PDA

View Full Version : Any idea what would cause this...?


precinct13
11-16-2007, 09:27 PM
I was synching up the day's footage when I ran into this:

http://www.precinct13video.com/images/A106_C007.jpg

http://www.precinct13video.com/images/A109_C001.jpg

They appear randomly over several shots. Also, some clips shut down QT and FCP when played...

Any ideas?

Thanks in advance!
Greg M
11-16-2007, 09:32 PM
what are you doing?
what are you running?

need more info to help.
Matt Uhry
11-17-2007, 12:05 AM
I've had that happen to me and was not able to identify the problem. ( tried many many variables too ) Seek Help. If you figure out what it was let us know. What camera # are you using ?

Matt Uhry
www.mattuhry.com
Rick Darge
11-17-2007, 02:45 AM
I had that happen to me to... way back when... couldn't isolate the problem
Joseph Mastantuono
11-17-2007, 07:31 AM
Triple check your FW cables, and your CF card reader. It's a data problem, not a sensor problem.

If the issue persists, call it in. They should get you up and running pretty quickly, once you've made sure that the problem exists on the card itself, and it's on the card itself.
Finner
11-17-2007, 07:45 AM
For everyone who has had this problem lets start with the exact model and brand of card reader used.
Matt Uhry
11-17-2007, 07:59 AM
Triple check your FW cables, and your CF card reader. It's a data problem, not a sensor problem.


Why do you think that ? What in particular about the problem makes you think, "It might be a bad cable or CF card" ?

Matt Uhry
www.mattuhry.com
precinct13
11-17-2007, 08:09 AM
Sorry to hear others have had the same problem...

It's Red #15. I'm not sure what firmware build. I'm just the assistant editor. I get handed a drive with all the material on it at the end of the shoot day. I'll try to find out build # and CF reader type.

This has happened over a couple of nights now. The only thing I can find that these shots have in common is that they are night time exteriors and its been very cold here in Ohio. But I'll keep digging...
Finner
11-17-2007, 08:17 AM
Is it getting worse with each new card shot?
precinct13
11-17-2007, 08:34 AM
I wouldn't say it gets progressively worse...it's pretty random. But it definitely happens most in the last third of the shoot day. Which, lately, has been midnight to 4am in 30 degree weather.
Greg M
11-17-2007, 09:03 AM
The firmware version is pretty important info here...can you find out?
jbeale
11-17-2007, 09:16 AM
With all the metadata encoded in the R3D file, maybe firmware version should be included too?
Joseph Mastantuono
11-17-2007, 09:35 AM
Why do you think that ? What in particular about the problem makes you think, "It might be a bad cable or CF card" ?

Matt Uhry
www.mattuhry.com

Sorry, I meant card reader. I have seen the 'crashing fcp' thing show up when copied with a bad cable, I replaced the cable in a dummy check and the second copy when fine.

But do contact red tech support asap, because if those aren't a bad copy, it means that that stuff is actually on the R3D file.

It is NOT a cold issue...
Adrian T.
11-17-2007, 09:53 AM
I found a similar problem with a sample clip in REDCODE. Some areas of certain frames show a weird pattern. On one frame it's there, on the next frame it's gone.

http://img232.imageshack.us/img232/5836/anomalyfd8.jpg

Note that I was working in "Full res. High".

Can anybody reproduce this?
Nick Shaw
11-17-2007, 09:56 AM
I saw this too. But if you go to the same frame in Red Alert it's not there. Reassuring that it seems to be a function of the decoding rather than the encoding of the footage.

I emailed Graeme about it and he's on the case.
Stacey Spears
11-17-2007, 10:45 AM
The waffle like pattern is there on frames 265 and 266 in full res high quality mode.
Brent J. Craig
11-17-2007, 11:34 AM
See this thread: http://www.reduser.net/forum/showthread.php?t=5987 Graeme says its a decoding issue. They are working on it
precinct13
11-17-2007, 02:01 PM
Well, if it still matters, I found out it is build 6 on the camera.

Also, I still see the patterns when I load the .r3d file in Red Alert.
Cüneyt Kaya
11-17-2007, 02:04 PM
Well, if it still matters, I found out it is build 6 on the camera.

Also, I still see the patterns when I load the .r3d file in Red Alert.

in red alert too !?

you should get in contact with graeme, send him a pm
Greg M
11-17-2007, 02:53 PM
Well, if it still matters, I found out it is build 6 on the camera.

Also, I still see the patterns when I load the .r3d file in Red Alert.

you might want to update to at least build 8
Geoff Reisner
11-17-2007, 08:04 PM
I had a similar problem with some footage from B camera in a recent shoot. We were running build 9 (I believe) and in FCP/Compressor, it'll crash when dealing with those specific frames. One frame is good, the next will be badly messed up. I figure I can cut around those frames in the actual edit, but right now it's just the matter of rendering the shots out into a useable format that won't crash the system.

Later tonight I'm going to try running the shots through RedCine. Perhaps I can render out new proxies or at least a TIFF sequence that can be rendered out to prores without a crash.
Brook Willard
11-17-2007, 08:33 PM
Were there any HMIs in play? I'm sure that's not the cause here, but I've seen some pretty funny things from angry HMIs.
Justin K Phillips
11-18-2007, 10:23 AM
The only thing I can find that these shots have in common is that they are night time exteriors and its been very cold here in Ohio.
A Red in Ohio? Where?! I must see it!
CJ Roy
11-18-2007, 11:13 AM
Were there any HMIs in play? I'm sure that's not the cause here, but I've seen some pretty funny things from angry HMIs.

HMI's causing interference with the Red? How so? Magnetic or electronic ballast?

Sounds like important info. Thanks.
precinct13
11-18-2007, 12:48 PM
A Red in Ohio? Where?! I must see it!
We're in Dayton, but tonight is the last night of the shoot. Email me if you want more details.
Still trying to figure out the problem. We switched camera bodies and last night's shoot went off without a hitch.
Neely O'Hara
03-02-2008, 07:28 PM
I'm working with precinct13 on the same film. Now that we're getting closer to a rough cut of the feature, I'd like to get our "problem" clips straightened out.

Has there been a solution to this problem yet? I see from previous posts that this is possibly a decode problem, but until it is solved I figure anything is possible.

I downloaded the latest version of RedAlert (1.5.4) and ran one of our problem clips through it. The QuickTimes still had the bad frames.

Here is what it looks like:

http://home.mindspring.com/~arotolante/BadFrame.png
Russ Fill
03-05-2008, 02:18 AM
Just jumping in here and may be way out of my realm of understanding but I got a disc or 4 from Silverado to get up to speed on REDCINE with a while back and noticed that some of the r3d. clips had a similar pattern in them. I also noticed that when I shuttled the clip across the pattern it usually made REDCINE crash or blue screen. Those clips were shot back in Sept. so Im sure they had a very old build on their camera. Maybe its a old build vs. new REDCINE build?????
Also just downloaded a clip from Redrelay.net of some girl in a wine celler and noticed that all the QT clips all had the same half frame electriczebra print.
For what its worth.
Charles Angus
03-05-2008, 03:37 AM
Reminds me of when FCP gets angry at DVCPro HD... Sometimes you get the black triangle and static, sometimes the weird solarization, you name it. Quicktime (and by extension FCP) is weird.

What about frame renders? Do they work? (ie. tiff, dpx)
Torrey Loomis
03-06-2008, 11:19 PM
Just jumping in here and may be way out of my realm of understanding but I got a disc or 4 from Silverado to get up to speed on REDCINE with a while back and noticed that some of the r3d. clips had a similar pattern in them. I also noticed that when I shuttled the clip across the pattern it usually made REDCINE crash or blue screen. Those clips were shot back in Sept. so Im sure they had a very old build on their camera. Maybe its a old build vs. new REDCINE build????.

Russ,

Most likely due to us shooting on older firmware or running an older version of REDCINE.

We'll have a "sampler" project released in about two weeks that will blow your mind in terms of size and scope. First of its kind--ever.

Torrey
-----------------------------------------------
Torrey Loomis
President & CEO - Silverado Systems, Inc.
Outfitter to the World's Foremost Apple Professionals
2600 East Bidwell Street, Suite 280
Folsom, CA 95630
(916) 760-0032 • FAX (916) 404-5258
tloomis@silverado.cc
http://www.Silverado.cc

Check out our Insider blog at http://silveradosys.blogspot.com