Page 2 of 2

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 6:16 pm
by Paul Provost
Well specs do say this
480p, 576p, 720p 60/50, 1080i 60/50, 1080p 60/50/24, 3840x2160p 60/50/24, 4096x2160p 24 *3

So 1080 is not native even though it uses 1080 chips? Does it downscale 4k input to UHD? Does it scale everything to fake UHD? Well for the money...
I was thinking more of a "Review" device and maybe not as a theatrical grade device. Maybe grade on LCD display then review and trim on projector. I don't know.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 6:22 pm
by Erik Wittbusch
The projector isn't the only thing you'll need for theatrical projection.

You'll need a big room, absolute darkness/dark wall+ceiling, 100% controllable light conditions, a good screen, cooling...

Might be more expensive than the projector in the end.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 6:53 pm
by Andrew Kolakowski
Paul Provost wrote:Well specs do say this
480p, 576p, 720p 60/50, 1080i 60/50, 1080p 60/50/24, 3840x2160p 60/50/24, 4096x2160p 24 *3

So 1080 is not native even though it uses 1080 chips? Does it downscale 4k input to UHD? Does it scale everything to fake UHD? Well for the money...
I was thinking more of a "Review" device and maybe not as a theatrical grade device. Maybe grade on LCD display then review and trim on projector. I don't know.


It crops 4K to UHD.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 6:55 pm
by Nimrod Erez
In test after test in the studio on the Eizo monitor we have, the P3 image looks much much better than the 709 picture, for whatever reason. Not using any LUTs for normalizing, just contrast and pivot on a third node in the line, all other corrections happen on the LOG material on a serial node prior to normalizing node. P3 image much smoother with richer colors. And this is just on an interview shot.

However that doesn't mean I'll end up grading the project in P3. Past conversions from 709 to DCP were super smooth and looked good on the big screen and the film will eventually exist streaming on Netflix = 709.

To me it sounds like only 98% of P3 on my monitor is not much of an issue because the what's important is not color gamut but hue, and Y gamut (As many F stops as the format supports basically.) As long as the hue doesn't shift, losing 2% of color is not such a big loss, especially on a historical documentary.

Will probably stick to 709 just to be on the safe side, for now.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 6:55 pm
by Paul Provost

The projector isn't the only thing you'll need for theatrical projection.

You'll need a big room, absolute darkness/dark wall+ceiling, 100% controllable light conditions, a good screen, cooling...

Might be more expensive than the projector in the end.

well yeah those are a given, but much less onerous with a home theater type setup.
believe me I know - I was halfway through moving forward with a barco, complete room build etc. before I abandoned the idea. just not a good business model in my market.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 7:02 pm
by Paul Provost
Nimrod Erez wrote:In test after test in the studio on the Eizo monitor we have, the P3 image looks much much better than the 709 picture, for whatever reason. Not using any LUTs for normalizing, just contrast and pivot on a third node in the line, all other corrections happen on the LOG material on a serial node prior to normalizing node. P3 image much smoother with richer colors. And this is just on an interview shot.

However that doesn't mean I'll end up grading the project in P3. Past conversions from 709 to DCP were super smooth and looked good on the big screen and the film will eventually exist streaming on Netflix = 709.

To me it sounds like only 98% of P3 on my monitor is not much of an issue because the what's important is not color gamut but hue, and Y gamut (As many F stops as the format supports basically.) As long as the hue doesn't shift, losing 2% of color is not such a big loss, especially on a historical documentary.

Will probably stick to 709 just to be on the safe side, for now.


stick to 709 is a good idea.
i think you might be being fooled by what you are seeing. it's kind of like when I'm doing a grade and client says, "I like more how it looks there!" (the app gui instead of the broadcast monitor)
a dcp made from your computer monitor preset p3 setting is a dangerous route. you really have no idea what you are looking at (gamma? viewing environment? luminance?) without proper profiling and a test of that pipeline.
also
do you even know what the white point and gamut are? have you measured it?

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 7:04 pm
by Andrew Kolakowski
Nimrod Erez wrote:In test after test in the studio on the Eizo monitor we have, the P3 image looks much much better than the 709 picture, for whatever reason. Not using any LUTs for normalizing, just contrast and pivot on a third node in the line, all other corrections happen on the LOG material on a serial node prior to normalizing node. P3 image much smoother with richer colors. And this is just on an interview shot.

However that doesn't mean I'll end up grading the project in P3. Past conversions from 709 to DCP were super smooth and looked good on the big screen and the film will eventually exist streaming on Netflix = 709.

To me it sounds like only 98% of P3 on my monitor is not much of an issue because the what's important is not color gamut but hue, and Y gamut (As many F stops as the format supports basically.) As long as the hue doesn't shift, losing 2% of color is not such a big loss, especially on a historical documentary.

Will probably stick to 709 just to be on the safe side, for now.


I'm also not convinced that this missing 2% is real issue. We are not talking abut multimillion budget film projects.
What side effect can it create at the end?

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 7:15 pm
by Nimrod Erez
Paul Provost wrote:
Nimrod Erez wrote:In test after test in the studio on the Eizo monitor we have, the P3 image looks much much better than the 709 picture, for whatever reason. Not using any LUTs for normalizing, just contrast and pivot on a third node in the line, all other corrections happen on the LOG material on a serial node prior to normalizing node. P3 image much smoother with richer colors. And this is just on an interview shot.

However that doesn't mean I'll end up grading the project in P3. Past conversions from 709 to DCP were super smooth and looked good on the big screen and the film will eventually exist streaming on Netflix = 709.

To me it sounds like only 98% of P3 on my monitor is not much of an issue because the what's important is not color gamut but hue, and Y gamut (As many F stops as the format supports basically.) As long as the hue doesn't shift, losing 2% of color is not such a big loss, especially on a historical documentary.

Will probably stick to 709 just to be on the safe side, for now.


stick to 709 is a good idea.
i think you might be being fooled by what you are seeing. it's kind of like when I'm doing a grade and client says, "I like more how it looks there!" (the app gui instead of the broadcast monitor)
a dcp made from your computer monitor preset p3 setting is a dangerous route. you really have no idea what you are looking at (gamma? viewing environment? luminance?) without proper profiling and a test of that pipeline.
also
do you even know what the white point and gamut are? have you measured it?


White Point on the monitor says 48cd/m, I'm not sure how it translates to D6x values. Gamut is set to DCI-P3.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 7:16 pm
by Paul Provost
that's luminance. it "says" p3 but you need to measure it OR 709 for something as critical as drp to theatrical release. call Dave Abrams

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 8:25 pm
by Nimrod Erez
Paul Provost wrote:that's luminance. it "says" p3 but you need to measure it OR 709 for something as critical as drp to theatrical release. call Dave Abrams


I got Dave's email address and will connect with him soon, thinking about getting you in here to consult for a day.

Re: ProRes HQ 422 to DCI-P3 workflow

PostPosted: Fri Mar 10, 2017 8:27 pm
by Paul Provost
well you know where to find me. I'm not a certified calibrator, but i have the gear and can check it out (i do my own calibration, but used dave in the past - great guy).