Dragon Non-Linear Color Response?

[Gavin Greenwalt]

With all of these questions I find it difficult to blame the camera for the skin tones I see on Netflix. There are too many other hands in the soup. I don't doubt that there is an issue with makeups out there, but to assume that we are seeing it in the final product of all of these programs casts doubt on the professional skills of a whole lot of people. Many DP's, Directors, Producers and Colorists would have to be screwing the pooch for that to happen. It doesn't seem likely to me.

Certainly there are many places for this to be introduced. One of the easiest is just through keying. A little despill gone awry and presto wonky skin-tones. There is a lot of really bad skintones in The Amazing Spiderman, particularly in Spiderman's house. That also reinforces the Tungsten theory.

I've seen it in raw footage though too. So it's certainly a color science characteristic. I'll be sure to save a r3d grab next time I see it come through my desk. I don't expect to see it often though--whenever I'm involved in pre-pro I always request to shoot daylight with RED (budget allowing) since it's so much cleaner. I've never seen it in daylight temperature lighting. So whether that's the high CRI of most daylight fixtures or color science I can't say.

 

[Eric Haase]

Points all well taken Hrvoje.

All I can say is we got green skin on a £2 sterling budget 5 week shoot with a very experienced make-up crew.

If you have seen this it is really obvious. It's on the LCD, not visible with the naked eye.

I would compare it to IR pollution in that respect (and I have suspected it may be related). The camera is seeing something the naked eye cannot.

Yep. Definitely feels like it's falling mostly on deaf ears. Oh well.

The makeup list I posted wasn't meant to say that the entire brand is problematic, but rather that in that instance there was a tone in the brand that caused a problem. Was meant as a starting place of further investigation. This forum is so frustrating. I was reluctant to post that list.

As Ivan and I have said, when you encounter this problem, you'll know it. Although maybe not- I've seen some hideous examples of this on the air on some Red MX shows- maybe the DP didn't notice- I don't know. I wish I could post examples but the work I have that shows this problem is with celebrity talent who would definitely not be happy with a green faced picture posted on a public forum. I'll try to gather original R3D files and send to Graeme.

 

[Hrvoje Simic]

Yep. Definitely feels like it's falling mostly on deaf ears. Oh well.

The makeup list I posted wasn't meant to say that the entire brand is problematic, but rather that in that instance there was a tone in the brand that caused a problem. Was meant as a starting place of further investigation. This forum is so frustrating. I was reluctant to post that list.
.

I don't see the reason for frustration.

It's falling on nicely hearing ears, the part between them is suggesting more firm premises for deduction.

If the list is to be relied upon, it is best to compare experiences and use more scientific approach by getting confirmed and repeatable results. With so many variables it makes sense to remove them out of the equation before making any firm conclusions.

If this would end up as a list alone, I don't see how checking every make-up artist's arsenal prior to the shoot would make sense in the long run. In my experience talents tend to have preferred make-up artists, people who they trust and who use preferred make up, so I'd like to avoid having to influence that part or having to wonder about that factor as well, especially if the cause of issues can be more precisely diagnosed. If we are doing this then it makes sense to get to the bottom of this, otherwise this continues as juggling unknown variables, which are bound to pile up eventually, further complicating things later down the road.

Regarding this particular example, if someone with new sensor and OLPF can meet someone with Alexa and they can spend an afternoon with an experienced make up artist and another person to model, a large amounts of time and guesswork would be unnecessary. Single light source for few main types shouldn't be a problem to arrange.
I'd bring a few DSLRs too.

 

[Ivan McCullough]

I don't see the reason for the frustration.....


Regarding this particular example, if someone with new sensor and OLPF can meet someone with Alexa and they can spend and afternoon with an experienced make up artist and another person to be in front of the lens, a large amounts of time and guesswork would be unnecessary. Single light source shouldn't be a problem to find.

Or how about RED get on this themselves as its their camera(s)?

Its only been a couple of years ( sarcasm intended ).

Again, people do not need to use RED's cameras, RED need people to use their cameras. Its RED's job to give people little or no reason to say no when their camera is suggested.

On the Job I was on the majority of the crew were UK crew, they had not worked with the RED much. This did RED no favours in the UK market.

 

[Scott Crawley]

What you are missing is DRAGON COLOR SCIENCE (and at times, adequate exposure.) ;-)

Don't worry yet.

Ahh Guys,

Am I seeing that Dragon just renders skin greener than Epic?
http://vimeo.com/86458847

What am I missing here?

 

[Scott Crawley]

Yep. Definitely feels like it's falling mostly on deaf ears. Oh well.

The makeup list I posted wasn't meant to say that the entire brand is problematic, but rather that in that instance there was a tone in the brand that caused a problem. Was meant as a starting place of further investigation. This forum is so frustrating. I was reluctant to post that list.

I'm sorry you feel frustrated, and I'm not sure who you are referring to as deaf because I think you are being heard. I for one definitely take it to heart. I don't have your experience with it so I lack information, and I perhaps apply different reasoning but I don't doubt your sincerity at all.

The list you offered is a great starting point. It is required knowledge if anyone is to get to the bottom of this so I say keep it coming. However, basic troubleshooting suggests that there is most likely a complex reaction at root.

If it were just the sensor we would be seeing this more often, and on people wearing no makeup at all. I have yet to see evidence of that. If it is out there I would like to see it.

If it were just the foundation we might see it all of the time, every time the particular foundation is used, or a least we would see it much more often. To my knowlege that has not been the case. Makeup is complex and often multi layered. We do a disservice to makeup people when we try to over simplify and blindly blame foundation alone. That was never my intention, and in fact I never really suspected foundation though it could certainly be a player.

 

[Hrvoje Simic]

Ahh Guys,

Am I seeing that Dragon just renders skin greener than Epic?
http://vimeo.com/86458847

What am I missing here?

A reference.

If your reference is left one, right one is slightly more green.
If your reference is right one, left one is slightly more magenta.

If none of them is a reference you can decide more objectively on which you prefer.





In my experience Alexa has more greenish feel than this.

 

[Elsie N]

I don't see the reason for frustration.

It's falling on nicely hearing ears, the part between them is suggesting more firm premises for deduction.

...

Sound and reasonable advice... that seems to be falling on deaf ears. "-(... for some. Others seem open to variable causes and even a combination of variables.

 

[Ivan McCullough]

Sound and reasonable advice... that seems to be falling on deaf ears. "-(

Ha ha, belittling people is so much fun. no?

In my experience Alexa has more greenish feel than this.

Than which? Left or right? or based on how your eyes see the world?

In my experience RED has a bad rep for doing things that are "not normal" to skin.

And before the high post count fanboys start, read that sentence again.

 

[Hrvoje Simic]

Than which? Left or right? or based on how your eyes see the world?

Than both. Speaking of uncorrected Alexa footage, and not throughout the full range and based on short experience with posted samples from this sensor. It is also important to note that "greenish" does not suggest exact green color but very subtle tonality in that part of the spectrum, which mostly goes away with proper transforms. In Alexa's case I don't see it as an issue and think its color science is deliberate taking into account a couple of things.

I've also had a few bizarre cases of make-up look on faces coming from Alexa and I don't have enough information to make any claims and I don't think it was the camera. Or camera alone.

If anyone is worried about this "greenish" feel from new sensor compared to prior one, R1M had much more greenish feel in first color sciences and turned out with pretty nice colors, and nicer looking skin tones compared to MX, IMO.
And that was with few stops of DR less, 3 times higher compression and 4096 times less precise color sampling.
And Red team has gathered a 7 year experience since then.

In my experience RED has a bad rep for doing things that are "not normal" to skin.

I've mentioned in one of the previous posts the potential reasons for this. I'm all for defining the cause what ever it may be, so I'm proposing more detailed research.

Let's be productive:

- In order to check whether this is a camera issue, test should include a camera from another manufacturer in same conditions
- In order to check whether this is only a make-up issue or make-up + specfic light source, problematic and approved make up list should include the type of light used
- In order to check whether the mentioned make-up is the only issue, or the human factor (make up application process) is also related, at least two different occasions with same make-up type and same light source should be used


If you had the issues you can check about the light and make up on the shoot and have Graeme check those r3ds.

In any case, the more information there is, the easier the issues can be pointed out and hopefully squashed.

 

[Brian Boyer]

Let's be productive:

- In order to check whether this is a camera issue, test should include a camera from another manufacturer in same conditions
- In order to check whether this is only a make-up issue or make-up + specfic light source, problematic and approved make up list should include the type of light used
- In order to check whether the mentioned make-up is the only issue, or the human factor (make up application process) is also related, at least two different occasions with same type make-up and same light source should be used.

I'd also add:

- Account for the light under which the makeup is being applied and vary it to see the results. Even if the camera is ultimately at fault, it may be exacerbated by the application environment. It may mean a lot of repetition, but I don't see the test being comprehensive without doing it this way.

• Applied under tungsten, shot under tungsten and daylight
• Applied under daylight, shot under tungsten and daylight
• Applied under tungsten and daylight balanced fluorescent, shot under tungsten and daylight

- Include a shot of talent without makeup. As you pointed, someone like Jim Broadbent might have naturally occurring yellow tonality to their face to begin with. At the very least, it's a control and will give insight into what the makeup artist thinks needs to be done to the skin.

 

[Elsie N]

Yep. Definitely feels like it's falling mostly on deaf ears. Oh well.

...

Ha ha, belittling people is so much fun. no?

No, just responding to the comment that Hrvoje was responding too. But now that I think about it, I do think you belong to the group I alluded to in the second part of my comment. That is, you seem bound and determined to make this a RED issue when post after post has provided evidence that this is a problem with all cameras... including your apparent personal favorite, Alexa. I'm just suggesting everyone consider that this is a PROBLEM, but let's not selectively place blame on one thing while others have shown that there are examples of many players sharing or being the cause of the problem.

 

[Ivan McCullough]

No, just responding to the comment that Hrvoje was responding too. But now that I think about it, I do think you belong to the group I alluded to in the second part of my comment. That is, you seem bound and determined to make this a RED issue when post after post has provided evidence that this is a problem with all cameras... including your apparent personal favorite, Alexa. I'm just suggesting everyone consider that this is a PROBLEM, but let's not selectively place blame on one thing while others have shown that there are examples of many players sharing or being the cause of the problem.

I don't have a favorite camera Elsie, can't think where you might have gotten that idea. Oh maybe because I am pointing out a weakness of your adopted baby and so must be a fan of her great enemy? gimme a break guy.

If anyone needs to do tests to establish the cause(s) of this issue its RED, why

-people like Elsie etc will not accept results from "independent" sourses because they might have a hidden bias
-even if conclusions can be drawn then RED will have to repeat the tests because of above and its best scientific practice.

 

[Hrvoje Simic]

If anyone needs to do tests to establish the cause(s) of this issue its RED, why

-people like Elsie etc will not accept results from "independent" sourses because they might have a hidden bias
-even if conclusions can be drawn then RED will have to repeat the tests because of above and its best scientific practice.

It would be appreciated if there was no generalization or suggested fanboyism in this case. If anyone fails to admit any proven weaknesses it's their chain of events and I don't think Elsie was suggesting that, or that there is any reason to polarize this discussion. Everyone (hopefully) wishes any issue solved.

I would say that Red already has done and is doing tons of tests. There are so many factors which can affect this, just the make up type X make-up application process X amount of underexposure X various light sources - gives tons of routes to test. I think it is much more effective to focus on documented examples so Graeme and guys can pin point the issue, if it's related to camera alone.

 

[Ivan McCullough]

I would say that Red already has done and is doing tons of tests. .

I would hope so, and now they have a better idea where to look.

Suggesting that individuals go out and shoot tests, then pass on relevant info is silly.

RED make the cameras, RED test the cameras. We use them.

 

[Elsie N]

I don't have a favorite camera Elsie, can't think where you might have gotten that idea. Oh maybe because I am pointing out a weakness of your adopted baby and so must be a fan of her great enemy? gimme a break guy.

No offense, but I'm not sure you have the bonafides to accurately point out a weakness of my Dragon or Epic X.

But I do owe you an opoligy. I thought in one of your posts on this thread you were touting the Alexa, but after re-reading your posts here your primary transgression was to call Scott Crawley ignorant. Maybe you could take my lead and apologize to Scott for making such a juvenile comment. (Oops! sorry for characterizing your comment as juvenile. '-)

If anyone needs to do tests to establish the cause(s) of this issue its RED, why

-people like Elsie etc will not accept results from "independent" sourses because they might have a hidden bias
-even if conclusions can be drawn then RED will have to repeat the tests because of above and its best scientific practice.

Actually, there are established posters here on Reduser whose results I will accept. And even Eric Hasse who seems convinced it is a RED only problem, at least he his working with Graeme to see if it replicates. Rather than cursing the perceived darkness, he is lighting a match.

 

[Scott Crawley]

Actually, there are established posters here on Reduser whose results I will accept. And even Eric Hasse who seems convinced it is a RED only problem, at least he his working with Graeme to see if it replicates. Rather than cursing the perceived darkness, he is lighting a match.

Without saying anything definitive Graeme did frame the issue as two-fold; makeup and color science. That's not to say that one or the other is responsible in total or in part, but that is where the solution will likely be found. At least, that is how I read between the lines. Also evident is the fact that he is both aware and looking into it.

Two issues here - one is raw development which is always under constant R&D for improvements, 2nd is makeup. Best advice for makeup is to always ensure the light used to apply the makeup is the kind of light used to illuminate the talent. I did some testing on makeup and found that it would appear different to the human eye depending on lighting conditions, and that difference was also seen by all the cameras I have here. I also discussed makeup with some top makeup people, and they refused to blame any camera for makeup issues, even with lots of prodding from me.

Graeme

I would also point out that both Graeme and Eric were there in the beginning, actively engaged in the discussion of Axel's problem back in 2011. Red is not ignoring this (Or Eric) at all.

Edit 6:55 PM EST
This might be a bit of a reach, but the fact that Dragon and the new OLPF no longer require special IR treatment may just speak to this issue both in terms of the makeup problem and the OP's questions. My money is on the new color science and the likelihood that it addresses both concerns in a substantial way.

 

[Gavin Greenwalt]

I would say that Red already has done and is doing tons of tests.

Yeah, but RED has had lots of problems in the past that they didn't own up to. Even if there was a pervasive problem with the RED sensor they will be the last ones to tell us. It's generally poor marketing to enumerate their flaws.

Autodesk complained that they used to post the entire bug lists of their products--but had to stop because their competitors were pointing at their transparency as proof of failure. Companies aren't going to admit to defects unless legally obligated.

 

[Eric Haase]

Yeah, but RED has had lots of problems in the past that they didn't own up to.

in a word, Redmote.

 

[Scott Crawley]

Yeah, but RED has had lots of problems in the past that they didn't own up to. Even if there was a pervasive problem with the RED sensor they will be the last ones to tell us. It's generally poor marketing to enumerate their flaws.

in a word, Redmote.

So the motive here is what... strictly punative, or are we really interested in good color? Are some of us looking for capitulation or color science?

http://youtu.be/sglyFwTjfDU

Sorry, I'm on a bit of a Python jag today.

But now I am wondering... does the RedMote weigh the same as a duck?