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 :)

RGB, YUV, & Film-out

T

Tommy

Well-known member
Joined
May 19, 2008
Messages
64
Reaction score
0
Points
0
I would really appreciate a bit of guidance as I try and iron out the final details in my plan... I plan on transferring the R3D files to Prores for cutting, going Crimson to Redcine, doing a one-light, and exporting DPX files using RedLog gamma & CameraRGB (or would Redspace be better?) color space. From there I will get a LUT from the film-out house and do the color grading in Color. I'll finish with DPX files on a RAID that I can take to the film-out house and have transfered. Now, to create an HDCAM SR master that can be screened at festivals as well as HDCAM, Digibeta, and DVDs can I grade my original DPX files (prior to exporting from Color for film-out) or do I have to start back at Redcine and export using Rec709 gamma and Redspace color space in order to have the master be YUV? Is there any reason I would need both a RGB and YUV HDCAM SR master? What is the best way to create both a version of the film for a film-out and an HDCAM SR master and the various deliverables I'll need? Finally, going back to the source, what color space and gamma settings should I use on the camera while shooting? Thank you so much for your advice. Tommy
 
Now that we have the native workflow in FCS, your concept looks overly complicated. If you want to grade in Color anyway, forget RC and Crimson.

Depending on shooting ratio and amount of FX to be used in FCP, use either wrapped (native) L&T or ProRes for going into FCP. Edit.

If you used ProRes, do a re-conform as described in the whitepaper. Export to Color, get your LUT for it from the lab and grade for film. Let Color render to a DPX sequence.

Now switch Color and your calibrated display to 709, grade for TV and render to ProRes. Export back to FCP and play out to HDCam or have it encoded for BluRay. Get a good card like AJA Kona3 and let it do the downscale for SD.

Done.
 
Still not that simple, color does not follow spanned R3D's there is a detailed thread about that:

http://www.reduser.net/forum/showthread.php?t=23105

and that without speaking of the colormanegmant you have to do to get the right LUT's for film out and rec. 709 indipendently. If you have to deliver DVD's as well you need a good downscale option and a good mpeg 2 encoder.
 
It's still unclear if the spanned files problem is present only in a workflow not recommended by Apple or in wrapped files too. I couldn't reproduce the bug here with wrapped files.

I agree with you that correct color management for film-out with Color is still quite challenging. It's very important to talk to your lab.

For free downscaling and good encoding to DVD, please see my resposes next door:

http://www.reduser.net/forum/showthread.php?t=23350
 
I'm beginning to think that the best plan might be to transcode, edit, and finish using 2K Prores HQ YUV that way I can create an HDCAM SR master and all my other various festival deliverables while staying in the Rec709 color space and when I'm ready to do the film-out (if we get distribution) I'll go to the post house and have the them color correct from DPX or the original R3D files. Does this sound like a more reasonable plan? I've been trying to figure out how to go about this process using DPX files but I get hung up on the RGB vs. YUV issues and the storage & performance issues (money). Thanks, Tommy
 
Stop me if this is crazy, but couldn't the lab filmout your SR master? Maybe not the best way to go, but simple and likely cheaper if you're already making an SR.

At the end of the day, your contact at the lab should be able to walk you through the costs of the different workflows, and you can pick from there.
 
I think your plan makes sense, Tommy. Why put time and money into preparations for a film-out before it's clear you will have one? And it's never a good idea to make an electronic version straight from a version graded for film.
 
nomad said:
IAnd it's never a good idea to make an electronic version straight from a version graded for film.
Click to expand...

It's done exactly that way by virtually every digital intermediate facility in the world. In fact, it's the only real way to get all of the deliverables - film and electronic - to look essentially the same. The base color correction is done to a film target using a print preview lookup table, and the electronic deliverables are created by swapping out to a Rec.709 LUT (using the same film target) and making minor, if any, level adjustments.
 
Very important thread, thanks.

I have one question:

Why is DPX the preferred format? Isn't exporting to 16-bit TIFFs the *ultimate* route in terms of color precision and image fidelity? Are there cons to using TIFFs [besides the bigger size]?

(OK, those were three questions. :biggrin:)
 
I admit that my answer was a bit short. Yes, you can deliver from the film version if you swap the LUT. If not, it's going to be unwatchable…

Regarding DPX: It's just more efficient than TIFF and widely accepted in professional DI film workflow.
 
Thanks everyone for your help! I believe the reason is that DPX files are the more widely accepted of the two in terms of software compatibility-but I'm not 100% sure on that. Hopefully someone can confirm that. To help me put this picture together... I would need to ask the post house what color space and gamma they want for the DPX files that I bring them. Then, what I would get from the post house after color grading is a DPX version (on a HDD) of the movie color corrected for a film-out and my video deliverables (HDCAM SR, HDCAM, DVD) color corrected in the Rec709 color space. Is that correct? Is there a standard color space and gamma used for DPX files?
 
Tommy said:
Thanks everyone for your help! I believe the reason is that DPX files are the more widely accepted of the two in terms of software compatibility-but I'm not 100% sure on that.
Click to expand...

DPX are widely used for a number of reasons, not the least of which is that they are specifically designed to carry metadata in a standardized way. The most common metadata used in DPX headers is SMPTE timecode, but you can also have reel numbers, and keycode numbers for film, among other things. Some systems even include audio in DPX headers. All of this information is usually included in film scans (not the audio, of course), and in files converted by Red's tools, time code and reel number are included. This facilitates conforming, visual effects, and other post production steps.
 
karapetkov said:
Are there 16[48] bit DPXs?
Click to expand...

Yes, but they are rarely used in post production.

As RED's files have 12-bit color, isn't something lost in a 10-bit DPX?
Click to expand...

Not really. Log encoding is usually the answer. In image processing, 10 bit log is generally considered to be the equivalent of about 14 bit linear. Red's Redlog gamma curve is specifically designed to retain all available information from the 12 bit linear original and place it in a 10 bit container.
 
When shooting you can use what you like – it's all metadata. If anyone is looking over your shoulder (or sitting in the video village), Redspace will give them a nice, punchy image. If you want to know what your camera can actually see, go linear.

Regarding RedLog: talk to your film-out service, as has been said a thousand times before. Some might prefer PanaLog, since it may fit their workflow.

BTW, those 11 or 11.5 stops the Red can deliver will fit nicely into 10 bit log.

Regarding cine vs. TV: in an ordinary living room lit normally TV can display about 5 to 6 stops ( so it's quite unrealistic to showcase a Panasonic Kuro in a pitch-black room). Cinema can deliver about 9, when there's still some light to find your popcorn or the emergency exit.
You could achieve even more in a home theatre with a JVC projector, walls painted black, seats in black velvet and yourself wearing black too…

That should make it very clear why you can't use the same images for all displays. TV viewers would just miss to much of the action in dark scenes, while even the cinema version would look milky on the JVC. You can adapt automatically with different LUTs, but some say that specific grading will make it an even better experience.
 
nomad said:
When shooting you can use what you like – it's all metadata. If anyone is looking over your shoulder (or sitting in the video village), Redspace will give them a nice, punchy image. If you want to know what your camera can actually see, go linear.

Regarding RedLog: talk to your film-out service, as has been said a thousand times before. Some might prefer PanaLog, since it may fit their workflow.

BTW, those 11 or 11.5 stops the Red can deliver fit nicely into 10 bit log.
Click to expand...

Depending on positive stock... you either have for Kodak films:

Kodak Vision 2383 maximum density from 3,3D to 3,55D thats 11 to 11,8 stops
Kodak Vision Premiere 2393 maximum density 3,9D to 4,3D 13 to 14,5 stops

The first gives a very good image and its dificult to get a good result, but its achievable with good color management.

The premiere is like a wild horse... if you want to ride it and get astonishing velvet blacks, you really need an exceptionally calibrated display device and lots of experience and very wide know how...

So to put the "11 stops" of RED in the correct RGB density range you really need a color managed pipeline to achieve it.

Karapetkov, always have in mind that LUTs are good only if they are derived directly from measuring your specific display devices... and taking account all the mismatches and hiccups for your monitor.. so no generic LUT's...

All the rest is as they say... start for film out and use LUTs to deliver for PAL and HD - REC709.
 
Regarding the FCS-Color workflow for filmout if you want to grade native R3D, Color makes at the moment a half debayer. So if you want to go on the big screen I think the DPX route with a full debayer would be the best choice. Monkey Extract does a really nice job when you take a EDL of your Prores Offline..... and you can take the dpx files and the edl and grade it on any CC.
 
You must log in or register to reply here.
Back
Top