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 :)

Weird RED RAW Glitch/Jitter

Status
Not open for further replies.
B

Ben Smallbone

Member
Joined
Feb 13, 2008
Messages
11
Reaction score
0
Points
0
Location
Nashville, TN
Hey Guys,

I've been shooting on the RED for over a year now and never experienced anything like this:

http://radiatefilms.com/Glitch-B004_C013_0118NL_M.mov

This clip goes on for another 2 minutes just like this. At first I thought I must have accidentally bumped record during setup. However, it would be virtually impossible for me to consistently shake the camera in the exact same way, for that amount of time. I then quickly realized I was missing some takes. 3 other takes of different lengths, shot over 7 minutes, have the exact same jitter/glitch. This was meant to be a beautiful shot of a father and young daughter walking through downtown Raleigh. I looked around the forums and haven't found anything similar. Any thoughts on what would cause this and how it can be fixed?
 
Does the file playback the same way on the camera? How about Redcine-X?
 
Camera was a rental so haven't tried that...but have opened the R3D files in Redcine-X and it still has the same glitch/jitter.
 
Interesting, were you shooting to a Red Drive? Have you tried opening r3d in other programs?
 
Yes, recording to Red drive. Tried opening the r3d in Red Alert as well as through FCP log and transfer with all the same results.
 
It seems like a recording error. Although check the drive, in case it was corrupted during transfer.
 
Don't assume it is a recording error just yet, for some reason the playback apps can trip over internal file pointers and form tight loops like this.

One thing that has saved me from a few bad clips is the fact that if you close RedAlert with the pointer somewhere in the timeline it will re-open at that same place. I'd try opening it in RedAlert, set the index pointer somewhere deeper in the file, close then re-open RedAlert.

You could also try to use RedLine to pull a single tif or dpx from deeper in the file to see if the content you shot is actually there.

Bob
 
Yeah.

It looks to me like the composition of the shot and the varying elements contributed to a type of codec loop. It's like the particular curve ball being thrown at it can't be hit.

What happened when you altered teh shot?

You can put a short piece of it on a loop and get a usable shot out of this as there is noting in the frame that is "moving" per say.

David
 
Try RedLine as noted above, also try Storm.

RedAlert & RedCine are more the same than different.

Any Quicktime based software will behave exactly the same as Quicktime and FCP L&T.

RedLine works differently, and Storm uses its own code.

So if it's a case of "the particular curveball" different software should give better results.

DISCLAIMER: I'm not on the RED Team, and only they actually know the truth of what I suggest.

Also... you may want to try changing platforms. If you are on Windows try OS X and vice versa.
 
Last edited:
benjamen81 said:
Hey Guys,
Click to expand...

Benjamen81,

This is a real names forum.

You need to use your real name when posting here.

For this instant you can add your real name to your sig.

Then post in the real names sticky thread, and an admin will update your account for you.

Thanks,
 
TomDannenberg said:
To me the mov looks faked. EarthQuake post processing?
Click to expand...

Are you trying to be funny?

While this is possible... it doesn't seem fruitful to me to impugn Benjamen81's issue or character.

If it is a real issue, then it does need to be addressed- I think that is clear.

It does occur to me that no one has asked for the .r3d file.

It would be useful to have the original .r3d. Any of us can test it and make sure its not an idiosyncrasy of your specific setup or a bad setting.

RED can also check it out and determine in far more detail than most of us if there are any problems with the footage itself.
 
Radiate films looks like a legitmate site. This could be a way of attracting attention to the site, but I think that Ben just has to sig his real name and also maybe put up the r3d snapshots or a section of the file that is giving him trouble.

A report ticket should also be issued to RED.

David
 
Alexander Ibrahim said:
Are you trying to be funny?

While this is possible... it doesn't seem fruitful to me to impugn Benjamen81's issue or character.

If it is a real issue, then it does need to be addressed- I think that is clear.

It does occur to me that no one has asked for the .r3d file.

It would be useful to have the original .r3d. Any of us can test it and make sure its not an idiosyncrasy of your specific setup or a bad setting.

RED can also check it out and determine in far more detail than most of us if there are any problems with the footage itself.
Click to expand...

I always try to be funny, but:
I figured out the black border on the right side in the mov. It is like moving a footage off the viewer's screen. Standard jitter plug ins also produces this effect if set to high.

So, if no plug in was used, what could be a reason?

If the sensor itself would jitter, then you did not have some black borders.
I also do not think that the codec itself generates this issue. Just think of artefacts when bytes went wrong in the footage.

Technically I see no reason why it should jitter. Maybe I am wrong.

But seeing the original footage may be helpful to find out if it is fake or not
and to find out a reason for this. If my cam would have such an issue, I did not post it here, but directly went to RED and discuss this with them.


Tom
 
Last edited:
Alexander, I put in a request to add my real name on the board before I posted anything. I also added my real name to my signature.

I went ahead and uploaded one of the short clips in R3D here: http://radiatefilms.com/B004_C012_01180X_001.R3D.zip

At this point I'm prepared to not recover the footage. I contacted the rental house where we got the camera from and they have never seen anything like it. I've tried opening the files in Redcine-X, RedAlert, FCP logo & transfer, and now Storm with all the same results. Would really appreciate you guys giving the R3D file a test on your system and see if you experience anything different.
 
Ben,

Is it possible at all this could be some external vibration?

We've never seen anything like this either.

Do you still have the original media? If so, it's worth running REDundead on it to see what files you can get back.

It's too bad that this was a rental camera, it would have been good to take a look at the camera log file.
 
I have seen codecs do this on renders sometimes. There is something (a one in a billion thing in the math) that can create this kind of looping effect.

I think it is just one of those kinds of anomalies.

David
 
Rob Lohman said:
Ben,

Is it possible at all this could be some external vibration?

We've never seen anything like this either.

Do you still have the original media? If so, it's worth running REDundead on it to see what files you can get back.

It's too bad that this was a rental camera, it would have been good to take a look at the camera log file.
Click to expand...

I'm glad this has come to the attention of RED Team.

It's one of those mysteries I really truly hate. Even if there is no way to fix it, I'd love to see it explained.

I am still pulling down the r3d, and will check it on an idle Windows 7 system. I am not hopeful based on Taylor's test.

UPDATE:
Tested on RedCineX 356 Windows 7 and OS X with identical results.

Aren't there any sort of "video disassembler" tools that can really let you look at the codec structure and such?
 
Last edited:
Status
Not open for further replies.
Back
Top