Choosing the ACES Output Transform for DLP-Based VFX Approval

Discussions - Using ACES VFX
1

Hello,

I work at a VFX company in South Korea and have recently taken on the additional role of operating our in-house screening room, where we review VFX shots. I would like to mention in advance that I am not a DI colorist, nor have I received formal training in color grading. My knowledge is quite limited compared to the experts here. but somehow i have to do this role(just setting the plate and play it on the screen, but i want make it clear of my job duty’s info)

Our VFX supervisor wants our deliverables to closely match the final DI environment, so our internal screening room has been configured to simulate theatrical projection as closely as possible. However, our VFX artists still work on Rec.709 monitors.

Since this project will be released in theaters, we try to visually review the VFX shots in an environment that closely resembles a theatrical projection. Based on that review, the artists then make their revisions.
That’s why we believe it is important for our in-house screening room to match the conditions of both the theater and the external DI facility — even though the creative color decisions ultimately belong to the colorist in the DI suite.

External DI Environment (Baselight)

File format: 16bit half, PIZ-compressed, ACES 2065-1 EXR (exported from Baselight)

Mastering color space: DCI, Gamma 2.6 / P3 D60

RRT: ACES RRT 1.1+

Projection system: Barco DLP (i don’t know Detailed Model Name, but I was told it’s set to a D60 white point their own self made PCF)

Internal VFX Screening Room Setup

System: Mac Studio + DaVinci Resolve Studio 20

Settings:

Color Science: ACEScct (ACES 1.3)

Input Transform: None

Output Transform: P3-DCI (D60 simulation)

Output: 4:4:4 SDI (only works via single-link; dual-link doesn’t function for unknown reasons)

Projector: Barco SP4K-12C

Target White Point: 2D-DCI

PCF Calibration Values:

White: x=0.3195, y=0.3559

Red: x=0.7205, y=0.2795

Green: x=0.1015, y=0.8727

Blue: x=0.1411, y=0.0332

here is My Questions

  1. Based on the above information, would you say that our current configuration closely matches the external DI environment?

  2. In DaVinci Resolve’s ACES Output Transform options, I chose P3-DCI (D60 simulation) instead of P3 D60, because P3 D60 produced a pretty strong green on the screen. Was this the correct choice?

  3. In Barco Web Communicator, the only available options for Target White Point are Rec.709/Rec.2020 and 2D-DCI. Since D60 is not directly selectable, I assumed that it is not supported. Is this a valid assumption?
    and what is acually means 2D-DCI’s white point? while rec.709/rec.2020’s white point is D65.

  4. When creating a PCF in Barco Web Communicator, we input white, red, green, and blue xy values, and also choose a separate Target White Point. Should I understand this Target White Point as a white-point correction or offset relative to the calibrated PCF values?

-When the DLP’s Target White Point is set to 2D-DCI, switching between D65 Simulation and D60 Simulation in Resolve’s Output Transform results in a clearly visible color difference.

However, when DLP’s the white point Target is set to Rec.709/Rec.2020, there is almost no visible difference between P3 D65 and P3 D60 Output Transforms.

5.Additionally, I noticed a subtle but visible difference between the following two combinations:

Target white point: 2D-DCI + ACES Output Transform: p3-DCI(D65 simulation)

Target white point: Rec.709/Rec.2020 + ACES Output Transformt: p3 D65

Is this difference to be expected, or is it an unexpected behavior?

Thanks to ACES, our actual image data remains consistent from input to delivery. However, what the human eye sees during the process is still heavily influenced by display and system settings.
Is there any further approach or best practice within ACES to minimize these viewing inconsistencies during production?

Thank you very much for taking the time to read this long post and my bad english.
Any insights or suggestions would be greatly appreciated.

ACES 2.0 Resolve OFX DRT - P3 D60
2

I would say so, yes.

Yes

Yes. Unless you have the option to manually enter the chromaticity coordinates for D60 white.

That I will have to leave for somebody more familiar with the projector specifics to answer.

Again, somebody more familiar with the projector can answer in detail. But a colour space is defined by a white point chromaticity as well as the chromaticities of its primaries.

That is expected the P3 D65 and P3 D60 Output Transforms are in tended for projectors calibrated to each of those white points, but you are sending it to the projector without changing the white point. The actual code values are only very subtly different, as the visual difference comes mainly from the difference in projector white point. When simulating D60 and D65 on a DCI white projector, the difference is baked into the signal.

It is expected. Additional scaling is needed to ‘fit’ D65 white when simulating it within a colour space with DCI white. This is not needed for a projector natively calibrated to D65.

1 Like
3

thank you so much Nick Shaw.

I have a small but great confidence in handling my work from your help.

For answers to the rest of the questions you have left, I’ll be waiting for answers from other experts who are familiar with the dlp projector.

Thank you again. Have a nice day!

4

Hi Keunyoung,

  1. I disagree with Nick here. DI uses a native D60 white point while you are using a DCI white point with D60 simulation. You will get some limits that DI will not get, notably getting a hue shift when whites are clipping. Ideally, you should do just like DI and calibrate the projector to D60 and not use the simulation in ACES.
  2. P3D60 is meant to be used with true D60 calibration in the projector. D60 simulation is meant to be used with a DCI white point calibration on the projector. The mismatch is giving you the green tint which is not correct in that case.
  3. It’s been a couple of years I touched a Barco but I believe you can generate one in Communicator Postproduction. You would have to reach out to your Barco rep. to get a license. As an alternative, you can probably create the white point as a TCGD from the TI DLP Control software. You will need someone like your projector service person to help on this.
  4. You might be correct on that one although I have no memory of that. I always hated PCF files. Sorry!

Next question: Nothing to add to Nick’s answer.

  1. Nothing to add except 2 things
    • You should avoid simulations and use the calibrated white point
    • Save yourself a bag of troubles and do everything in D65: rec709 on the floor and P3D65 in DI and review room. It will also help if you have to master in HDR.
1 Like
5

It’s certainly true that the ideal would be to set the projector to P3-D60 2.6 gamma, and use the appropriate Output Transform. Hence my comment:

Unless you have the option to manually enter the chromaticity coordinates for D60 white.

My response was on the basis that the OP was only able to set the projector to DCI white, in which case “P3-DCI (D60 simulation)” is the next best thing. You are right that the appearance will vary slightly from that of a true D60 projection at the very top end.

2 Likes