Dragon Sensor Green Shift

[Kyle Kelley]

Does anyone else notice that the Dragon sensor tends to have a bit of a green shift? Whereas the MX tended to lean a bit magenta, I find that I need to set my tint to about -10 in order to get neutral results. That said, I do like the image from the Dragon quite a bit more than the MX, just wondering if anyone else has noticed this.

 

[Brian Merlen]

With no ND or hot mirror?

 

[Daniel John]

I find with no ND or Hot Mirror that I get a pretty clean image without any real noticeable color shift. However, with filtration, I usually do seem to get a slight green shift. Been using the
Formatt Firecrests and they seem to minimize it more so than other ND's I've tried in the past.

 

[Jeremy Hildebrant]

I usually have a -1.9 tint dialed in. -10 seems a bit much unless you are under fluorescent practicals.

 

[Kyle Kelley]

To clarify, the problem is most pronounced with standard NDs. I was planning to purchase some IRNDs or Hot Mirrors before I upgraded to the Dragon, but was under the impression that the dragon would eliminate the IR pollution magenta shift. It seems that it has done this but has gone in the opposite direction.

And when I'm shooting without NDs, I still find I need to set my tint to somewhere in the -2 to -5 range.

 

[Nick Morrison]

To clarify, the problem is most pronounced with standard NDs. I was planning to purchase some IRNDs or Hot Mirrors before I upgraded to the Dragon, but was under the impression that the dragon would eliminate the IR pollution magenta shift. It seems that it has done this but has gone in the opposite direction.

And when I'm shooting without NDs, I still find I need to set my tint to somewhere in the -2 to -5 range.

What OLPF you on? LLO or SKIN?

 

[Phil Holland]

I would need to see images and/or .R3D snapshots. Ideally information about the light source, lenses used, and a shot color chart is extremely helpful.

 

[Tristan Houghton]

I get the green shift too, haven't done a side by side test with and without ND but have shot with IR ND and a stiffen variable and both time i'm needing to add a few points of magenta.

 

[Akin A]

Dragon using the STH OLPF does have a green tint that needs correction in post. LLO has a slight orange cast.

Also, with either OLPF, the less light that hits the sensor, the more green the tint. You can test this by closing down the aperture. It's pretty simple to fix it in post. But yes, both STH and LLO require adjustments in addition to just setting the temperature, if you want accurate color.

If you notice a green shift with ND vs no ND at the same aperture, it might not necessarily be the ND to blame. It could be just the fact that less light is hitting the sensor with the ND on.

 

[Les Hillis]

I've ended up leaving -10 Tint dialled in as default at times, but it does vary.

Don't want to muddy the waters, but just as a quick example to show when I wouldn't, here are some pics -

0 Tint -10 Tint


0 Tint +10 Tint


Scarlet Dragon
STH-OLPF
5K
ISO250
5600K
f2.8
1.2ND (Formatt Firecrest)

 

[Brian Merlen]

https://www.hurlbutvisuals.com/blog/2015/10/what-is-ir-pollution-and-how-do-you-combat-it-tiffen/

 

[Phil Holland]

Tint

Interesting examples Les. I do actually prefer the 0 tint images on this display.

https://www.hurlbutvisuals.com/blog/2015/10/what-is-ir-pollution-and-how-do-you-combat-it-tiffen/

I really wish I was involved with Shane's test here. Especially since I'm so damn close to his shop.

Short take on this, and I do feel I'm speaking from a real solid place as I've spent literally years now testing various ND filters.

1. Regular NDs have WILD color variation occasionally. Especially when getting past 0.9 densities. 1.2-3.0 (for those who even make those deep filters) tend to be where the wiggle is. I've found Tiffen go green often, some Schneiders go blue. Cheap brands are all over the place. I have no idea how many filters I've tested at this point, but I test them about as hard as I test lenses. So maybe 200 or so. However, I don't feel the need to test at all more recently as solutions are evident.

2. Dragon has built in IR Blocking/Cutting, however, if your filters are tinted in any way color shift is going to show up. Often more magnified than to the naked eye at stronger densities.

3. I personally use Format Hitech (and I guess now Kenko?!) Firecrest ND filters and a set of Schneider regular NDs that I personally hand picked color matched versions of. This is something I learned to do from a few very discerning DPs over the last decade and a half.

I did a write up on my Firecrests here:
http://www.reduser.net/forum/showthread.php?123352-Close-Look-Formatt-Hitech-Firecrest-ND-Filters&highlight=

Specifically take note of the Clear and 0.3-2.7 images:





And these gray patches:



I haven't had to worry about ND filtration under any lighting condition since about one year ago thanks to that technology.

 

[Anton Shavlik]

I've also noticed that using STH OLPF I need a little negative tint, somewhere between -3 and -10 usually. Lightsource, ND and it seems ISO do all have some effect on what tint will look correct.

 

[Ulf Krentz]

Interesting examples Les. I do actually prefer the 0 tint images on this display.

Same here, tint 0 looks best to me. How is your display set up, is it calibrated to judge color? I find the camera calibration to be quite consistent even between different brains. And looking at your tests I see this reflected again. I usually get all color settings done in front of the camera (including color temperature filtration), feed the sensor with with its native color and don´t need to alter the tint settings, but I´m picky with my light sources.

Cheers

Ulf

 

[Brian Merlen]

Phil have you tried the built in Arri ND on Amira or mini and what do you think of those? Just curious based on my cited link and trying to educate myself more