Welcome to our community

Be a part of something great, join today!

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 :)

EMERGENCY - Scarlet Footage Problem (on location now)

chris morgan

chris morgan

Well-known member
Joined
Nov 23, 2011
Messages
84
Reaction score
0
Points
0
Location
Venice, CA
hey guys...

i'm on a set where my scarlet is being used and i just got a redmag of footage back from the DP and AC and there's a major problem with some of the shots. the two links here are from the same setup, lens, and everything... but as you'll see one of them is soft and basically useless.

when the DP was metering, he was having a hard time getting the stoplights to show clipping - which was odd, as we're shooting in bright sunlight.

i now have a mag (and an hour's worth) of footage that's useless. does ANYONE have any idea what's going on? stuck in an area with barely internet, and about to go drown myself.

here are the shots:

http://www.darkenergymedia.com/red/bad.png

http://www.darkenergymedia.com/red/good.png

PLEASE help.
 
Sorry Chris, your links don't seem to be working. Can you try uploading the shots again?
 
hey will / all

sorry, had to re-upload the images... the faces are blacked out on purpose, obviously... not the problem... they should be up there now...

thanks!
 
Okay see them now. I must have tried downloading when it was still uploading. The bad one looks like a "Raw" image versus the "Medadata" version. If that's not it, did you "Black balance" before the shoot? The image might be salvageable with some color correcting.
 
What lenses are you using? Did the lens get swopped between takes? How is focusing being done?
 
hey guys...

both shots are the same lens, same shot, literally within seconds of each other. i was with the dp when the problem "corrected" itself and we got the "good" image you see.... but then they shot an entire mag without me being there and EVERY take exhibits the issue the first shot has.

i did witness the image "pop" into the correct colour balance and sharpness in the above example, which was odd, but the rest of the shots have had the problem.

we are using vintage takumar lenses on canon adapters with an EF mount. we're also using a Tokina 11-16 which doesn't exhibit the problem.

...running back into the lion's den, will be back. appreciate you guys taking a look.

cheers,
c
 
So, you only have the problem when using the takumar lenses with the canon adaptors and not with other lenses? If you can't reproduce the problem with other lenses then it's likely the adaptor/takumar lenses. Why would the DP keep shooting if he saw there was a problem on the monitor? Did he not shoot a test and play it back to see if the footage was bad?

Either way, make sure you save your log files so RED can take a look if it proves necessary.
 
just back to point out again that when the problem was exhibiting itself, it seemed impossible to overexpose the image - almost as though the sensor wasn't understanding that it had too much light - which i think is the crux of the problem, if that makes any sense.
 
probably cleaned the lens between shots, pulled out a dirty ND filter, who knows... sure they won't admit it.
 
The lens adapters are not good quality & something moved. What monitors are you looking at whilst shooting, this should be obvious......
 
i had the same with may tamron 17-50 f2.8 lense...the apeture was set by sidehandle to a very hight setting. this happend with one of the first OS 3 releases. Now its fixed. With 3.3.18 problem is solved :)
 
ok... maybe i'm not explaining myself well here. =)

the two shots you're seeing are within SECONDS of each other, while i was watching. no lens changes, no lens cleanings, nothing. the image "snapped" back into proper colour and sharpness seemingly on its own.

the DP is unfamiliar with red, though the AC is familiar... granted this is something that should have been caught but this is a pretty tense shoot as it is with a tight schedule (hence me being ready to hang myself at this point.)

i'm not ruling out a lens adapter issue, but i've never seen this before when shooting with these lenses MANY times.

it's also worth mentioning that this is a high heat / humidity environment. the camera has been shaded and i'm running the fan full all the time as we're MOS.

will report back more when i know more.
 
You might want to peek at the metadata in REDCINE-X to look at the metadata of the bad shots. I had this sort of scares previously when my picture turned pink and desaturated... turns out the metadata was whacked (very different from what I set in camera). Resetting all the metadata to default did calm the nerves. The bad shot does look a little bit as if the Clarity slider has been pushed way too left...
 
uploaded another image that's showing the problem a little differently. you can see that there's a purple cast to the image in this one... so a similar problem.

http://www.darkenergymedia.com/red/bad2.png

i'm realising now that there's not much point in figuring it out at the moment - i'm now on set and monitoring things personally and there haven't been any problems.

i appreciate you guys taking a look... and will check in again soon.

cheers,
c
 
looks to me that on the bad/good shot the aperture was opened (out of focus is more distinct) and they dropped in another ND - you got heavy IR pollution here - can be fixed, but not easy......
 
To me it looks like too strong ND or POL filtering of the wrong kind - IR pollution.
Most probably not the camera´s fault, unless there was a malfunction.
 
I think I would second the diagnosis of infrared contamination from forcing shallow depth-of-field using strong ND without IR filter.. Because of the long wavelengths, the image will soften and be impossible to focus. It will be correct only for visible light or for IR light but not both together.

Other seasoned RED veterans please correct me here as I running with dementia memory here. I understand maybe wrongly that the appropriate filter is the Schneider Optics IR786. Other users here have reported a preference for using a "Hot Mirror" filter. The IR786 also works in a nicely subtle way with the SI2K to sweeten the greens and pale yellows just a little more towards natural.
 
chris morgan said:
ok... maybe i'm not explaining myself well here. =)

the two shots you're seeing are within SECONDS of each other, while i was watching. no lens changes, no lens cleanings, nothing. the image "snapped" back into proper colour and sharpness seemingly on its own.
Click to expand...

Maybe it's worth repeating again....
 
If it's indeed something happening with literally nothing physical or optical being changed, then it would point to a camera fault - but then the image in the LCD or EVF would surely look magenta and out of focus too?

Basically need to recreate the fault to start eliminating possibilities, but definitely check to see if there is any odd metadata settings in the bad clips compared to the good ones.
 
every now and again in life, you learn lessons the very, very hard way.

this post was originally posted out of sheer panic and helplessness in the middle of a pretty grueling day (including seat-of-the-pants drone shots among other things.)

a 45-minute air conditioned van ride allowed everything to click in to place.

of course it's massive IR contamination. we were shooting in bright sunlight with no IR filtration whatsoever.

this is very much an asshat / derp / self-immolation type situation. i did briefly mention to a few people involved that IRNDs might be a good call, but nobody seemed too worried, and i'd never experienced contamination to this level before.

what tipped me off was reviewing some of the footage in my head, a few clips of which included foliage, and i remember now that the foliage exhibited the same purplish cast. which, of course, it would. even my clumsy forays into shooting with IR-modified 5Ds taught me that meatpuppets and plants reflect massive IR.

what was throwing me was the first shot i posted, which seemingly "clicked" into being OK. what i now realise is that likely someone was adjusting a bounce (all our lighting was via bounces and reflectors, thereby compounding the problem), or some other lighting condition changed and lessened the amount of IR spilling into the lens at that exact moment. hence suddenly having a shot that worked.

so now i've got the task of trying to correct this stuff. the other IR artifacts of slightly glowy skin and such might actually work OK... but i'm not sure i can pull this stuff back.

needless to say at this point, this was a very, very painful lesson learned. truthfully, the DP or AC should have caught this as it was happening as well, but again this was an extremely high-pressure shoot.

so i'm gonna suck it up and take the heat for it. luckily i'm the guy doing post on the whole piece, and i've got a couple of months to work with.

man, today really sucked.

and hey, the less "well no s**t, dumbass!" you guys dish out, the better. i'm already beating the hell out of myself for this.

what doesn't kill us makes us stronger. blah.

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