Media for DCP Color space xyz or res709

[singhasubhajit.i]

Hi alI [emoji1]
I had just finished a short film. It is now in color correction facility. They ask me about the delivery format. Is it worth to collect j2c image sequence in DCI xyz colours pace instead of 10bit liner res709 colours pace as we normally do. The final goal of us to make Dcp from the master material. Do anyone have any idea?* *


Sent from my Lenovo A328 using Tapatalk

[Cornucopia]

What can you afford?

Depends on your DCP authoring package, but if it accepts DCI xyz color space (not colors pace), that would seem a more direct, straightforward way to go, since that is where you want to end up.

Do you also intend to have other output destinations? DVD? BD? Internet streaming? Those would likely use 709 (either 8 or 10bit, depending). You may need BOTH (which is why I asked the affordability question).

Going straight from the CC process allows one to retain the extended colorspace & dynamic range that was created there. Going to 709 and then back again to XYZ will never fully recover from the loss.

Scott

[singhasubhajit.i]

What can you afford?

Depends on your DCP authoring package, but if it accepts DCI xyz color space (not colors pace), that would seem a more direct, straightforward way to go, since that is where you want to end up.

Do you also intend to have other output destinations? DVD? BD? Internet streaming? Those would likely use 709 (either 8 or 10bit, depending). You may need BOTH (which is why I asked the affordability question).

Going straight from the CC process allows one to retain the extended colorspace & dynamic range that was created there. Going to 709 and then back again to XYZ will never fully recover from the loss.

Scott

Thanks, and sorry for spelling mistake.
Instead DCP for other distribution like BD DVD and Internet streaming we use HD prores HQ mov as source.
So according to you, for DCP j2c in xyz color space is best option. I am going to use DCP o Matric for making DCP. Is it OK with j2c?

Sent from my Lenovo A328 using Tapatalk

[Cornucopia]

Quite understandable to use HD ProresHQ as a source. The thing is, when the Color Correction is done, quite often it "expands" the range of used colorspace - quite possibly BEYOND what is workable with just standard YUV (Prores' native space). Think: floating point math vs. rounded integer math. If you were to attempt restricting the resulting spectrum to just YUV, you would be irreversibly crippling it's impact. That's one area where cinema distribution still way outshines video distribution.

I'm assuming you meant j2k (as in jpeg2000), not j2c. And DCP-o-Matic. I've not used that software (have previously used easyDCP), but wouldn't expect it to have any difficulty with j2k, as that is the ONLY visual format that is DCP-compliant.

Scott

[singhasubhajit.i]

Quite understandable to use HD ProresHQ as a source. The thing is, when the Color Correction is done, quite often it "expands" the range of used colorspace - quite possibly BEYOND what is workable with just standard YUV (Prores' native space). Think: floating point math vs. rounded integer math. If you were to attempt restricting the resulting spectrum to just YUV, you would be irreversibly crippling it's impact. That's one area where cinema distribution still way outshines video distribution.

I'm assuming you meant j2k (as in jpeg2000), not j2c. And DCP-o-Matic. I've not used that software (have previously used easyDCP), but wouldn't expect it to have any difficulty with j2k, as that is the ONLY visual format that is DCP-compliant.

Scott

Thanks Scott. I tried DCP-o-matic and find the result is impressive. Far more better then the native conversion technique used by DCP-o-Matic. The projection image is similar to our CC image. More importantly making dcp from J2K in xyz is not a time consuming job at all, a 100min film took just 1houre to finish as dcp. Thanks again.

Sent from my Lenovo A328 using Tapatalk