Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Corrupt Frame Killes all Programs

D

Damian D

New member
Joined
Aug 27, 2009
Messages
2
Reaction score
0
Points
0
I received 15 files from a client to fix, all of them have at least one corrupted frame. I have attached an example of the issue. Of course all of the original mags have been cleared. I am hoping that there is a way to fix this issue.

They are crashing every single program I have at my disposal. The only work around I have found is rendering out a DPX sequence up to the frame before the corrupted frame and then the frames after, but its time consuming is breaking up good takes right in the middle of needed footage.

Thanks

Damian D.
 

Attachments

  • Screen-shot-2010-07-30-at-3.56.jpg
    Screen-shot-2010-07-30-at-3.56.jpg
    37.2 KB · Views: 0
Damian, welcome to the club...!!!

I color graded a feature with MANY corrupted files, and I had similar nightmares. The list is long, but one worth to mention: Redcine-X, if we date to include a file with at least ONLY one corrupted frame, it stalled the entire scene, so we needed to remove early on the file so the XML will not push it through the pipeline in order to avoid this issue. :nono:

Bottomline:

A two week job was draged to almost four weeks to long hours and a lot of frustration. At the end the reward was phenomenal, but it costs me (or to the director) many tantrums... :rant:
 
There is no way to fix the corrupt frames. The corruption happened somewhere between original media and offloading data. Most of the time it seems to be the connector cables or the the Cf card reader/drive. You are allready using the best workflow I know, by transcoding arround those frames. To make the clips usable you could try to fix this one frame by simply duplicating the before or after frame. Most of the time this won't be noticeable. Or you could interpolate the missing frame by "merging" the before and after to one new. I would try to do this after effects time remapping or something like twixtor.
 
This is a tech problem because the corruption is an avoidable issue.

This can be solved with a checksum verification and a quality check. Eye-balls to footage!

I have 3 CF readers so if I find one corrupting footage, I switch it out. It has alot to do with Heat. They were never intended to download as much data as we can in a day.

Once data is corrupted all you can do is cut around it.

Goodluck,

-
Dane
 
I've had to do the same thing: cut around the problem. I have had issues with bad frames before from the red rocket and had to export out just that one clip without using the rocket as well.
 
We got this kind of corrupted files from a feature shoot as well, it was caused by the combination of the LEXAR FW800 CF Reader together with LACIE RUGGED 500GB Disks over the same bus. We where lucky, cause they backed up the files also on a nextoDI and we could recover about 80% of the corrupted files from the nexto.
The rest was as discribed, cutting arround and exporting dpx sequences.

After we discovered the issue, we suggest to change the FW800 Reader and didn´t have one more corrupted file ever since on that project. I think they changed to an Express34 CF Card reader.

Another important thing is, alsways use 1 Generation files. On another job, they didn´t want to buy Lacie Rugged drives for the transfer between post and set, so they used one Lacie rugged on set and copied the files after shooting from the Lacie disk to another Drive, this also generated some corrupted files.

best regards
Manuel
 
I use ice packs on set to keep CF readers cool. They get hot and start to drop frames. I have 4 readers and constantly rotate them as well as ice them down. REI has waterproof plastic bags that are easy to fill with Ice and seal tightly with a built in clamp.
 
Dane Brehm said:
This is a tech problem because the corruption is an avoidable issue.
This can be solved with a checksum verification and a quality check. Eye-balls to footage!
Click to expand...

Have had plenty of instances where the checksums never caught corrupt frames, leading me to believe it's a camera to drive/card error in most cases. I've had this problem 30+ times (from every single different rental house in LA), and I've only had the card/drive to offload corrupt file error maybe twice.

As you say, eye-balling it by watching every frame is the only real way to catch it... sometimes, depending in which beta software you're using. And of course the QT proxies show something completely different.
 
CJ Roy said:
Have had plenty of instances where the checksums never caught corrupt frames, leading me to believe it's a camera to drive/card error in most cases. I've had this problem 30+ times (from every single different rental house in LA), and I've only had the card/drive to offload corrupt file error maybe twice.

As you say, eye-balling it by watching every frame is the only real way to catch it... sometimes, depending in which beta software you're using. And of course the QT proxies show something completely different.
Click to expand...

Checksums being done on the computer side can give false positives because the data going into the checksum could be corrupted by the reader/cable. The way around that is to do a separate pass, ideally on a different reader and computer and compare the checksums. If they differ, then you know somethings bad on the copy.
 
interesting... have this problem here at the moment as well.... but I have the bad frames also on the original cf card. so it is a problem on the camera side... if there is a problem with the cf reader on the pc, most of the time r3d manager tells it. but as said before... the bad frame is also on the card which comes direct from the camera... probably a problem with the cf module there... if not, we need to change the cam....
 
i see it clearly when looking to my 24inch HD-SDI monitor hooked up to my rocket, on camera playback it is hard to recognise....
 
Hey guys,

Not saying that everyone should run out and buy SCRATCH (well, yes... that is what I'm saying, but whatever...) :beer:

SCRATCH has a pretty good way of dealing with errant frames, and the app almost never crashes, but instead puts a black frame (or just the corrupt data) in the place of the bad frame/s.

Lucas

Lucas Wilson
------------
Director, Business Development
ASSIMILATE, inc.
LA, CA, USA
 
Rainer, had this problem twice in the last two years. first was a loose connection inside the Red, the CF module is connected via a standard connector to the main body and it got loose probably during a rollercoaster ride with the red hard mounted to the coaster.

2nd time was on a long term shoot when file corruption started to increase over time.
turned out to be a bad CF Module and got it replaced by red. Thank god RED allows change of CF module by user so you don´t need to send the camera in.

Manuel
 
Hi Manuel !

You are absolutely right, the SATA connection from the cf module is a pain in the ass. I made couple of months ago a post regarding this.
I changed I think now three times CF modules.... why is it not possible to use SSD on the CF port. It would be big enough to place it there. Nearly no rental has RED-RAMs....
RED certified SSDs would be nice... so you can decide if you use the CF port or mount the SSD port there....

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