Click here to go to the first RED TEAM post in this thread.   Thread: REDWideGamutRGB and Log3G10

Reply to Thread
Page 2 of 11 FirstFirst 123456 ... LastLast
Results 11 to 20 of 103
  1. #11  
    Senior Member
    Join Date
    Nov 2009
    Location
    Helsinki
    Posts
    178
    Quote Originally Posted by Thai Christen View Post
    Quick test this morning with the new IPP2 in camera preset (Epic-W). Without pixel peeping my first feeling was that there are subtle improvements in skin tones which is nice.

    But what I don't understand is when I load the files in RedCine the color profile shows me DragonColor2 and Redgamma4.
    So, is the IPP2 in camera for preview only?
    You need the "unlisted" beta version of RCX, the download link is in the first post here

    Or use the lut Anton mentioned.
    Reply With Quote  
     

  2. #12  
    Senior Member Thai Christen's Avatar
    Join Date
    Feb 2013
    Location
    Switzerland
    Posts
    475
    Quote Originally Posted by jussi rovanpera View Post
    You need the "unlisted" beta version of RCX, the download link is in the first post here

    Or use the lut Anton mentioned.
    I totally missed that. All set. Thanks!
    Reply With Quote  
     

  3.   Click here to go to the next RED TEAM post in this thread.
  #13  
    Quote Originally Posted by Anton Esteban View Post
    Hey Graeme I have a question about LOG3G10. If you used it as your recording space could you theoretically get more dynamic range out of a camera because of the number of stops LOG3G10 has above mid gray? Right now when you click on tools and see the RAW sensor info I'm assuming that's a traditional LOG image. Or is its primary purpose to be a large holding color space that is used to easily convert an expanded raw color space into different working color spaces?

    I've been experimenting with RWG and LOG3G10 a great deal since it was released in Da Vinci Resolve. I can clearly see the benefit of RWG especially in conjunction with a saturation boost from a LAB color based node. I've been getting much better color contrast and speration without colors ever becoming too garish and over saturated. I still have not found, however, the real benefits of using log3g20 vs RED Log Film. To my eye, once you normalize the footage, the dynamic range between Red log film and LOG3G10 looks the same. One just uses more video gain then the other to get there.
    Log encodings are just containers. The key feature of such a container is that it's big enough to hold what you want to put in there! REDLogFilm adheres to the Cineon spec, which means it can encode a linear light value up to about 13.5, which gives you a bit over 6 stops above mid grey before it runs out ( log2(13.5/0.18)=> 6.2288 ) So if your float image always has values lower than 13.5, REDLogFilm works fine. Similarly, if you're grading a REDLogFilm image and you don't need to produce or grade float values (as equivalent log code values) greater than 13.5, you're not constrained by the log encoding.

    On the camera side, ISO calibration is done by constructing a linear gain such that scene mid grey gets mapped to 0.18 in float. Any element in the scene that is brighter than mid grey will get mapped to higher and higher float values, so as you increase the ISO from 800 to 1600 to 3200 to 6400 you're adding a stop of gain each time and those bright areas of the image are getting brighter and brighter. Depending on camera, sensor and OLPF the max gain you can apply before REDLogFilm runs out as a container space is going to be different, but it's inevitable you'll reach it at some point.

    On the grading side of things HDR is increasingly important. SMPTE2048 spec encodes a signal up to 10,000 nits. Say we reasonably put scene mid grey at 18 nits, that means the 10,000 bit peak brightness of the spec is just over 9 stops above mid grey ( log2(10000/18)=> 9.1178 ) and thus if you're trying to grade a log image that represents linear light data, you're going to want it to be able to contain more than 9 stops above mid grey.

    Given the above necessary container size, Log3G10 encodes 10 stops above mid grey and thus means increasing ISO won't constrain the dynamic range of the image, and in grading you can push highlights beyond SMPTE2084 spec. But it's no bigger than it needs to be either, and that's good for precision.

    So you may not have come up agains the limits of Cineon spec yet, but sooner of later you will, and hence Log3G10 makes sense as a larger container format.

    Graeme
    www.red.com - 8k Digital Cinema Camera
    Science enables stories. Stories drive science
    IPP2, Image Processing, Colour Science and Demosaic Algorithms
    Reply With Quote  
     

  4.   Click here to go to the next RED TEAM post in this thread.
  #14  
    Quote Originally Posted by Thai Christen View Post
    Quick test this morning with the new IPP2 in camera preset (Epic-W). Without pixel peeping my first feeling was that there are subtle improvements in skin tones which is nice.

    But what I don't understand is when I load the files in RedCine the color profile shows me DragonColor2 and Redgamma4.
    So, is the IPP2 in camera for preview only?
    IPP2 Preview targets a monitor with REC709 primaries and BT1886 gamma. That is why when you open your R3D in REDCine-X IPP2 Preview, it will set colour space and gamma space to REC709 and BT1886. Similarly The various RedColors and gammas target a monitor with REC709 primaries and BT1886 gamma, and thus when you open in normal REDCine-X, you'll see it coming in as DragonColor2 and Redgamma4. For the IPP2 preview build of REDCine-X (which also expands your set of targets to include other colour space and gamma options) go here: http://www.reduser.net/forum/showthr...-Preview-Build


    Graeme
    www.red.com - 8k Digital Cinema Camera
    Science enables stories. Stories drive science
    IPP2, Image Processing, Colour Science and Demosaic Algorithms
    Reply With Quote  
     

  5. #15  
    Senior Member Thai Christen's Avatar
    Join Date
    Feb 2013
    Location
    Switzerland
    Posts
    475
    Quote Originally Posted by Graeme Nattress View Post
    IPP2 Preview targets a monitor with REC709 primaries and BT1886 gamma. That is why when you open your R3D in REDCine-X IPP2 Preview, it will set colour space and gamma space to REC709 and BT1886. Similarly The various RedColors and gammas target a monitor with REC709 primaries and BT1886 gamma, and thus when you open in normal REDCine-X, you'll see it coming in as DragonColor2 and Redgamma4. For the IPP2 preview build of REDCine-X (which also expands your set of targets to include other colour space and gamma options) go here: http://www.reduser.net/forum/showthr...-Preview-Build


    Graeme
    Gotcha, thanks a lot Graeme!
    Playing with some test files now.
    Reply With Quote  
     

  6. #16  
    Senior Member Christoffer Glans's Avatar
    Join Date
    Jun 2007
    Location
    Stockholm, Sweden
    Posts
    3,979
    Just started up the new RCX, is there any new settings to apply or is it just the basic function of the software?
    "Using any digital cinema camera today is like sending your 35mm rolls to a standard lab. -Using a Red is like owning a dark room."
    Red Weapon 6K #00600
    Reply With Quote  
     

  7.   Click here to go to the next RED TEAM post in this thread.
  #17  
    Quote Originally Posted by Christoffer Glans View Post
    Just started up the new RCX, is there any new settings to apply or is it just the basic function of the software?
    The IPP2 Preview REDCine-X build is just that - it only does IPP2 preview and if you want to go out to REC709 / BT1886 there's not much else you need to do. It only does IPP2 so there's no need to turn on a switch to access it. You can select from the named standard colour spaces and also gamma specs. As we move towards the full IPP2 implementation we'll separate out the technical and creative controls, and also give you more of a colour managed workflow using the full IPP2 system.

    Graeme
    www.red.com - 8k Digital Cinema Camera
    Science enables stories. Stories drive science
    IPP2, Image Processing, Colour Science and Demosaic Algorithms
    Reply With Quote  
     

  8. #18  
    Senior Member Christoffer Glans's Avatar
    Join Date
    Jun 2007
    Location
    Stockholm, Sweden
    Posts
    3,979
    Quote Originally Posted by Graeme Nattress View Post
    The IPP2 Preview REDCine-X build is just that - it only does IPP2 preview and if you want to go out to REC709 / BT1886 there's not much else you need to do. It only does IPP2 so there's no need to turn on a switch to access it. You can select from the named standard colour spaces and also gamma specs. As we move towards the full IPP2 implementation we'll separate out the technical and creative controls, and also give you more of a colour managed workflow using the full IPP2 system.

    Graeme
    Alright, but how does this apply to RWG? Up until now since RWG was announced I've been using RWG/Log3G10 with an output LUT. Should I instead use the output LUT with rec709 color space? Isn't that mapping the colors wrong if the LUT is made from RWG space? How will all this be applied to other software like Resolve? Do you know if they are working to implement all of this? They use to be pretty fast with including new things.
    "Using any digital cinema camera today is like sending your 35mm rolls to a standard lab. -Using a Red is like owning a dark room."
    Red Weapon 6K #00600
    Reply With Quote  
     

  9.   Click here to go to the next RED TEAM post in this thread.
  #19  
    Quote Originally Posted by Christoffer Glans View Post
    Alright, but how does this apply to RWG? Up until now since RWG was announced I've been using RWG/Log3G10 with an output LUT. Should I instead use the output LUT with rec709 color space? Isn't that mapping the colors wrong if the LUT is made from RWG space? How will all this be applied to other software like Resolve? Do you know if they are working to implement all of this? They use to be pretty fast with including new things.
    So IPP2 has a number of stages in the pipeline:

    1) Primary: take sensor data to RWG/Log3G10 via ISO / exposure, kelvin and tint white balance.
    2) Grading: Typically CDL or grading in log space in a grading app
    3) Development: Transform RWG into the desired output colour space (say REC709, P3 or REC2020) via a gamut mapping algorithm that stops out-of-gamut colours going funky
    4) Output: Takes the developed image (still in Log3G10) out to the gamma of the display, typically BT1886 for a standard dynamic range (SDR) HD monitor or SMPTE2084 for a HDR monitor.

    The IPP2 preview cube LUT takes all the math from stages 3 and 4 above (Development and output) and boils it down into a form people could use for testing purposes, or for when the workflow involves grading in a 3rd party application like Resolve.

    The IPP2 Preview build of REDCine-X does all the stages using transform math, and thus is flexible and can be configured for different monitor outputs or colour spaces.

    Graeme
    www.red.com - 8k Digital Cinema Camera
    Science enables stories. Stories drive science
    IPP2, Image Processing, Colour Science and Demosaic Algorithms
    Reply With Quote  
     

  10. #20  
    Senior Member Christoffer Glans's Avatar
    Join Date
    Jun 2007
    Location
    Stockholm, Sweden
    Posts
    3,979
    Quote Originally Posted by Graeme Nattress View Post
    So IPP2 has a number of stages in the pipeline:

    1) Primary: take sensor data to RWG/Log3G10 via ISO / exposure, kelvin and tint white balance.
    2) Grading: Typically CDL or grading in log space in a grading app
    3) Development: Transform RWG into the desired output colour space (say REC709, P3 or REC2020) via a gamut mapping algorithm that stops out-of-gamut colours going funky
    4) Output: Takes the developed image (still in Log3G10) out to the gamma of the display, typically BT1886 for a standard dynamic range (SDR) HD monitor or SMPTE2084 for a HDR monitor.

    The IPP2 preview cube LUT takes all the math from stages 3 and 4 above (Development and output) and boils it down into a form people could use for testing purposes, or for when the workflow involves grading in a 3rd party application like Resolve.

    The IPP2 Preview build of REDCine-X does all the stages using transform math, and thus is flexible and can be configured for different monitor outputs or colour spaces.

    Graeme
    Yeah, that process is what I've been using with the output LUT. But I'm a bit confused about the preview builds workings. Is it supposed to calculate for a color space automatically? Because I still seem to need an output LUT in order to see a proper image.
    Also, did I miss some link to a IPP2 Preview cube? Is that the latest in the Cube development we were talking about earlier?
    "Using any digital cinema camera today is like sending your 35mm rolls to a standard lab. -Using a Red is like owning a dark room."
    Red Weapon 6K #00600
    Reply With Quote  
     

Tags for this Thread

View Tag Cloud

Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts