OpenEXR is a high-dynamic range (HDR) media format supporting multi-resolutions and an arbitrary number of channels and channel types, such as specular, diffuse, alpha, RGB, normals, in a single file. Lustre supports OpenEXR 16-bit RGB only (files with other channels cannot be read). This format is sometimes referred to as "16-bit half float".
Lustre can obtain 16-bit floating-point media in four distinct ways. First, it can acquire the media by way of a Wiretap server. That is to say, by browsing the clip libraries and associated framestores of Visual Effects and Editing applications. Second, it can acquire the media directly, by importing OpenEXR media using a Wiretap Gateway server. Third, OpenEXR files can be read directly without using the Wiretap Gateway server. Fouth, media shot with a RED camera in HDRx mode and where "High Dynamic Range" is selected in the HDRx settings section of the Transcode tab also enter Lustre as 16-bit floating-point. In all cases, you need to tell Lustre how to work with the half float media.
As indicated, OpenEXR represents pixels as floating-point numbers. Internally, Lustre uses a 16-bit integer buffer format (although processing is done in floating-point). Thus, before you can grade OpenEXR media, it must be converted from floating-point to Lustre's special extended-range integer format. This is done by setting a floating-point conversion LUT for the project. This setting is located in the calibration tab of the project settings. Several floating point conversion LUTs are provided as examples.
| LUT | Description |
|---|---|
| log_default_v2 | Converts scene-linear floating-point media into an extended range logarithmic space suitable for grading using Lustre's logarithmic grading controls. Unlike log_default, log_default_v2 avoids problems with lifted blacks when using ACES colour transforms in Setup > Calibrate. This is the default float conversion LUT, and is the recommended way of grading HDR media. |
| log_default | Converts scene-linear floating-point media into an extended range logarithmic space suitable for grading using Lustre's logarithmic grading controls. |
| log_black_limit1 | Similar to log_default, but accomodates media from some digital cameras which may have spurious negative or very low values. |
| lin_default_v2 | Converts the scene-linear floating-point media into an extended range video space suitable for grading using Lustre's linear grading controls. Unlike lin_default, lin_default_v2 avoids problems with lifted blacks when using ACES colour transforms in Setup > Calibrate. |
| lin_default | Converts the scene-linear floating-point media into an extended range video space suitable for grading using Lustre's linear grading controls. |
| lin_black_limit1 | Similar to lin_default, but accomodates media from some digital cameras which may have spurious negative or very low values. |
| ACESproxy_CDL_workflow | Similar to log_default, but the range of scene exposures presented to the ASC CDL match the ACESproxy range. Values outside this range are not clamped. |
| ACESproxy_CDL_blacklim | Same as ACESProxy_CDL_workflow over the official ACESProxy domain, but has more robust handling of very dark values below the limit of ACESProxy. |
| float_identity | Useful for projects with EXR files that have gamma or log encoding baked in. For a single clip instead of a whole project, you can instead use float-to-int-identify.ctf in the Transcode settings. |
Whenever loading half float media, Post Input LUT should be selected in the Image/CDL menu.
Similar to logarithmic media, HDR media requires a viewing LUT in order to look correct on a display or to render to a video color space. This is sometimes referred to as a tone-mapping LUT. This LUT should be selected in the Output tab of the Render settings. See Rendering Shots.
You can use a tone-mapping Colour Transform such as the ones in the RRT+ODT directory. Alternatively, you can use one of the legacy 1d-LUT tone maps:
| LUT | Description |
|---|---|
| lin_default_tonemap | An example tone-map output LUT to be used with the lin_default floating-point conversion LUT. |
| log_default_tonemap | An example tone-map output LUT to be used with the log_default floating-point conversion LUT. |
| identityExtended | An output LUT which avoids clipping of HDR information when rendering back to floating-point. |
When your goal is to output a floating-point image, either directly to an OpenEXR file, or via rendering to Wiretap as part of an interoperability workflow with an Autodesk Visual Effects and Finishing application, use identityExtended as an output LUT to preserve the extended-range floating-point information. For example, if you are pre-grading OpenEXR media preparing for VFX work, use the identityExtended output LUT during the render.