Artifacts with my Scarlet

[Mikey Piliero]

Just got back from a shoot and these strange artifacts have appeared on a lot of footage.

Here's a couple things to note:

This has happened before but never this high of a frequency. It happened on probably a total of 5 clips of footage before this. However its in about 60 percent of clips from this project alone.

I had just upgraded to the newest firmware.

I did not blackshade.

The artifacts only ever appear on the far right side of the frame.


I have looked through all the threads trying to find someone with the exact same issue and had no luck. Most people's issues look a bit more digitized and have crazy colors.

 

[Patrick Grossien]

Haven't seen this before. With strange artifacts like that, I'd first contact RED support directly and see what they have to say.

What settings did you shoot with? RC / resolution?

 

[Sam Winzar]

Looks like it could be the result of a bad copy. Do you still have the original media? Have you tried playing it directly off the card?

 

[ericyoung]

It's pretty subtle - definitely contact Red Support.

 

[Mikey Piliero]

I can confirm it is in fact in the original footage on the Red Mag... . This is very sad news.

 

[Robert Hart]

Firstly, please ignore my comments except for a chance that they prompt better practitioners than I in a helpful direction of thought.


To my rheumy old eyes, there seems to be two things going on. It does not look like a black calibration issue as that normally yields a fixed parttern artifact. That does not seem to be apparent in your uploaded footage. The issue may be a defect in the camera but to me it seems it might be working properly. Image compression on vimeo makes it a bit difficult to study the image closely.

I think your available light might have been a bit low. It seems like plenty but may not have been as much as it looks. The whole area seems to be relatively uncontrasted which creates the impression of adequate available light.

The artifact which seems to initiate from right edge towards left might be the beginning of system noise which occurrs when your camera is struggling for light. From memory the noise propagates from right to left and is apparent as horizontal random noise lines across the image. It may become more apparent if you try to pull detail out of darker areas.



I find that in a low light circumstance with the SI2K, a camera of superficially similar behaviour, that a brighter area not less than about one/eighth of the frame within the image helps to suppress system noise.

You may have to crush the blacks a bit to make it go away - maybe only partially at best.


There is another artifact evident across the entire image if you look hard enough. It is a power frequency effect in the lighting and most apparent in the reflected light on the shiny floor. Crushing black may aggravate this artifact. There might be random frequencies going back into the power supply from machinery which might be causing the fluctuations if it is not an off-frequency issue or incorrect choice of frame rate. I am prompted to that thought by the localisation of the artifact on the areas lit by the strips overhead the workers. Was the building supplied from mains utility power or a factory generator?


Because your uploaded sample shot is dynamic, any technique which involves blending frames such as stacking layers of fractional opacity successively delayed by one frame, will cause an unpleasant smear effect on motion.

Blending frames in combination with crushing the blacks a bit may help or may aggravate the system noise. It may help reduce the power frequency artifact. There may be softare plug-ins which can achieve more. Pllease heed the advice of those responders who are better at this game than I.

 

[Mikey Piliero]

Thanks for the replies everybody.

Robert, I appreciate the in depth analysis. You're very right about the frequency issue and if it was just this footage that it ever happened to I would chalk it up to that but it actually has happened to footage taken outside during daylight. So I don't think it would be a frequency or low light issue.

 

[Phil Holland]

I had just upgraded to the newest firmware.

I did not blackshade.

Mikey, in the future it is recommended by Red to blackshade after updating your firmware.

Don't know exactly what's going on here, but if the field isn't uniform on the calmap something could be funky.

 

[ericyoung]

How are you determining it is on the original footage - by viewing the files direct of the RedMag in Redcine-X Pro? If so, that would rule out a transfer/copy problem which can cause frame corruption.

Looking at it on a brighter screen it looks like it's in a discrete vertical band on the right, and the blotchy pattern is very reminiscent of other people's problems where unfortunately they had to send the camera in to have an internal board replaced. Hopefully it is still under warranty?

Here are some previous threads:
http://www.reduser.net/forum/showthread.php?97004-Epic-Frame-Artifacts-Corrupt-Frames
http://www.reduser.net/forum/showthread.php?90360-Big-problem-artifacts-in-the-picture-red-epic-x
http://www.reduser.net/forum/showthread.php?96490-Strange-Artifacts-from-drive-transfer

 

[Mikey Piliero]

Mikey, in the future it is recommended by Red to blackshade after updating your firmware.

Don't know exactly what's going on here, but if the field isn't uniform on the calmap something could be funky.

Good to know. I'll make sure to do that from now on.

 

[Mikey Piliero]

How are you determining it is on the original footage - by viewing the files direct of the RedMag in Redcine-X Pro? If so, that would rule out a transfer/copy problem which can cause frame corruption.

Looking at it on a brighter screen it looks like it's in a discrete vertical band on the right, and the blotchy pattern is very reminiscent of other people's problems where unfortunately they had to send the camera in to have an internal board replaced. Hopefully it is still under warranty?

Here are some previous threads:
http://www.reduser.net/forum/showthread.php?97004-Epic-Frame-Artifacts-Corrupt-Frames
http://www.reduser.net/forum/showthread.php?90360-Big-problem-artifacts-in-the-picture-red-epic-x
http://www.reduser.net/forum/showthread.php?96490-Strange-Artifacts-from-drive-transfer

I determined it by viewing it off the mag from the 5 inch touch screen. I can actually clearly see it there. Yeah I already checked out all those previous threads. I was curious though with those cases if it happened all the time or just randomly.

Unfortunately my scarlet is out of warranty and which totally sucks because I've only logged a total of 120 hours on it and been so anal about taking care of it. I did make sure to purchase it on an american express and supposedly they match the warranty. I'm not sure if anyone has had any success with getting something covered for their Red through a credit card company though.l

 

[ericyoung]

Well, no harm in asking how much Red will charge for the repair, and American Express if the will cover it. So the camera is more than a year old?

 

[Mikey Piliero]

Yeah I received the camera somewhere around April 2012. I'm gonna do some rigorous testing to see if I can figure out what may make it show up.

It's strange that recently when it does happen it shows up on like 80% of the footage, and if I don't see it within 5 minutes of shooting it won't appear at all.

 

[Robert Hart]

Just as a wild-card suggestion, does your camera have a facility for providing phantom power to a microphone? If it does, go into menu, select phantom power "Off" and check to see if that affects the artefact.

 

[Mikey Piliero]

what do you mean facility. Is this just an option I should have off on Scarlet if I'm not using a mic? Has it been reported in the past that this causes problems?

 

[Robert Hart]

Mikey.

　
Please forgive my wordstuff. I get a bit complicated saying things sometimes - writer's habit. -

Phantom power when selected "on" with some older non-RED cameras sometimes introduces a faint buzzing noise in the audio, sometimes only to headphones. Sometimes dirty pins are indicated when moving the connecting plug will make the problem go away.

My understanding is that 48v phantom power for camera-powered external mics is produced from a lower voltage source via an inverter circuit. An inverter circuit includes an alternating current stage. I have a vague notion that 48V phantom power may be AC which is rectified in the mic. Hopefully somebody else will put me right on this point.

There have been past mentions of the small fan under the mount introducing faint artifacts into the audio and I think also the image in some instances in the Scarlet/Epic family. Powered lens mounts when a non-servo lens has been mounted have also been reported as causing subtle problems.

If phantom power for camera-powered mics is provided by the Scarlet camera through front mic input, then it might be that if a self-powered mic is plugged in, the mic plug is wired incorrectly for the camera or the mic input is not used at all, that phantom power - if switched "on" might be involved. The Scarlet audio socket is apparently wired to the camera circuit in an arrangement unique to the camera. Incorrect connection of wires of a cable into the plug may cause some problems even if the audio does work.

As there is an AC stage involved, it may be remotely possible that this might interact with the image as a frequency artifact if other things are not right.

I don't know if phantom power is available on the Scarlet via the front mic input. If it is, then it may be user-selectable in the camera Menu.

My comments are not based in experience with the Scarlet, only some other cameras and previous bug reports on this forum. It seems likely that your camera may have to go back to RED. My wild thought was to suggest another line of checking and provoke other replies here before you have to send the camera back.

 

[Patrick Tresch]

Looks like compression / decompression artifacts. Blocking stuff.
Wonder what RED says.
Keep us informed. THX


Pat

 

[Mikey Piliero]

Mikey.

　
Please forgive my wordstuff. I get a bit complicated saying things sometimes - writer's habit. -

Phantom power when selected "on" with some older non-RED cameras sometimes introduces a faint buzzing noise in the audio, sometimes only to headphones. Sometimes dirty pins are indicated when moving the connecting plug will make the problem go away.

My understanding is that 48v phantom power for camera-powered external mics is produced from a lower voltage source via an inverter circuit. An inverter circuit includes an alternating current stage. I have a vague notion that 48V phantom power may be AC which is rectified in the mic. Hopefully somebody else will put me right on this point.

There have been past mentions of the small fan under the mount introducing faint artifacts into the audio and I think also the image in some instances in the Scarlet/Epic family. Powered lens mounts when a non-servo lens has been mounted have also been reported as causing subtle problems.

If phantom power for camera-powered mics is provided by the Scarlet camera through front mic input, then it might be that if a self-powered mic is plugged in, the mic plug is wired incorrectly for the camera or the mic input is not used at all, that phantom power - if switched "on" might be involved. The Scarlet audio socket is apparently wired to the camera circuit in an arrangement unique to the camera. Incorrect connection of wires of a cable into the plug may cause some problems even if the audio does work.

As there is an AC stage involved, it may be remotely possible that this might interact with the image as a frequency artifact if other things are not right.

I don't know if phantom power is available on the Scarlet via the front mic input. If it is, then it may be user-selectable in the camera Menu.

My comments are not based in experience with the Scarlet, only some other cameras and previous bug reports on this forum. It seems likely that your camera may have to go back to RED. My wild thought was to suggest another line of checking and provoke other replies here before you have to send the camera back.

Thanks for the explanation. I'll definitely weigh in every possibility I can before what seems like inevitably sending it to Red... I really can't afford a repair at the moment, that's what makes this is annoying...

I'll keep everyone informed on what they say.

 

[Kaan A]

I had this same issue with firmware 3.3.14 on a couple of clips. I updated the firmware to a newer version and the problem has no yet reappeared. Just out of curiosity, what is the exact version of firmware you are using?

 

[Mikey Piliero]

Really? Well this is some what comforting.

I was using the latest 4.0.12. I have now rolled back to 4.0.10... I haven't seen the problem pop up again, however I haven't done extensive shooting with it because of the lack of free time. Fingers cross that this roll back fixed it.

Thanks for letting me know Kaan.