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 :)

DCP- from Scratch ( workflow ) ?

K

kito

Member
Joined
Dec 11, 2008
Messages
16
Reaction score
0
Points
0
Location
maroc-casablanca
Website
www.youtube.com
hi RED forum,

just finished the movie here , and now i have to make the output for DCP. The guy in cinema just required me as JPEG2000 files.
what i wanna ask here :

1. what is the good method for DCP from scratch ? - guess DCP base 24fps , so i must make the dpx output as 24 fps ?
2. or i send to FCP as DPX files based on 25fps then later on converting inside FCP into QT-jpeg2000?
3. should i apply LUT ?
4. and what the rest i must concern about ?
hope someone with Master Guru level here can help me .

I tried to look on the site about this case , without any convinced method.
so i hope anyone here could make a liitle light for me, please.

best regards

kito
 
Last edited:
easyDCP?

easyDCP?

When you go from 23.976 or 25 to 24 you don't frame rate convert, you adjust the length of the audio WAV files. Frames don't have a frame rate.

Some DCP seem to be at gamma 2.6 (does anyone know why?), so you need to take that into account in making your DPX frames, DPX can be graded 100% or LOG-C 95-685 and other ways, so you need to know how you are going to get to the final DCP projection frames and what LUT are in use.

Have you used easyDCP player before (to check the DCP)?

http://www.iis.fraunhofer.de/en/bf/bv/dc/pp/
 
hi Dan, thank you for your quick respons, i appreciated much

1. When you go from 23.976 or 25 to 24 you don't frame rate convert, you adjust the length of the audio WAV files. Frames don't have a frame rate.


- so my project based on 25fps, so i dont have to convert it into 24 fps (is that your point right,Dan ?)

2.Some DCP seem to be at gamma 2.6 (does anyone know why?), so you need to take that into account in making your DPX frames, DPX can be graded 100% or LOG-C 95-685 and other ways, so you need to know how you are going to get to the final DCP projection frames and what LUT are in use.

- so inside Scratch, i wanna set all my DPX files as Gamma 2.6, so how about the LUT ? should i apply it or no ? and if it should, Whut most LUT commonly used for DCP transfer? based on my DPX files.


3.Have you used easyDCP player before (to check the DCP)?

- this is a great information Dan. no i never use this.


ok Dan, i just need a little more steps about the LUT, that's all.

and Thank you for your effort.

cheers
kito
 
What I ment about the fps issue, is that if you output frames shot at 23.976 or 25 to be used in a DCP as 24.000fps, you can just output the frames as whole frames, if you try to do interpolation to convert the total frames count to be as if it was shot at 24.000fps then you get interpolation artifacts. There is the issue of the motion and sound being a bit off if you go from 25 down to 24 or 23.976 up to, but that needs to be compared to frame interpolation artifacts going the other way. There is also some pitch shift with the sound, if you use non-pitch shift then you can get some artifacts from the sound length adjustment. When you use a frame set shot at one fps and played at another fps, the SMPTE time code from the camera will display faster or slower on the playback, so you cannot get the total run time for the edit from the camera fps time code (unless you figure by the ratio of the camera to projector fps).

About the gamma used in the projection frames, you will need to look at the easyDCP Creator documentation to see what options it has for its conversion (ingest) of the DPX into the needed result. I was just giving a heads up about the projection gamma since if you show gamma 2.2 images on a gamma 2.6 projector the shadow areas will be too dark, unless the projector has a LUT set in it to compensate.

Post back again and tell what you learn when you get things working.
 
Some DCP seem to be at gamma 2.6 (does anyone know why
Click to expand...

The DCinema spec defines 2.6 as the display gamma. In October, the ITU agreed upon 2.4 for the display gamma for 601 and 709. It's been undefined until now. They stated that a display gamma was not needed because everyone was using CRTs, which were inherently 2.4. Now that everything is transitioning to digital displays, the gamma needed to be defined, because they are all over the map. Anywhere from 1.9 to 2.8.

Right now the ITU is not finished. While they defined the display gamma, there are still several items undefined. Hopefully that will be resolved before the end of the year.
 
Stacey Spears said:
In October, the ITU agreed upon 2.4 for the display gamma for 601 and 709. It's been undefined until now.
Click to expand...

Thanks for the info. I wonder if that puts off the 709 results from REDCINE-X? RED said they use about 46% for 18% gray before, at gamma 2.4 it would need to be 0.489 or 49%?

What is the default gamma for SD and HD video and monitors (I guess 2.4 now)?

Do you have any links about this new gamma 2.4 "standard" and what it means for making "monitor gamma" RGB files?

I found this on the ARRI site,

http://www.acvl.org/digital_intermediates/dicompanion/ch05.html

Quote: [ The Digital Cinema Initiative (DCI) recommends to run the projector with a gamma of 2.6. The characteristic curve of a DLP tuned to this recommendation is similar to the curve of a CRT monitor.]

But they say the 18% factor for ITU is 0.45 not 0.49, so is that just for encoding not the monitor its viewed on?

Quote: [The recommendation ITU-R BT.709 specifies the encoding of luminance values with a power function where the gamma value is 0.45,]

1/0.45 = 2.22, or gamma 2.22 correction

Is the 709 encoding value still 1/2.22 and the display 2.4? Or will the encoding value also be changed to 1/2.4?
 
I will follow-up after NAB. it's not a lot of fun to type on an iPad.

sRGB defined both source and display gamma. 709 only defined the source gamma. 709 LUTs are based on that source value.

Condumer LCDs are all over the map when it comes to gamma. Computer LCDs are another matter.
 
Ok, then I read that as the encoding gamma is still 1/2.22 or 0.45 for 709?

And for DCP its encoded 1/2.6 or 0.3846?

that makes the question, why do they want the monitor at 2.4 and the encoding at 1/2.22 (I guess because they never matched in the CRT days as well, so keep the mismatch into the future), and if they wanted the DCI projector to match a CRT monitor then why is it 2.6 and not 2.4.
 
Hi Dan and Stacey,

thank you very much for your appreciated here. by the way, coz it is sunday, the cinema is closed, then i called the people in there, so we re gonna arrange the appointment and they gimme times to discuss about this case. i will bring my small files , the files that you required here then bring them to the DCP. and I would tell you and report a.s.a.p.

again, this is my first experience which the materials will deliver to DCP, normally i just do print out to film. fuuihhh...whut a hard day., isn't it. take it or leave it, we must accept the progression of technology now. cause in this movie, i do graded 1617 shots (not so special grading), do compositing about 192 vfx shots-- and me also as a Lead compositor just got help from 4 peoples, then make graded back -all the final Vfx comp. in scratch. and in the end, i just dont want the color have a big different look between mine and DCP. if just a bit shift, i cna take it, but hopefully not so a big gap.

anyway thanks again guys-- have a nice day
and i am so sad couldn't go to NAB this year, if any of you have some news about DCP latest news please let me know guys.

cheers

kito
/Vfx artist/Vfx Supervisor/Senior Compositor/DI artist
 
DCP being for gamma 2.6 projectors

DCP being for gamma 2.6 projectors

kito said:
- so inside Scratch, i wanna set all my DPX files as Gamma 2.6, so how about the LUT ? should i apply it or no ? and if it should, Whut most LUT commonly used for DCP transfer? based on my DPX files.
Click to expand...

I noticed this in the side part from the ARRI link above:

Quote: [A digital projector has a higher gamma than a monitor because projected images in a dark surround need another characteristic than images viewed in a dim surround. The system transfer function resulting from using a DLP projector instead of a monitor is shown in the blue curve.]

It seems that they are saying that no compensation is needed, but when I see DLP projection here, the image do look at bit dark and lacking in shadow detail, maybe because they have so many lights on in the theatre to avoid trip and fall issues.

I would shoot a 18% gray card under the lighting you are using, and check that it does not read too low in the graded frames, if you convert the graded frames to BMP you should read 18% gray as about 0.458 to 0.462 or RGB= 116 to 118 or so on the 0 to 255 scale. The shareware program IRFANVIEW has a readout if you put the cursor on a spot on the image and click the mouse button it then reads out the R= G= B= value for that pixel,

http://www.irfanview.com/

If you are near midtone then it should not matter so much what gamma your viewing monitor is at, the pixel value is a better way to tell because the monitor brightness and contrast change the way the image looks anyway, you would hope the DCP projector is setup right in the theatre...
 
Hi Dan,

what a flashest respons from you.
one day, if i got in CA, i will call you and hook up in the bar, guess i owed you a lot of beers, buddy.

thanks you very much Dan, and have a good day - take care wherever you are
about this case, i 'll hook 'em up at the cinema, in the morning, maybe on tuesday.
I'll let you know the result asap. and some pics preview.

God Bless You, Dan

kito
Vfx Supervisor/artist | Senior Compositor | DI artist
 
Dan Hudgins said:
that makes the question, why do they want the monitor at 2.4 and the encoding at 1/2.22
Click to expand...

The reasoning behind a system gamma (input transform and output transform concatenated) that is not 1 is well explained by Pointon in the PDF Stacey linked to (thanks for that Stacey. I had not read that one.)
 
Just digged into my archives:

XYZ 2.6 gamma

The Digital Cinema specifications are often referring to a 2.6 gamma. When applied to XYZ
coordinates, this gamma has nothing to do with a “reproduction” gamma or with the 2.5
gamma of the CRT monitor (primed capital letters – X’Y’Z’ – are used to designate gamma
corrected values).
This 2.6 gamma is an encoding scheme which purpose is to embed in a limited digital space of 12 bits the maximum possible levels that match the human visual sensitivity.
The human visual system is much more sensible to small variations in dark areas. At the other end of the scale, large brightness variations are not detected.
This is true for almost every human sensation and the relationship between the physical magnitudes of stimuli and the perceived intensity of the stimuli.
This is sometimes called the Weber Fechner law, which is based on the logarithm function. It has been superseded today by Stephen’s power law.
Extensive tests conducted by the DCI have shown that the visual system sensitivity, in the viewing conditions of a cinema projection, is following more or less a power law with a 2.6 gamma.
The result is that:
Coding luminance values with a 1/2.6 power law allows a good balance in the number of codes representing low and high brightness values. This is the most efficient coding for transmitting the luminance details that satisfy reproduction on a cinema screen.
The 2.6 gamma, applied to the XYZ, found in the Digital Cinema equations is a coding gamma and has nothing to do with the reproduction gamma.

From Doremi "Gamma for Digital Cinema"
 
You must log in or register to reply here.
Back
Top