Andy Minuth

Hi Nikola, you would definitely render everything into an OpenEXR with a Linear colour space. The colour primaries of this space depend on your workflow (709, AP0, AP1, E-Gamut, etc.). During this render In BL you have the option to smoothly bend in out of gamut colours, which might cause problems while comping or grading down the line because of negative code values. Use a slight CompressGamut for that. Then you need BLGs for the full look or at least the correct viewing transform for the content to look right. You can export the viewing transforms as LUTs out of Baselight. Let me know if you need help with that. Cheers, Andy

Hi Mark, You should avoid a mismatch of Viewing Colour Space in BL and the settings of your display. Otherwise your renders will look wrong. Actually there is no need in BL to set your viewing colour space to P3 gamut while grading in your case, but if you want to do it, then please use the 'Dolby: ST 2084 PQ / P3 D65 / 1000 nits' colour space for viewing. Alternatively set your viewing colour space to 'Rec.2020: ST 2084 PQ / Rec.2020 / 1000 nits' and the X300 accordingly. Please verify then in the Colour Space Journey that your Mastering Colour Space is 'Dolby: ST 2084 PQ / P3 D65 / 1000 nits'. It will limit your output to P3 gamut. Just try to see Rec.2020 as a container gamut, compareable to X'Y'Z'. And as a sidenote: Please use the 709 colour matrix in the X300 in both cases. In BL 5.2 there will be an option to change the colour matrix for the RGB -> Y'CbCr conversion, but until 5.1 the 709 matrix is used. Cheers, Andy

@Margus: Are you using ACEScc or ACEScct as a working space? ACEScc might make noise in under-exposed shots a lot more visible, because it increases the contrast in the blacks if you push up a shot heavily. Curves like ACEScct, T-Log, LogC, etc. avoid that by using a toe in the shadows and are better suited as working space. The out of gamut colours that Jussi mention need to be dealt with accordingly. The T-Log / E-Gamut working colour space is optimised for typical out of gamut colours of current cameras. Alternatively working with the native primaries of the main camera sources (LogC / WideGamut, 3G10Log / REDWideGamutRGB, S-Log3 / S-Gamut3.cine, V-Log / V-Gamut, etc.) might avoid them on input side. But when you are running into out of gamut artefacts on the output side (e.g. in the RRT) during grading you need a tool to deal with that. In Baselight we have the CompressGamut tool which deals with out of gamut problems in both cases. It gently moves colours back into a target gamut without affecting the overall look. And it can bring back even negative code values, so it also works nicely in smaller working gamuts like AP1.

Glad to hear that Julien.

Hi Margus, there is a built in Grain tool with parameters to tweak. Cheers

Join us for live presentations and practical discussions on look development, colour space management and new grading tools – followed by a networking ‘Happy Hour’ at the Fimwirtschaft to celebrate the one-year anniversary of our Munich office. https://filmlightcolourtourmunich.eventbrite.com/ Cheers, Andy

I can't comment on that specific model. But 72% NTSC gamut is not a helpful information.

For that budget focus on a decent sRGB/709 display. And forget about NTSC. Larger Gamuts you want to cover are P3 and Adobe RGB, depending on the type of projects.

Seems okay on the first look. 🤙

If you are only delivering 2.4 Gamma / 709, then I would calibrate just this.

In most cases Full to Legal scale will be the right choice. It takes 0 and maps it to 64. In your case 10bit CV -3 will also be mapped to 64, because everything below 0 will be clipped. It maintains the information in the shadows and highlights. The soft clip operation basically takes everything below CV 64 and above 940 and clips it. So you will loose picture details there, but the image will look more contrasty compared to Full to legal scale.

We are organising a day of exciting presentations at IBC this year. The speakers are: Maxine Gervais, Senior Colourist, Technicolor, Hollywood; she will talk about her recent work on Black Panther, The Meg and Alpha. Sylvain Canaux, Colourist, St Louis, Paris; Sylvain will explain his grading apporach on high-end beauty commercials. Mike Nuget, Freelance Colourist/Editor, New York City; Mike will present his recent experiences live using the new Baselight for Avid tools. Daniele Siragusano, Image Engineer, FilmLight, Munich; In this more techical session Daniele will show what it takes to produce natural images on modern wide gamut displays They will present live from a Baselight system and go into the details of the colour space management and finishing toolset that helped them to create beautiful grades, while navigating modern challenges such as HDR and multiple delivery formats. Seats are limited. Please register at: https://goo.gl/agS9Wh

log / film-style controls. Nice comparison 🤙

Visiting NAB this year? I would like to draw your attention to the following event: https://www.eventbrite.com/e/arri-and-filmlight-workshop-at-nab-2018-registration-44545711422 We will show an effortless BLG workflow between the Alexa and the FilmLight tools. And after that I will present new and upcoming features of Baselight. There is a longer description of the event in the link. See you there, Andy

For Kodak I remembered it like this, but no guarantee: The starting number 1 is 35mm nitrate 2 is 35/70mm polyester 3 is 8/16mm polyester 5 is 35/70mm acetate 7 is 8/16mm acetate Don't know about 4? The second number indicates camera film (2) or lab film e.g. print (3) The last two digits seem more arbitrary to me.

You need to work with spatial tools, to rework the edges. In Baselight I like to use Sharpen (negative range) and DFuse a lot for that. In Baselight 5 there are now even more powerful tools for that with Texture Equaliser and Texture Blend. They give You a lot more precise control over Your edges, because they disassemble the image into several different frequencies (sharp edges -> soft edges) and let You control each of these frequency bands individually.

Log encodings were developed to effectively store high dynamic range images with integer codings. Integer formats are limited in dynamic range to 0..1. OpenEXR is a much more advanced idea and was developed for the computer-graphics industry, where linear floating-point coding is a native format. It combines a lot of advantages: For example the effective coding of log, with the almost unlimited dynamic range of floating-point, etc.

One should make sure to save Linear data to OpenEXR files, to maintain the best color fidelity. A log colour space saved to an OpenEXR file might work, but will only use a small fraction of the available bit depth.

Hi Amada, > Does the ARRI Photometric DRT eliminate the need of the regular LogC to Rec709 LUT provided by ARRI? Yes exaclty. The transform from that LUT is inside the DRT, but in a more generic way, so it can be applied to and from different colour spaces. > Does the Truelight Film 1 LUT mimic the film stocks precisely? See dermots answer. It is based on real measurements taken from a print stock and not "graded" to look like film. But it is not intended to be used as an accurate simulation of a film-out, that should be provided by the lab doing the film-out. > You wrote that the DRT output the image a bit flat. Does that mean it will only transform the image 'halfway' to the IRE range known as within the REC709 standard? The 'flat' comment was only regading the 'Truelight Video 1' DRT. There we are going to Rec.1886/709, but we are letting the colorist add the last bit of contrast manually. I hope that helps

I guess we helped You via support.

Better late than never. Here is a version of the used show-lut. The original one goes to P3 D60, but for most of You Rec.1886 (709) is probably more suitable. It expects Arri LogC WideGamut as an Input. HW_ShowLUT_LogC_2_Rec1886_16.cub

Generally speaking: Especially when You are using several different input colour spaces, it makes sense to tag Your footage correct. I recommend You to use the 'Automatic' setting or set it manually per clip for transcoded footage. All the footage will then be matched to one log flavour (your working CS) and grades can be reused much easier. For Your specific case, I ask You to solve it with the BaseLight support baselight-support@filmlight.ltd.uk. Then we can have a detailed look at Your scene and help You solve all the issues.

Hi Soumitra, I just replied in the other thread about this topic. Best, Andy

Hi Soumitra, it makes me happy to hear, that You like BaseGrade. What You are describing is Telecine Style grading. Generally I would recommend You to exercise scene-referred film-style grading. The scene settings for that would look like that: Input CS: Defined by You footage, e.g. ARRI LogC WG Working CS: Your preferred scene-referred space, e.g. Filmlight T-Log / E-Gamut or ARRI LogC WG Grade Result: From Stack Display / Render CS: Your target display, e.g. Rec.1886, DCI XYZ, DCI P3, etc. DRT: Choose one that suits Your artistic and technical needs, e.g. Truelight CAM I know it is hard for a colourist to switch the style especially on an important job. But in the long run, it is worth it. With the arrival of HDR displays and multiple deliveries, the telecine style heads into a dead end kind of. Additionally You could add a reduction of contrast as a first layer (set to 'No Bypass'), to show the client a washed out image in the beginning of the grade. Having said that, BaseGrade can also be used in a Telecine style scene. Best practice for that looks like that: Use the same colour space settings as mentioned above. But then insert a 'Colour Space' layer to the stack, set to 'Identify Colour Space' and the colour space You are manually grading into, e.g. Rec.1886 or DCI 2.6 Gamma / D60. Your image will then look wrong (washed out) as desired. The important thing is then to put the Colour Space layer at the right point in the stack. Usually Your grade will start with adding a lot of contrast to the washed out image. As soon as the image brightness and contrast look okay on the display, the necessary colour space conversion is finished. After that layer You should put the 'Identify colour space operator'. Basically it tells BaseLight that before that operator, the image is in Log colour space, and after it, it is in Display Colour space (e.g. Rec.1886). I attached a picture that shows how the stack might look. BaseGrade will then work 'correct' on the image. Please note that most of the new tools in Baselight 5 like 'Boost Shadows', 'DeNoise', etc. are colour space aware and work best with correctly tagged colour spaces. Happy grading, Andy

In Baselight 5 You can use the matte brush in the paint tool, which creates a matte for a layer with a brush. Brushstrokes can be tracked. Then just add the desired correction (brighter, darker, etc.) to the layer and You will paint it on to the image.