Welcome to our community

Be a part of something great, join today!

Register 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 :)

Keeping it 12 bit

J

John Sellars

Member
Joined
Oct 13, 2012
Messages
24
Reaction score
0
Points
1
Location
Los Angeles
I have a project shot on Epic that I would like to keep 12 bit all the way to DCP. It has to go back to FCP7 for titling and additional edit tweaks, so I will render in ProRes 4444 from R3D's (at 2048x858). Questions I have are:


  • Does Resolve render 444 in YUV or RGB, and is it really 12 bit without alpha?
  • In FCP, does "Render in RGB" pass through the 12 bit for clips that are untouched?
  • Can I make a 2.40 1080p from this project for SR output and BluRay, or return to Resolve for the downscaling ?
  • Does it really matter that I keep it 12 bit?


Maybe these are questions for an FCP forum, but most editors seem to deal with 10 bit at most…
 
John, how long do you have for delivery? Your job seems like exactly the type of thing that Davinci v10 would be great at tackling, saving that one PITA trip back to FCP right at the end. I certainly can't say how long you'd have to wait, but if you look at my thread right below this it seems like we might be getting pretty close to getting our hands on v10 (a bit of optimistic hope there...).
Best,

-Harry
 
Unless you plan on grading the footage later, you're not going to see a difference between 12 and 10-bit. I'd wager you'd have a problem telling the difference (at least for video) between 8 and 10-bit.
 
This is an easy one. Do your offline in FCP at whatever offline Quicktime ProRes or whatever codec you transcode to. When your cut is locked, export an EDL or XML, send to Resolve. Import only the RAW media not the transcodes into Resolve Media Pool. Import the XML/EDL and conform your timeline in Resolve based on that EDL/XML and compare against a rendered ProRes file of your timeline so you are sure that your conform is done properly. Now you're working with your online media ie RAW 12bit media. Color to taste, and render to 16bit Tiff files at 2048x858. Now, depending upon which DCP software you decide to use you may or may not apply the Rec709-DCI XYZ Lut in Resolve upon output to transform your RGB to XYZ. Easy DCP does a great job in the 709-XYZ conversion in software. Open DCP I don't trust so much so If thats the way you are going to go I'd recommend applying the LUT on output of your Tiff render and skipping OpenDCP's color conversion.
Create the DCP, put it on a Flash Drive or Ext2 or Ext3 hard drive (preferrably with its own power supply, not bus powered) and send to your local cinema / whoever is screening your film. Basically, don't do it the way you are planning to do it and you can finish in 12bit.
 
I always understood (possibly incorrectly) that 12 bit is usually only necessary when doing heavy compositing and color correction.

I know we shoot a ton of green screen, and rendering to QT4444 made keying a breeze, then when we started keying the R3d's in AE it got even easier!
 
Mike - In your workflow, when are the opening credits composited with the textless, and when do I add the end credit crawl? (unless I take Matthew's suggestion). Also, I am working in reels, so they have to be assembled for the 5.1 mix (sound house wants a flat file of the whole film). I know I could assemble in Resolve, but I really want the responsibility for the proper sound mix/sync to fall with the editor and sound house. And, lots of speed effects, which need to go back to FCP for proper rendering. I will be using CuteDCP, which I have used before successfully (it uses AE in 16bit). Rendering in 16bit tiff and finishing in AE might be an option, but I don't think the 5.1 will play out in RT with the picture for client approval…
 
If you export from Resolve as un-compressed .mov, you are sure it will 10 bit, the same as one of the DPX options. But if you do it from FCP, is 8 bit.

Now, if you have Glue Tools, you can even export your DPX sequence as 16 bit from Resolve, and the conformance and final export will be kept at 16 bit, which is the preferable bit depth for DCP authoring.
 
What about titling/Credit crawl in AE - Presumably there'd be workflow there that'd allow you to stay 16bit RGB and then do an XYZ conversion after?
 
I couldn't get OpenDCP or jp2k to give me more than an 8bit-j2c sequence. I *think* you can manually get jp2k to render out the sequence (from Adobe Media Encoder) at whatever bit depth you want (up to 16), but then you're no longer using their "digital cinema" pre-set, so I shied away from it since compatibility was my main concern.

Also, OpenDCP has sRGB and Rec709 to XYZ conversion and both did a proper job for me. I rendered out the same project in all three modes -- No conversion, sRGB-to-XYZ (since that was what we monitored), and rec709-to-XYZ -- just to see what would happen. The sRGB was closest to the original (indistinguishable from our reference monitors), the rec709 was only only slightly more contrasty, and the non-conversion looked straight balls -- super magenta, since there was no compensating for the DCP/Projectors output (which was to be expected.)
 
Last edited:
Mike P. said:
I couldn't get OpenDCP or jp2k to give me more than an 8bit-j2c sequence. I *think* you can manually get jp2k to render out the sequence (from Adobe Media Encoder) at whatever bit depth you want (up to 16), but then you're no longer using their "digital cinema" pre-set, so I shied away from it since compatibility was my main concern.
Click to expand...

OpenDCP creates 12-bit JPEG2000 images per the DCI specification. The images are not 8-bit. Also, you should not create 16-bit JPEG2000 images for use in Digital Cinema.
 
Oh. I opened up the jp2000 files in Photoshop and used inspector/info (whatever it's called) for file details, which is where it said 8bit… that's the reason I was under the impression that they weren't 12-bit. Conversely, no banding or anything like that was visible in the resultant DCP when projected.
 
Just wondering if the reason the jp2000 files open as 8 bit in Photoshop is a restriction of Photoshop, I've only ever seen 8, 16 or 32 bits as bit depth options in Photoshop?
 
Photoshop, Apple Preview, and many other applications will show the files as 8-bit because they only report 8, 16, 32-bit.
 
Just saw the last two responses today. Excellent tidbit of info (that some programs don't report anything other than 8 or 16bit). As I said, the DCP worked flawlessly and looked exactly as intended, so I was content just knowing I could (quickly/easily/inexpensively) make my own DCPs.
 
You must log in or register to reply here.
Back
Top