-
EasyDCP creator and DPX Color Spaces
So, I’ve been testing EasyDCP creator to see about doing some limited DCP mastering here and I had a question.
I made a test reel (which I’m happy to share with folks) which outputs every format that Da Vinci can do that works with EasyDCP (Tiff and DPX, RGB, XYZ, etc.) then goes through the EasyDCP process at all available settings.
Then I went to my old school (USC), which has a calibrated DCP theater (which they are super finicky about and I trust) and screened my DCP. For good measure I’m going to screen at AFI through another friend to be sure.
My Da Vinci suite is set up for Rec. 709 grading, so I was expecting something like TIFF Rec 709 converted to XYZ to look the most accurate to how it had looked in the grading suite. But we watched it twice, and I swear, the most accurate setting was to render DPX 16bit out of Da Vinci and then transcode it using EasyDCP with EasyDCP set to sRGB as the source codec and transform it to XYZ. Looked spot on.
What I’m trying to figure is, why? Does Da Vinci write it to sRGB DPX files, doing a bit of an image transform on it? I had always kind of assumed that when writing DPX from Da Vinci (something I don’t often do) it would write Rec. 709 DPX files, but it doesn’t seem to be doing that.
I do tend to do DPX for FX work (our in-house FX guys like working in frames, makes their rendering easier), then master it back to ProRes444 for broadcast delivery. So, is that going through a 709-sRGB shift and then back again?
I’ve done before/afters (shot with/without FX), sometimes cut against each other (for our promo reels and whatnot) and there isn’t an appreciable shift, so it seems that if there is a color space change going on there, it’s Da Vinci is pretty good at matching it both directions.
Charles
If anybody wants some info on EasyDCP, found it here, from the Tao of Color newsletter:
https://www.thefilmbakery.com/blog/creating-a-feature-film-dcp-using-opendcp -
6 Replies
-
Your last sentence got me puzzled. Are you using easydcp or opendcp?
To my knowledge the two colorspaces you are talking about are almost identical so that could be why your seeing such good results with those settings.
Furthermore I also tested Resolve’s various outputs and also concluded that it doesn’t do a proper rgb to xyz conversion for DCP usage.
Editor/Colorist, Amsterdam, The Netherlands
http://www.pepijnklijs.nl -
sRGB and Rec.709 share the same color primaries.
R x=0.64 y=0.33
G x=0.30 y=0.60
B x=0.15 y=0.06with D65 White Point
x=0.3127 y=0.3290That’s why many people use sRGB and Rec.709 interchangeably.
BUT gamma curve of sRGB and Rec.709 are really different.“Rec.709 gamma curve” is only used in Camera.
“sRGB gamma curve” is for monitor.If your monitor’s gamma curve is sRGB the DPX rendered out from Resolve will be sRGB not Rec.709.
-
Hello,
I’m user of EasyDCP creator+. I output DPX in Rec709 from Davinci and use Rec709-> XZY conversion in EasyDCP creator+, it’s about 100% the same in the theater (I have a Theater for that too, calibrated in DCI – Barco + Dolby server).
Then the workflow is very simple Rec709->XYZ, of course if you grade with a DCI colorpsace on a DCI projector, you have to change the type of conversion in EasyDCP creator+.
Best
Robin
réalisateur, scénariste, monteur, étalonneur
http://www.robinerard.ch -
While I’m unsure about EasyDCP, many programs misinterpret Rec.709 and use the rec.709 camera encoding curve as a decoding curve for the material. This is wrong. Rec.709 material should be decoded using a 2.2-2.4 gamma curve, not the rec.709 encoding curve. I know some DCP authoring software gets this wrong, including DoReMi CineAsset, which we use.
This would explain why things look basically correct if you interpret the input files with an sRGB curve. sRGB is not a pure gamma curve, but its overall gamma is about 2.2, so it would look much closer to correct than the rec.709 encoding curve.
Your “rec.709” monitor which you’re using to color correct has 2.2-2.4 gamma, so that’s why it basically matches.
What you’re seeing is correct, but you should try to use a gamma 2.2-2.4 curve instead of the sRGB one in your DCP authoring software to be 100% correct. sRGB should not be used. Experiment with gamma values between 2.2 and 2.4 to find the one you’re happiest with. I think we’re using 2.3 at the moment for this.
—
Joakim Ziegler – Postproduction Supervisor -
This sounds like a case of working with and rendering files that are not 709 themselves. So even though you are grading to a 709 environment there is nothing about your files that are 709, this using any 709 to XYZ transform is not going to work unless you first bake down your graded/rendered files to 709.
Example: If you were working with full range RGB DPX frames but colored out to YUV 709 (head range) monitor then rendered full range tif a 709 transform really would only apply to what your seeing on your 709 SDI output.
-
We were working on .R3D source media, and I’m sorry, it was OpenDCP not EasyDCP.
Thanks for all the great feedback.
Reply to this Discussion! Login or Sign Up