Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums Apple Final Cut Pro Legacy FinalCut Pro RGB 4:4:4 not matching 4:2:2

  • FinalCut Pro RGB 4:4:4 not matching 4:2:2

    Posted by Carl Skaff on October 28, 2008 at 9:37 pm

    Hi All

    I’m a Colorist working on daVinci 2K+ & Resolve.
    We’ve been getting a lot of requests to deliver graded film as Uncopressed QuickTimes in 4:2:2 and latly also RGB 4:4:4.

    So I’ve done this; Got a MacPro, with a Kona3, running the latest FinalCut Pro. Connected to our Bright SAN.

    When I do 4:2:2 QuickTimes everything works fine. But not when I do 4:4:4…

    Since there isn’t a “1920×1080 25i RGB 444” setting I took the “1080i25 422”-one and change the codec to “AJA RGB 444”.

    When I compare the two they don’t look the same at all. The 444 is brighter, and if I check on a external scope it looks like the 444one is lifted “linjerly” (if thats a word) about 7%ish. It doesn’t look like a gamma or CGR/SMPTE issue. The whole signal goes straight up showing to much in the blacks and is clipping in the whites.

    I tryed changing the RGB<>YUV setting between “Full (CGR) 0-1023” and “SMPTE 64-940” but that didn’t make any difference at all.

    So if anyone has a tip or solution, let me know.

    /carl

    Gary Adcock replied 17 years, 6 months ago 6 Members · 21 Replies
  • 21 Replies
  • Sean Oneil

    October 28, 2008 at 10:09 pm

    Go into sequence settings, video processing. Change it to “Always render in RGB.” That may be the problem.

    Sean

  • Carl Skaff

    October 28, 2008 at 11:04 pm

    Does that mean that the file is actualy OK, I’m just viewing it wrong?

    Although, when I checked the two file to eachother I did that in both QuickTime Player, by double clicking on the clip in the bin to see it in the left player-window & dropping it downon the timeline and making a split.

    I’ll try to change that tomorrow.
    Let you know if its that.

    /carl

  • Rafael Amador

    October 29, 2008 at 1:18 am

    Hi Carl,
    If you are going from YUV to RGB or from RGB to YUV you will always see differences in the picture.
    By default FC will render in the same color space of the sequence.
    RGB is always 444. But don’t set the AJA RGB 10b. FC can render RGB only in 8b.
    Rafael

    http://www.nagavideo.com

  • Carl Skaff

    October 29, 2008 at 12:23 pm

    Hmm, that didn’t change it.
    It’s still not looking the same as the 422.

    But when you say that I’ll always see a diference between YUV and RGB…

    You can’t mean that in the end a YUV and RGB should look diferent?
    It shouldn’t matter in the end wether i’m working in one or the other.
    Or are do you mean that FinalCut will always Display the two files different but when I lay of to tape it is actualy the same…

    /carl

  • Rafael Amador

    October 29, 2008 at 1:28 pm

    Hi Carl,
    Is not a matter of FC. This is an unavoidable problem when you send YUV stuff to any application that works in RGB (AE, Shake, Colors,..). The two colors spaces they have different properties. There are colors in the RGB space that do not exist in the YUV space, and colors in the YUV space that do not exist in the RGB space. The best policy is to try to avoid going from a color space to the other when possible”.
    if you want to know more I recommend you an article of Charles Poynton (he knows everything) “Merging computing with studio video: Converting between RGB and 4:2:2” (www.pynton.com).
    cheers,
    rafael

    http://www.nagavideo.com

  • Chris Borjis

    October 29, 2008 at 4:15 pm

    [Rafael Amador] “FC can render RGB only in 8b.”

    when I read about that this last weekend I was appalled.

    Apple needs to step up and make that 10-bit (least) or 16-bit or better.

  • Gary Adcock

    October 29, 2008 at 5:19 pm

    [Rafael Amador] “RGB is always 444.”

    NO IT IS NOT.

    RGB has nothing to do with 4:4:4 -YUV recording at 4:4:4 is more common than with RGB data

    RGB is a color space and 4:4:4 is a method of color sampling where every luma sample has an equal number of color samples. 4:2:2 is 2x as many luma samples as chroma.

    gary adcock
    Studio37
    HD & Film Consultation
    Post and Production Workflows

    Inside look at the IoHD

  • Gary Adcock

    October 29, 2008 at 5:25 pm

    [Rafael Amador] “Is not a matter of FC. This is an unavoidable problem when you send YUV stuff to any application that works in RGB (AE, Shake, Colors,..).”

    It is not FC it is a quicktime issue, it is not an unavoidable since it is not an issue on platforms like Avids, autodesk, quantel, baselight or in most of the wintel edit platforms – nor is it an issue in the CS4 suite on mac.

    [Rafael Amador] “f you want to know more I recommend you an article of Charles Poynton (he knows everything) “Merging computing with studio video: Converting between RGB and 4:2:2″ (www.pynton.com).”

    I can assure you that he does not know everything, non of us do.

    gary adcock
    Studio37
    HD & Film Consultation
    Post and Production Workflows

    Inside look at the IoHD

  • Carl Skaff

    October 29, 2008 at 7:11 pm

    Gary: When Rafaela says:
    [Rafael Amador] “RGB is always 444.”

    I gues he means that RGB is always 444, and can’t be 422.
    Not that 444 is always RGB.
    From what I know RGB has to be sampled equal for R,G & B otherwise the image will look weird.

    So, no other ideas how or why my 444-QT looks so diferent then my 422-QT?
    I aware that RGB & YUV is diferent colorspaces. But the diference between those is not what I’m seeing. My 444-QT is a lot brighter, more then 5%.

    /carl

  • Chris Borjis

    October 29, 2008 at 10:29 pm

    carl, in your quicktime preferences do you have:
    “maintain color compatibility with final cut studio”
    or something to that effect?

    I noticed if that is not enabled you get random
    washed out video with any ProRes codec .mov files. (played in quicktime player)

    it may or may not be related.

    I wish apple would damn-well fix the obvious flaws in quicktime.

Page 1 of 3

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy