommendation at that time was to pull the card directly out of the adapter without going through the 'eject' process.

Do you have NEXTO drives (or even PC's) that you could use to offload the data on set rather than using MacBooks? I use both (NEXTOs and VAIO computers) and have never had a 'card unreadable' message on re-insertion into my NanoFlash.

Also -- have you verified the cards as readable on set with the NanoFlash there - after doing all your on-set transfers - before sending the cards to the post house? This should be the last step before shipping! If a card shows up as readable on one NanoFlash, it should be readable on the next one (with the same firmware) as well, as long as it hasn't been in any other device during the intervening time.

Best,
Dave S.

[Chris Braga]

Sorry about the late reply,

Its Monday now here in Aus and I didn't have my comp over the weekend and just now got all your replies and thanks for them.

Hi Dan,

The cards are SanDisk 64 GB ExtremePro CF cards we have 20 of them, and they were all purchased by Panavision around November last year, as to where Panavision bought them, I wouldn't know. But I'm sure the guys at Panavision know where they are buying there products and wouldn't buy dud ones.

Hi Ron,

The reason why post won't edit the files straight of the card is due to the problem they do not like having to "stich" the files which go over 3.77 gigs back together again (they say it takes too much of their time up), which means they would of charged the production a lot more which would of made the production not want to use the Nanoflash.

Hi Dave,

With my MBP I do not use the trash bin ejection method, I use right click on the file and eject. With the situation with pulling the cards without ejecting them, I was under the impression that could damage cards?

I do not have any NEXTO drives or PC's on set, I use my MBP as the computer onset to do all the checking and if my Mac was the problem, then why is it so intermittent?

Yeah I verify the cards onset to make sure they are indeed readable, I have a spare Nanoflash box sitting with me next to my comp and I will ofter put the card into the box to see if it reads before transfer and after and I've never come up with a problem.

Thanks for the replies, I usually get them a day later due to you guys replying while i'm asleep hehe.

Cheers Chris

[Dan Keaton]

Dear Chris,

The Mac operating system writes files to the CompactFlash cards during the Eject process.

The best procedure is to just pull the card out (or pull out the CompactFlash card reader with the card still in it), without performing any File|Eject process, or dragging and dropping of the Card Reader Icon to the trash can.

You will get an Appler error message which can be safely ignored.

The best procedure is to use a Nexto. However, with a Nexto, this is the best procedure as the Mac then does not write the files to the card.

If you are sending nanoFlash CompactFlash cards to your post house that uses Avid, the files wll still be in 3.72 GB chunks, unless you are rendering them in your Mac to a different disk subsystem.

As I understand your process, you are sending the original CompactFlash cards, the ones created in the nanoFlash to your post house.

In post, one can drag and drop a group of our files to the timeline in one operation, and they should fall in place on the timeline next to each other. (I know this works for many NLE's, but I have not confirmed that this works for AVID before I wrote this.)

[Alister Chapman]

So the post house thinks it takes more time to join s few files together than it does to make a tape dub of the material from the NanoFlash and the digitise the tape into the NLE??? let's also not forget the quality drop involved in the decompress from NanoFlash, encode and record to tape, play from tape and decompress, re-compress into NLE quality hit.

You really have to wonder about the people running post houses sometimes.

[Chris Braga]

Hey Dan,

Well the good news is the pulling the cards out instead of ejecting has seem to have done the trick, no more Unknown CF card messages. And the answer to your question is yes I am sending the originals in with the runner picking up the previous days ones and sending them back to me, so I rotate the cards.

When I went to the post house a couple of days ago to drop of rushes, I brought up the question why don't they edit the problem straight from the cards, isn't the timecode embedded, and I was told the timecode keeps slipping when they do that giving them lots of syncing issues. And also they don't like stiching the files back together.

Hey Alister, yeah I know how it sounds but they find it easier, but today they've called my DP and me and explained that they thought the timecode was ok when doing it that way but now they have found it even on tape it will slip, so they well want to see if we can fix that problem (it seems like i'm doing all the fixing and there just complaining x( (BTW this post house is the biggest in Australia, so they have weight to there complains and it means a lot of DOP's will not be able to use set ups like ours, due to the fact they don't like it, so its my job to make it usable for them, and try to change there thought as my show is the first major long form drama to use the data card system, I'm sort of the trial for all of Aus... eep!.)

Oh well now i'm going to start a new thread to for another problem, thanks to all who helped so far its working, now lets see if we can sort my other problem now (you guessed it post problems x( ).

[Dan Keaton]

Dear Chris,

Please let us know more about "Time Code Slip".

All non-linear editors that we know of, use just the initial timecode value in the file header, not the timecode value that is embedded in each frame.

Thus, the NLE's just calculate what the timecode value is for each frame based on the starting timecode for the file.

What camera are you using?

What firmware version are you using?

If you are using a tape based camera, then you should be using our 1.6.248 firmware with Timecode Trigger Delay set to 1 second or higher.

This solves the problem of many tape based cameras sending us bogus timecodes while they are repositing the tape prior to actually starting to record.