REDcolor3 and REDgamma3...

[Zack CC]

Ugh.

Ugh.

I'm sitting on set right now doing all this color work, and I just know that on the last day of shooting Graeme is gonna drop some mad science on us all, and I'm going to want to go back and regrade the whole shoot... XD

 

[Rob Ruffo]

I'm sitting on set right now doing all this color work, and I just know that on the last day of shooting Graeme is gonna drop some mad science on us all, and I'm going to want to go back and regrade the whole shoot... XD

True that!

 

[Lewis-M Soucy]

How did you do that? Looks amazing!

 

[Fred Beahm]

I guess clicking refresh every five minutes doesn't make it go faster... Can't wait to try RC3/RG3

 

[Rob Ruffo]

I guess clicking refresh every five minutes doesn't make it go faster... Can't wait to try RC3/RG3

Ha Ha! I know the feeling! Remember what they say about a watched pot and how slowly it boils.

I would happily pay $5000 for this, and it will be free!

 

[Les Dittert]

What are people thinking this will 'do' ?
I see it as a slight modification of the color space, an application of a mild 3d lut. Is it more than this ?
I didn't see a post showing otherwise.
I don't see it changing the metamerism problem with skin tones, as that is a sensor spectral response issue.

Ha Ha! I know the feeling! Remember what they say about a watched pot and how slowly it boils.

I would happily pay $5000 for this, and it will be free!

 

[Ian B. Johnson]

I don't see it changing the metamerism problem with skin tones, as that is a sensor spectral response issue.

Could you elaborate on this? Haven't heard the term metamerism applied to video before. Thanks.

 

[Les Dittert]

If you watch the Hobbit / Peter Jackson video where they show some behind the scenes stuff, you will see that they are futzing with the skin makeup to compensate for a color problem.
This is classic metamerism in action. The spectral summation of the lights/reflective surface ( face) /sensor is not adding up the same as what they are used to. They modified makeup to compensate.

Could you elaborate on this? Haven't heard the term metamerism applied to video before. Thanks.

 

[Craig Parkes]

If you watch the Hobbit / Peter Jackson video where they show some behind the scenes stuff, you will see that they are futzing with the skin makeup to compensate for a color problem.
This is classic metamerism in action. The spectral summation of the lights/reflective surface ( face) /sensor is not adding up the same as what they are used to. They modified makeup to compensate.

Les, that issue has been clearly and accurately been explained to be due to the color offset due to the mirror rigs for 3D, not the sensors on the Epic.

 

[Bob Gundu]

Explained by Jarred Land; "The Extra RED makeup has nothing to do with the camera color science... the mirrors in the 3D rig desaturate magenta from one eye ( tints green ) so instead of boosting back saturation in just one eye in post, they prefer to lift both eyes in saturation then remove it as necessary."

"The exact same thing would happen to any camera through the 3D rig beamsplitter. one eye goes green , one eye goes magenta. Many ways to compensate for that, this is just one."

"In regards to Peter Jackson and this video, and the way he is choosing to saturate not only makeup but wardrobe and sets, is to compensate for a saturation change coming through the beamsplitter. "

If you watch the Hobbit / Peter Jackson video where they show some behind the scenes stuff, you will see that they are futzing with the skin makeup to compensate for a color problem.
This is classic metamerism in action. The spectral summation of the lights/reflective surface ( face) /sensor is not adding up the same as what they are used to. They modified makeup to compensate.

 

[Les Dittert]

That's good news, I missed that....
Thanks!

Les, that issue has been clearly and accurately been explained to be due to the color offset due to the mirror rigs for 3D, not the sensors on the Epic.

 

[Nir Shelter]

FYI when ever splitting a beam your bound to reduce the percentage of certain frequencies differently on the other end of the medium, nature of the beast can't argue with physics.
As a rule of thumb in most cases the reflection of green light i.e ~500-570nm sits at around 50 - 55% and magenta i.e ~380-400nm sits at around 40% where as the transmission of green light sits at around 50% and magenta at around 60% therefor the disparity between the color of the eyes.

Nir.

 

[Wayne Morellini]

Sweet!

Got to say that working with these Gamma look applications provide a great view for the client as well as positive results for media capture on set as potential offline content and even online media on projects that don't require or have budget to inline with RAW media!

Clients have been amazed that this is just the low resolution footage and it looks pretty great without any one-lighting etc. Looking forward to Gamma3, and thanks for your continual support of this awesome product.

Below is a recent frame grab that I think shows off your EPIC-X sensor... amazing job in immense detail throughout the exposure range... maby 14 stops? I thought the sky's and range there was spectacular. Thanks for that! These are dumbed down for web.

Be Well,

Zach thanks, the untouched footage is the sort of color I have seen in older Red movies, and the improved looks like the best colorisation I have seen on modern films with these cameras. Can they do better with the colour saturation? I thought it would do better with the sun through the clouds though.

Just going back to what I think I am seeing in many movies, the relatively low saturation, is resulting in flatter color tones in faces also affecting the luminance rendering, as the satuation is lifted. It might be an algorithm thing, is there anyway to improve this. Actually, why is it so desatuated, shouldn't the filter array and software filtering remove more washout in pure primary colors?

 

[Dane Brehm]

Looking forward to seeing it with my RED ROCKET ENABLED workstation on Mavericks

best

 

[Patrick Tresch]

What are people thinking this will 'do' ?
I see it as a slight modification of the color space, an application of a mild 3d lut. Is it more than this ?
I didn't see a post showing otherwise.
I don't see it changing the metamerism problem with skin tones, as that is a sensor spectral response issue.

;-) I think Graeme is paid to do more than just to put a 3D lut.

He actually is doing the colors of RED images.

Pretty huge job to be the Master of colors!

Pat

 

[Gavin Greenwalt]

Les, that issue has been clearly and accurately been explained to be due to the color offset due to the mirror rigs for 3D, not the sensors on the Epic.

That being said, there are a few makeups which RED doesn't like due to metamerism. But that's going to be true of any sensor.

 

[Les Dittert]

No need to tell me of color issues. I was 'Color Czar' at Dreamworks Feature Animation for 4 years, responsible for color matching there.
Those artists were picky. Animation is more critical than live action, as there are more defined pallets.

So, other than a new color LUT, what do you guys think is up with red gamma 3 ?
I can compare it to old red color, and show the difference as a color cube with sparse vectors inside, to help visualize it, if any one cares.
But I'm busy getting the r3d debayer done in a different way right now. RCX needs more detail IMO.

;-) I think Graeme is paid to do more than just to put a 3D lut.

He actually is doing the colors of RED images.

Pretty huge job to be the Master of colors!

Pat

 

[Edward Blythe]

I can compare it to old red color, and show the difference as a color cube with sparse vectors inside, to help visualize it, if any one cares.

I care Les

 

[Gavin Greenwalt]

No need to tell me of color issues. I was 'Color Czar' at Dreamworks Feature Animation for 4 years, responsible for color matching there.
Those artists were picky. Animation is more critical than live action, as there are more defined pallets.

So, other than a new color LUT, what do you guys think is up with red gamma 3 ?
I can compare it to old red color, and show the difference as a color cube with sparse vectors inside, to help visualize it, if any one cares.
But I'm busy getting the r3d debayer done in a different way right now. RCX needs more detail IMO.

Well RC3 is more interesting to me than RG3 but I assume RG3 is going to be an adapative 3DLUT like FLUT that morphs based on the input range.

 

[Roberto Lequeux]

(...) the relatively low saturation, is resulting in flatter color tones in faces also affecting the luminance rendering, as the satuation is lifted. (...)

When you put it that way it seems to make a good deal of sense to me... Not sure why though...

Of course we haven't seen RC3 & RG3 yet, but it would seem they will change just that. Apparently, they are slated to provide a more saturated and contrasty base out of the box, though probably also providing as much/almost as much grading malleability..?

Well, we shall find out soon enough. One thing is certain: This is going to be a big step in the right direction. Wonder what Graeme will do next..??!? Starting to hit the ceiling... Probably work on ways to allocate the additional DR from DRagon so that it rolls highlights more than anything else?

Are we there yet? Can we see samples?