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 :)

Rolling Shutter Fluorescent?

A

abook

Member
Joined
Apr 23, 2008
Messages
19
Reaction score
0
Points
0
We shot a very quick test in a Hallway for colour correcting purposes and found what we think to be a rolling shutter issue. The old fluorescent bulbs in the hallway created bands on the raw file as two different shades of Green (watch the walls) The posted clip below has the problem outlined (beware the off focus). Notice the bands move with the handheld shot, not the bad paint job of the walls. When the daylight shows up, the band disappears. Thoughts?

http://www.mediabutton.com/proofs/misc/red_band.mov

abook
 
True, but usually it's a flicker effect. Have you seen the banding before?
 
I understand that, but what I'm worried about is if I change the shutter or rate the tubes (basically rate for the frequency of the bulbs), the CMOS still scans one frame from top to bottom, not as a single frame of emulsion. The same as with flash bulbs creating a straight horizontal line when the flash goes off on the red. With a film shutter, it is either open or closed, so you either see the bulb flash or you don't, you see the flicker or you don't.
 
common problem even on film

common problem even on film

Evin Grant said:
This is a common problem even on film. You can try to change the shutter speed to minimize it but your only total solution is to replace the fixtures with flicker free ones.
Click to expand...

This camera should overcome film problems, it is suppose to be advanced, otherwise we are going in circles.
I am sure problems like that will be solved in the future.
Question is when.
 
i've been testing RED with various strobes and flashes and barely notice the rolling shutter artifacts, if at all. I do see it with flo's a little bit.
 
Greenish gels on tungsten lamps...

Greenish gels on tungsten lamps...

abook said:
When the daylight shows up, the band disappears. Thoughts?
Click to expand...

Greenish gels on tungsten lamps can full up the down time on the fluorescents, and maybe shoot with a FL-D or FL-B on the camera.

Fluorescents are not the same color all the time, they flash bluish light from the Mercury vapor and then glow yellow from the phosphor, you can see this if you shake you fingers from side to side quickly, you see yellow and blueish lines.

I think I remember reading that ILM at one time used 3 phase power so 1/3 of the light tubes were on a different phase, that made the light more even, maybe.

Best to frame the fixtures out of frame, or over expose them and use lots of fill filtered to match the fluoresents to Drown the flicker and bars.

Its worse if you are shooting at 23.976+ rather that 24.000 since the bands move.

Another idea would be to put Christmas tree lights in diffuser in the fixtures if they have a diffuser cover, and leave the Fluorescent tubes out...
 
Stoopercraft said:
The artifacting you are seeing has nothing to do with the shutter, it is a compression artifact. It's either RedCode or the way it was compressed somewhere along your post flow.
Click to expand...

I know it is hard to see 4k in a small work station situation but think about the way the CMOS works. Scanning pieces of the frame till the end. Not a single frame like film, which I find better. I'm going to do some more tests, but I'm pretty really sure it's not a compression thing. But again, more tests. I think everyone should be aware that it may be happening during the scan and that there are limitations to the RED. It definitely has unforeseen issues.

abook
 
Thanks for all the posts, I'll just have to do more tests. I don't think I can explain it any differently but we trouble shot post workflow. The day light is kind of a dead give away too, but I'll work on it.

abook
 
Miltos Pilalitos said:
I just saw the quicktime and like the others didn't notice any rolling shutter problems. The banding looks like a compression problem to me too.
Click to expand...

Agreed, I don't see any sign of the rolling shutter, the banding certainly looks like a compression problem. Any chance of seeing a 4K still or two? Maybe post the R3D up?
 
Ya for sure, I'll see if I can get something up tomorrow.
 
I see some specular blue from the door reflected in high gloss wall paint.
It appears to chromatically affecting the walls colour.
Your eyes might not see this.
Camera may see this.
$0.02 worth.
 
The banding I saw was indeed the same as I've seen from bad florescent ballasts.
 
gdv said:
This camera should overcome film problems, it is suppose to be advanced, otherwise we are going in circles.
I am sure problems like that will be solved in the future.
Question is when.
Click to expand...

gdv... did you really post this?

Jim
 
abook said:
Scanning pieces of the frame till the end. Not a single frame like film, which I find better. I'm going to do some more tests, but I'm pretty really sure it's not a compression thing.

abook
Click to expand...

When a film frame is exposed on a 'normal' film camera, the shutter wipes
across the frame to start the exposure and then the trailing edge of the shutter wipes across to end the exposure. With a short shutter angle, it's more like a trapezoid that wipes across the frames. A single frame is not exposed all at once but is exposed with the shutter spinning across the frame.

Our shutter works in a similar way in that the 'blades' of the shutter are equivalent to the start of the exposure and end of the exposure. The main difference is rate at which the blades move across the frame.

Deanan
 
abook said:
Thanks for all the posts, I'll just have to do more tests. I don't think I can explain it any differently but we trouble shot post workflow. The day light is kind of a dead give away too, but I'll work on it.

abook
Click to expand...


Maybe do a test where you pan up and down, showing that the scan either follows or does not follow the frame. Also try one locked off. It looked to me that at the end it seemed to have moved up, which would be a phase issue, thus a rolling shutter issue. Like when you try to film a computer screen and there's that hideous black bar that scrolls across.

I don't know that much about the rate at which fluorescents flicker, but it seems to me that the anomoly shows up in the top, disappears and then reemerges further down the frame, as if when the rolling shutter moved across the frame, it caught a part of the ficker and then lost then caught it again.

Just my thoughts, def needs to be further tested though..
 
You must log in or register to reply here.
Back
Top