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 :)

ELYSIUM: a conversation with the DI colorist, ACES mentioned

Brett Harrison

Brett Harrison

Well-known member
Joined
Oct 15, 2012
Messages
1,083
Reaction score
0
Points
36
Location
Melbourne, Australia.
Sidney L. Plaut said:
although M Most is right about Oblivion, the film here mentioned, is " Elysium"...:)
Click to expand...

Yes, but the question was "is anyone else working with ACES and Red?". I took that to basically be two questions, neither of which involve Elysium.
 
M Most said:
Yes, but the question was "is anyone else working with ACES and Red?". I took that to basically be two questions, neither of which involve Elysium.
Click to expand...

Elysium is shot on RED Epic and according to the article, the DI was done with ACES color.

Perhaps Mike has glanced at 'Elysium' and read 'Oblivion'.

I should have been clearer: 'working with ACES and Red' should be 'using ACES in conjunction with RED'.
 
Last edited:
Brett Harrison said:
Perhaps Mike has glanced at 'Elysium' and read 'Oblivion'.

I should have been clearer: 'working with ACES and Red' should be 'using ACES in conjunction with RED'.
Click to expand...

No, I read exactly what you wrote. Since ACES has been discussed on numerous threads here, I thought you might be referring to "theatrically released pictures that went through an ACES pipeline," so I responded accordingly. If you were specifically referring to ACES in conjunction with Red, I would say that Red has rather actively discouraged that, and as a result, I don't know of any pictures that have gone that route, except for Elysium. I'm not sure exactly what they were trying to accomplish in going that route, but it might have to do with wanting an open format for archiving that is less restricted than a 10 bit DPX file. I'm just guessing on that, though.
 
As Mike had mentioned, for reasons unknown, Red has not welcomed ACES with open arms. In the past there was a mention of a technical problem affecting specifically Red material using ACES color space. Additionally, there is a significant bug affecting Resolve working in ACES color space. It has to do with not being able to use curves without severely clipping the signal. One can work around it, as I just did, but it's a bit annoying. So far, I haven't seen mention of a fix for this problem from BM. Hopefully it will be taken care of, as otherwise ACES shown, at least to this user, great promise in greatly simplifying the modern workflow, especially while using different cameras.
 
The concept underlying IIF/ACES is all about codifying a digital container that accommodates virtually any image type without truncation and can be interchanged using known values. Achieving true fidelity in the pipeline requires precision transforms per device in the chain - IRTs for each camera/sensor, a completely neutral RRT and ODTs that compensate for inherent display characteristics.

So far, the IDT for a RED MX RC3/RG3 source has not been a good match for the RRT, but AFAIK the RRT is still being tweaked. I am not a member of the committee working on this so I can't provide a proper progress report. That said, it seems likely that either:

1) the RRT will evolve into more of what its intended to be - completely neutral, or

2) RED will a mint a new IDT for ACES matched with the character of the "release version" of the RRT

Cheers - #19
 
Thanks Blair. I did a bit of reading on Charles Pynton's site and came to this article:

http://www.poynton.com/w/ACES/Poynton-ACES-AbelCine-Deluxe.pdf

he talks a lot about the advantages to CGI which would seem to be a primary justification for using ACES in this case.
 
"So far, the IDT for a RED MX RC3/RG3 source has not been a good match for the RRT, but AFAIK the RRT is still being tweaked. " - RC3/RG3 or any colour or gamma space has NOTHING to do with the RRT (or IDT for that matter). For RED there is effectively NO IDT anyway because in our pipeline our data starts off in linear light (so no transform is needed to get it to linear light) and our first step is always to convert to XYZ and the transform from XYZ to ACES is standard and we've checked and double checked that we've implemented that aspect correctly and to spec. The only stage after that is to dump the linear light ACES data to EXR and again, there's no problems there.

I've been through a good number of RRT iterations with the Academy. If there are any issues with ACES they lie in the realm of the RRT and that's a ball that's entirely in their court.

So yes either 1) as you mention, the RRT will actually become neutral (whatever that really means) or people will "roll their own". Your 2) is based on false premises as outlined above.

Graeme
 
ACES is still a work in progress.....Let them actually deliver a final "shipping" version and we can have an opinion on that.

Graeme
 
I got a chance to listen to Andrea's 50 minute interview on the ColorCast podcast. ACES was apparently used as a way to collaborate with Image Engine to secure look management between facilities. Andrea opined that there hasn't been such a close collaboration between a DI facility and a VFX facility on a technical level. They also live one block away in Toronto (edit: Vancouver). Just 8 people worked on the DI.

Thank you Graeme for clarifying about ACES - appreciate the directness of your answer.

Mike, upon reflection I completely misconstrued your comments so apologies for that.
 
Last edited:
Sorry to necro-post....could have sworn I saw a newer thread for this topic but whatever. Did he just say Pan and Scan version? :O Anyways Great article...I posted the same question on the article BUT perhaps one of you can answer my question: Was interesting to see they shot anamorphic and instead of going from 3296x2700 to 6592x2700 then "down razzing" to 4K for the edit they just went to 3296x1350…I mean in the film days anamorphic got you more resolution and I understand on a digital sensor of you don't get as much resolution gains (if any) but maybe someone can enlighten me? Having never worked with anamorphic Epic footage or even anamorphic digital, does anamorphic digital produce unwanted artifacts when de-stretching or something? I find anamorphic really interesting and hope people continue to use it.
 
Graeme Nattress said:
ACES is still a work in progress.....Let them actually deliver a final "shipping" version and we can have an opinion on that.

Graeme
Click to expand...

Interesting stance coming from Red. My color company was about to embark on an ACES workflow for an Epic shot feature and our only option would be to transcode the footage to an open EXR format first. The Red IDT of the r3d has some serious flaws which makes it unusable. Maybe the fcp plugin maverick has met his limit? ACES and Red remain incompatible.
 
Florian, the math for translation of R3D to ACES was done with the assistance of the ACES team at the Academy and they checked and verified it was correct.

Graeme
 
You must log in or register to reply here.
Back
Top