F35 ACES workflow, IDTs. LUT, etc

I suppose I like the ACES route more than film print emulation as a starting point for the flexibility. But definitely something to choose on a per-project basis. That clip was with an Arri Alura and the Hyperdeck. The Kodak LUT is the Rec709 D55 one built into Resolve 10.
 
They look like revised ACES 10bit and 12 bit transforms for the F35.
Software developers can integrate these ctls (color transform language) into their software programs.
 
So - quick recap, since Resolve doesn't have custom/updateable IDTs, we should set the project up as ACES Log (or the other ACES option) and then instead of selecting an IDT, we choose the 3D LUT listed above and set the appropriate ODT?

Clearly ACES is waaaay different than my previous YRGB with a SLog to Cineon 3D LUT, because even without correction applied the clips look very different. Dark and super contrasty and super saturated. Is this normal or do I have a setting off?
 
Film Scientist, nice article and implementation of various tools to allow working with ACES on the F35.

Here's a question I love to ask to those who've done their homework on the concepts behind ACES and perhaps practicality of working with it - realistically how viable can it be to use ACES to unravel camera unique color science into one that is "universal" when no camera specific IDT can account for the color science of the lens(es) used in front of the camera?

We all know glass has unique properties and characteristics of its own. ACES doesn't take that into account.
 
CaptainSKA said:
So - quick recap, since Resolve doesn't have custom/updateable IDTs, we should set the project up as ACES Log (or the other ACES option) and then instead of selecting an IDT, we choose the 3D LUT listed above and set the appropriate ODT?

Clearly ACES is waaaay different than my previous YRGB with a SLog to Cineon 3D LUT, because even without correction applied the clips look very different. Dark and super contrasty and super saturated. Is this normal or do I have a setting off?
Click to expand...

Hi Captain Ska, in Resolve I typically work in regular ACES, not ACES log. Since Resolve does not have a built in ACES IDT for SLog/F35, I roll my own through OCIO and bring it in as an input LUT. It simply serves as a starting point for color correction. Once I balance for white, all the colors begin to line up, and it looks very close to what my eye saw when I shot it. If ACES is viewed without a proper gamma curve (as linear data) then it can look super contrasty and saturated.

Hi Dennis, I'll try answer your question.
ACES is just a really big bucket to perform color transformations in. It guarantees there is no information lost throughout the image manipulation process because the 'number of crayons' is so big. ACES space encapsulates more than what our eyes can perceive and more than what any camera is capable of capturing. An IDT is basically an inverse transform of the gamma curve and color science applied by the camera. Ultimately the encoded/recorded signal is the sum of the entire system (light, scene contrast, lens mtf, camera sensitivity), each photosite/pixel simply averages out to a point value of RGB, which is translated into linear data (a discrete number) as its' converted to ACES. Think of each lens or camera as a filter that reduces the accuracy of the signal captured. What this means is that better cameras will be able to capture more accurate data, and the lens is just another filter that may or may not allow it to perceive a certain color (RGB number). The information captured by s*%tty cameras won't map into larger numbers (values) because they are not there (encoded). Really bright reds, or more subtle cyans are simply the result of capturing more refined numbers that are mapped onto the linear scale, and working at a linear scale is what allows everyone to agree on the same color red (in cg, in real life, and what the camera captures).
 
Last edited:
You must log in or register to reply here.
Back
Top