For those who don’t know the tool: FilmLight’s Baselight is a professional colour grading and finishing system used across film, episodic and advertising pipelines. It integrates with Avid, Nuke and Flame, and extends to Baselight for macOS for flexible desktop workflows.

Complexity simplified

FilmLight has released Baselight v7, a major update to its colour grading and finishing platform. The release focuses on matte handling, machine-learning-assisted segmentation, and performance for complex, VFX-driven projects.

A redesigned matte and channel architecture underpins the release, with the new matte channel picker offering list or thumbnail previews of internal and external mattes, including cryptomattes and depth maps. Up to 128 inputs are now supported in Matte Merge, which displays thumbnails for every matte input, simplifying the selection and combination of layered mattes.

Smarter mattes, finer edges

Machine-learning tools in Baselight v7 include the Segment Anything Flexi Effect, which isolates objects using simple region selections or control points. Artists can layer multiple selections and manage visibility via the integrated Blackboard panel interface. Edge handling has been improved with an updated Edge Filter and the new Matte Refiner, which recovers fine edge and hair detail for better composite blending.

The Depth Map generator automatically extracts depth from live-action footage. Its output feeds directly into the new Depth Keyer and Depth-based operators such as Haze and Bokeh, enabling depth-informed grading and atmospheric or optical effects. All generated channels can be passed downstream or exported in multi-channel EXRs.

Streamlined finishing tools

Baselight v7 adds the Transition operator, offering multiple transition types including Flow Blend, Dissolve, Wipe, Transform and Dip to Colour. Flow Blend can repair missing or damaged frames and smooth jump cuts, effectively replicating Avid’s Fluid Morph behaviour. Additional tools include Spill Suppression for removing colour spill on chroma-keyed material and a new Chromatic Aberration operator for optical correction or simulation.

Transform Matching automates alignment between offline reference and conformed shots by analysing scale, rotation and position. It can process individual frames or full sequences, automatically generating keyframes when needed. Track roles can now be assigned, such as “Offline” or “Messed up by the DIT”, simplifying comparison workflows.

Under the hood

Texture Smoothing provides subtle skin cleanup while maintaining fine texture detail. A revised caching system improves responsiveness and reliability during client reviews. Cache protection prevents critical files from being overwritten, while the new Cache View provides visibility into active and protected scenes. Processing-intensive effects, including Flexi-based operators, now support default strip caching to pre-render complex operations. Enhanced animation graphs allow per-keyframe interpolation control across transform and retime operations. Baselight v7 also integrates ACES 2.0, Dolby Vision review, and OpenTimeline I/O.

Flexi: the plug-in brain

Flexi, FilmLight’s effects and machine learning framework, now extends to developers. It supports custom models, including Segment Anything, and exposes interface primitives such as point lists, keyframe animation and tracker-linked regions. According to FilmLight, this allows facilities to integrate their own AI-based tools directly into Baselight’s stack architecture.

System and support

Baselight v7 runs on FLOS 8.4 or later and macOS 14–15 (or macOS 26 Tahoe) on Intel or Apple Silicon. Minimum recommended memory for machine-learning features is 48GB VRAM or 64GB unified memory. NVIDIA NVS 510 and K600 GPUs are no longer supported.

FilmLight describes the update as a refinement aimed at handling modern post-production complexity. Colourists are encouraged to test v7’s machine-learning features within controlled environments before full deployment.

FilmLight Baselight v7 Datasheet
https://www.filmlight.ltd.uk/products/baselight/overview/

The post Baselight v7 refines the colourist’s toolkit first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

Filmlight promises seamless operation with Apple’s CPU/GPU architecture, describing it as a plug-and-play experience that enables smooth 8K playback without performance compromises. Both Baselight S and Baselight M support FilmLight’s Slate control surface, as well as Tangent Wave and Element panels. Baselight M additionally works with Blackboard Classic.

Baselight for macOS: Subscription-Based Grading Arrives

FilmLight has launched Baselight S and Baselight M, the first software-only subscription versions of its high-end color grading system. Running exclusively on macOS, these new offerings bring the same advanced grading tools as Baselight’s Linux-based systems, now in a more accessible package.

Baselight S is tailored for independent freelance colorists, delivering the core grading features of Baselight without advanced collaboration tools. Baselight M, on the other hand, is aimed at facilities requiring multi-user workflows, offering integration with Baselight CONFORM and Baselight ASSIST, allowing studios to scale their grading and finishing capabilities.

Support and Availability

Subscribers to Baselight S and M will have full access to FilmLight’s renowned 24/7 support, the same service relied upon by existing Baselight users worldwide. Licenses are available in 3-month and 12-month options. FilmLight continues to offer its turnkey Linux-based systems, but with Baselight for macOS, the company expands its reach to a broader range of colorists—from freelancers to full-scale grading suites.

More details can be found at baselight.filmlight.ltd.uk.

The post Baselight Goes Mac: FilmLight Introduces Baselight S and M first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

When we stumbled across Nara, FimLight’s new workflow tool, we had questions – specifically, for Steve Britnell, Nara’s Head of Technical Development and Sam Lempp, the Head of Business Development for Nara. The topics? Nara, Colors, Server Racks and modern compressions. If you want to have a look at the website, you’ll find everything at nara.stream and can even directly book a demo. If you want to know what others think about it: Light Iron has it already in day-to-day use!

Recent key updates include direct media streaming with full color accuracy, deeper Baselight integration for remote scene access (but you don’t need Baselight to use it – Nara talks to pretty much anything), and enhanced project management tools. A tiered user system ensures security and streamlines onboarding. Built on a back-end index with extensive codec support, Nara natively decodes and plays back high-resolution media without the need for transcoding. Its color-accurate streaming ensures consistency across remote sessions, while its flexible project pipeline supports industry standards like ACES and many, many more. Designed for speed and efficiency, Nara optimizes workflow coordination for editors, colorists, VFX artists, and production teams.

DP: Congratulations on Nara—it looked incredible in action. Let’s start with the basics: What’s the idea behind Nara?

Steve Britnell: Nara was created out of FilmLight’s imaging expertise to transform how media is accessed, reviewed and processed. In today’s media industry, massive and complex image files are the norm—whether they’re RAW camera codecs, EXR files used in visual effects, or DCP formats for theatrical distribution. These media formats contain a wealth of information which results in large amounts of data, but this abundance of data often leads to hefty file sizes and intricate workflows. Teams must store, process, and track these files through multiple project stages, which demands robust infrastructure, significant expertise, and sizable budgets.

Traditionally, even a quick review meant enduring time-consuming transcoding—a cumbersome process that wastes resources. By contrast, Nara delivers a streamlined, secure, and fully colour-accurate viewing experience for anyone, anywhere. It eliminates the need for costly conversions, enabling creative professionals to view media in its original fidelity.

DP: Modern post-production pipelines can feel like juggling flaming torches. What challenges inspired the creation of Nara?

Sam Lempp: Modern post-production pipelines can indeed feel like juggling flaming torches, as they demand a complex web of tools, custom integrations, and ongoing maintenance just to keep projects on track. Combined, Steve and I have decades of experience in the media world and experienced first hand the amount of hours and dedication it takes to get things running efficiently – and how quickly things can take a turn. This constant balancing act was precisely what inspired the creation of Nara. Facilities have long relied on a patchwork of products—each solving only part of the problem—while devoting countless hours to manual upkeep and integrations.

By offering a cohesive framework that tackles these problems, whilst still supporting bespoke add-ons, Nara alleviates the burden of juggling multiple systems. As a result, teams can focus on the creative process rather than wrestling with fragmented workflows.

And we handle the integration with our technology partners – currently that is AWS, IBM Spectrum Scale, MASV, pixitmedia, Quantum and Qumulo.

DP: On a typical day in post-production or VFX, where does Nara sit?

Steve Britnell:  Nara’s user interface is intentionally straightforward, presenting users with a clean and visually appealing web experience. The software enables users to browse their storage by project or folder, complete with thumbnails for each media file and a detailed metadata view. At any time, any media file can be streamed directly to the user’s browser, allowing teams to quickly review the content they are working on.

Nara has a unique ability to support every media file format under the sun (almost). Supporting 160+ codecs and formats, including Camera RAW (OCF), intermediary files like EXR, TIFF, JPG as well as complex deliverables like IMF, DCP and Dolby Vision mastered files – all without the need to transcode the media beforehand. It also supports standard web friendly formats like MOV and MP4. This feature makes Nara unique in its available applications, as it can be introduced as a tool at any stage in the pipeline.

DP: How do artists, editors, and colourists actually interact with Nara?

Sam Lempp: Nara enables users to organise files and folders by “Projects,” where a single project can span multiple folders across various “storage silos”. While it may sound straightforward, the ability to consolidate all your files—regardless of their physical location—and make them fully searchable, filterable, AND streamable is a genuinely unique offering. Other applications might allow you to organise or stream files, but seldom both in a single solution and never with such a diverse set of supported file types.

By serving as the central project hub, Nara provides a unified space where every member of the facility can locate and review ANY media. Teams can also share links to ensure everyone references the exact same file, preventing confusion and streamlining collaboration. Users can also securely stream their media to external clients or vendors. This allows them to service their clients in a unique and powerful way. Let’s say their client wants to see a VFX shot in its native format, they no longer have to book in a session to review that shot, or compromise by transcoding it and uploading it, they can simply put that media in front of their client instantly.

DP: Let’s talk technical ecosystems. How does Nara connect to existing post-production pipelines?

Steve Britnell: If you were to ask an editor, colourist, or VFX artist, “Who is most frequently knocking on your door with daily requests?” the answer is likely, “Producers.” According to current Nara users, the platform helps save valuable time in the suite by offloading simple tasks—such as quickly checking a shot, reviewing newly delivered VFX assets, or confirming that a client-reported issue is resolved in the latest WIP. Instead of tying up costly systems and suites for these routine tasks, teams can use Nara to handle them independently. As a result, they gain full autonomy over smaller tasks, freeing the suite for the in-depth creative work that truly demands specialised resources.

Nara does have some very unique and powerful integrations with Baselight which allow users to browse and stream active Baselight projects, as well as add comments and notes to the timeline, giving users a way to work even closer to their artists and get a true picture of their current project. We also plan to release integrations with other NLEs and finishing software in the future.

DP: From an IT perspective, what does a Nara system look like? Is it a box in the studio, a cloud application, or something else entirely?

Steve Britnell: Nara is primarily an on-premise solution, consisting of a 1U rack server, GPU and NVMe. Fitted with a fast network card supporting anything from 1Gbe to 100Gbe depending on the facilities infrastructure. We can also deploy in a facility’s cloud estate, as an AMI. Whether in the cloud or on premise, Nara supports all types of storage such as NAS, SMB, S3 and Native distributed storage filesystems such as IBM scale, Quantum StorNext, Weka, Qumulo etc. Nara is not shy to storage, if you are unsure if we support it, the answer is we most likely do – just ask our friendly support team!

And, regardless of how Nara is deployed or what storage it mounts, the user will always interact with it via a browser – meaning there is no requirement to have an installer on the browsing device.

DP: What hardware set-up is necessary to run Nara and achieve full-quality playback?

Steve Britnell: We have a minimum spec for a Nara server which mostly puts a minimum spec on the CPU and GPU capability. A top spec GPU is required due to the intensive decode requirements for a lot of the media we support, and the CPU needs to be able to keep up with the index demand and multiple user access. However the device used by the Nara user can be any Laptop, desktop or tablet which supports Safari, Chrome or Firefox. We currently do not support mobiles. 

Nara does have the ability to play out to a JPEG-XS decoder and output to a high fidelity calibrated monitor. In this case there is a hardware requirement on the client end. This uses our FL-REMOTE software which you can find information about on our website.

DP: Let’s talk connection speeds. How fast does my internet need to be to enjoy a “butter-smooth” experience with Nara?

Sam Lempp: Nara allows users to customise their streaming output according to bandwidth requirements—whether that means lowering bitrate for slower connections or maximising it when faster speeds are available. Within the browser, Nara supports bitrates anywhere from 1 Mbit/s to 100 Mbit/s. In fact, we’ve successfully streamed images on notoriously poor trade-show Wi-Fi; if you can stream 8K R3D media at peak NAB, you can stream it anywhere (laughs).

The real bandwidth demand lies on the server side (where the magic happens). As long as Nara resides on the same production network as your storage, it will have sufficient bandwidth to stream the media. Additionally, Nara locally caches data on the server, reducing the frequency of reads from your production storage and ensuring a smoother experience overall.

DP: How does Nara compress video for streaming while maintaining quality?

Steve Britnell: Bandwidth will always demand compromise in bit-rate, but image quality isn’t solely about bitrate—accuracy is what plays the most significant role. FilmLight has over 25 years of colour expertise, bringing unparalleled know-how to Nara. Our colour engine is fully integrated into the platform, ensuring that every frame is precisely managed from source to stream. Users can even configure their own custom colour pipelines without sacrificing simplicity or accuracy. While bitrate is important, it isn’t always the key factor in achieving an accurate image; what truly matters is viewing the content consistently, with accurate colours maintained throughout.

DP: How do you ensure that colour, dynamic range, and resolution are preserved so the viewer gets the “original”?

Sam Lempp: Nara leverages the very same colour engine as Baselight—trusted by professionals worldwide to deliver Hollywood blockbusters and record-breaking OTT series. This powerful engine uses camera-native SDKs to ensure that each file is processed precisely according to its manufacturer’s specifications, while also adhering to industry standards for codecs and formats. With Nara, users can customise their colour pipelines per project, enabling advanced workflows such as ACES, RED IPP2, ARRI REVEAL, or any bespoke pipeline they choose.

DP: Colour needs calibration—what’s “good enough” for screens receiving streamed footage, and how do you ensure it?

Steve Britnell: For the highest level of accuracy, we always recommend calibrating your monitor. However, modern displays—like Apple’s XDR—already offer impressive precision. Nara automatically detects each device’s colour capabilities, so it will not, for example, stream HDR content to a screen lacking HDR support. In most cases, users can simply access Nara through a web browser without needing to calibrate their display, and still benefit from a highly accurate image.

DP: What’s the reaction from users so far? Any feedback that made you think, “That’s why we built this”?

Sam Lempp: We’ve had overwhelmingly positive feedback, and it’s been incredible to discover just how versatile Nara really is. At FilmLight, we’re naturally immersed in the realms of colour and finishing, so seeing Nara applied to such a broad spectrum of workflows—beyond what we’d typically be involved in—has been the most rewarding aspect of its release. We’re working closely with our customers and are always developing our solution to ensure Nara meets their specific needs – we look forward to continuing this journey.

DP: Every product ends up with some unexpected “off-label” uses. Has anyone done something crazy or unexpected with Nara yet?

Steve Britnell: Given Nara’s unique capabilities, we’ve had the opportunity to demo it in some unconventional settings. It never gets old seeing the reaction when Nara streams back high resolution OCF media or a finished DCP whilst we’re drinking coffee at a local coffee shop using our 5G phone Wifi, or whilst on a train back from IBC during an impromptu demo. This really resonates with people interested in how they can give more autonomy to remote staff and how they can improve their workflows between remote facilities.

DP: What’s next for Nara? What can users expect in the short term—say, in the next year?

Sam Lempp: We’re making significant progress on a range of new features, with a major update planned around NAB. One key area involves expanding our secure streaming capabilities—currently, users can stream media live to clients, and we’re now working on making this asynchronous. In other words, clients will soon be able to access their media whenever and wherever they choose, without needing a live session. All whilst leveraging their existing infrastructure.

We’re also introducing transcoding within Nara, allowing it to serve as an internal render engine. This lets teams integrate it directly into their core pipeline, for use cases such as easy rendering of stills or QuickTime clips from specific timelines or shots without adding extra load on artists.

Furthermore, we’re adding tools for semantic data generation, giving users the option to manually tag media and folders or leverage ML models to automatically tag faces, places, and objects for improved searchability. Alongside these features, we’ll be rolling out enhanced user access controls, so facilities can organise teams by project and assign features accordingly. And there’s still more to come—but we’ll keep a few surprises under wraps for now!

DP: Let’s get ambitious. Where do you see Nara—and production pipelines in general—in 2035? Will it be all AI and holograms by then?

Sam Lempp: It’s remarkable that in 2025 the status quo is that viewing media in a facility without Nara is often a painful experience, with countless steps required to achieve something that is seemingly so simple yet in reality is extremely complex. So yes, we have big ambitions for Nara and absolutely see it as a central part of M&E workflows moving forward.

Steve Britnell: ML and AI are of course huge topics of discussion, with more and more mind bending technology being released each week. It’s still unclear how these developments will ultimately shape our industry, but something will always remain – we will always have a desire to create our own stories. For us to continue doing this, we have to make sure those working on said stories are able to collaborate effectively and efficiently on the media they are using, wherever it may come from or whatever form it might take.

The post FilmLight sends your frames! first appeared on DIGITAL PRODUCTION and was written by Bela Beier.

A rough understanding of the evolution of colour correction helps in the evaluation of basegrades. The origins of colour grading as we know it today go back a long way, to the early days of television and cinema. For TV, video signals from television cameras or from a film scanner have always had to be levelled or corrected. The profession of colourist originated in the telecine, where film material was converted into a pleasing video signal.

VideoGrade

The technicians provided the colourist with four basic technical parameters with which he could process the video signal: Lift, Gamma, Gain and Saturation. These are still among the most popular grading tools today. In Baselight, they can be found in the VideoGrade Operator. With Lift, which is sometimes also called Pedestal, the colourist adjusts the black level and with Gain the white level. Experienced Photoshop users will find it easy to visualise the resulting gradation curve: Lift sets the starting point (bottom left) and Gain sets the end point (top right). When working with VideoGrade, these are the most important reference points. Gamma is technically a power function whose only parameter is the exponent. The gamma function determines the curvature of the gradation curve between the two end points.

These three parameters can be adjusted not only for brightness, but for all three colour channels of the video signal (RGB). However, the operator is not usually presented with individual controls for red, green and blue, but one for the brightness signal and a two-dimensional one for the colour component. This results in the basic structure of all grading panels: three spheres, which adjust the colour in two dimensions, and a rotating ring around or next to them for the luma setting. Lift is on the left, gain on the right and gamma in the centre. Colloquially, these three parameters are often referred to as shadows, mid-tones and highlights.

A very popular and solid working method used by video style colourists is to first adjust the black level and white point of the image. This involves balancing all three channels in both black and white just before the clipping points. The brightest point in the image is then pure white, i.e. without a colour cast and with maximum brightness, and the darkest point is pure black. This is often referred to as “clean” black and white. The gamma parameter is then set. This regulates the “airiness” and “heaviness” of the image. In other words, a combination of brightness and contrast. If you want to colour the image, for example, you often do this via the gamma, as black and white then remain “clean”.

Baselight provides two modes for VideoGrade. The standard mode is RGB. In YCbCr mode, the luma channel is processed in isolation. Changes in brightness then have no effect on the colour and saturation of the image. Over time, VideoGrade has also become the most important tool for telecine-style colourists. The colourist manually converts an image from a colour space with a high contrast range, such as log coding, into the output colour space. VideoGrade is therefore now not only applied to images in a video colour space, as originally intended, but also to images in a log colour space.

Lift, Gamm, Gain is probably the most frequently used grading operator in the video sector at present. However, it originates from a time when the handling of specular highlights, for example, was of secondary importance. Burned-out windows or overbright skies were tolerated as long as the faces were recognisable. A soft clip, i.e. a smoother transition to overshoot, is not possible with VideoGrade alone. Over time, video colourists therefore developed various techniques to meet this aesthetic requirement. Gradation curves, luma keys, blend modes or dedicated soft clip operators were used for example.

Another disadvantage becomes apparent in VFX workflows. Contemporary compositing works most realistically in a scene-linear colour space. The pixel values are proportional to the photons on set. Pregrading is nevertheless often helpful so that the basic brightness and white balance are correct and the individual shots in a sequence match each other. Unfortunately, lift, gamma and gain destroy the scene linearity and make VideoGrade unusable for this type of VFX pregrading.

FilmGrade

Long before the telecine colourists, the profession of the film light setter emerged. They influenced the look of a cinema film via the intensity of the copy lights and the chemical processes. With the introduction of the digital intermediate process around the year 2000, the new profession of DI colourist emerged. This person processes cinema images digitally before they are exposed on film material and copied onto print material.

The source material was also usually a film scan. The digital intermediate gradually replaced the analogue intermediate process and thus the creative part of determining the light in the copying plant. Film is the all-determining factor in the DI process, as you can only produce colours that can also be reproduced on film material. The analogue process up to the cinema copy is therefore simulated live in grading using a so-called Film Print Emulation LUT. The obligatory film LUT, which is used in the preview output, means that the image reacts differently to the colourist’s inputs than in telecine or video mode.
The manufacturers have developed new grading tools for DI processing, which are based on analogue light determination with copy lights. In Baselight it is called “FilmGrade”. FilmGrade is designed for processing images in the Cineon-Log colour space. After the colour correction, a conversion to a display colour space takes place, for example classically via a LUT or, since Baselight 4.4, via shaders with truelight colour spaces.

FilmGrade consists of a total of six tools, which are divided into two tabs. The main page consists of Exposure, Contrast and Saturation. The second page consists of Shadows, Midtones and Highlights. All tools offer adjustment options via a ball and a rotating ring. The most important tool is Exposure, which can also be adjusted in RGB copy light steps, so-called printer points, via the blackboard panel. The colourist uses Exposure, also known as Offset in other grading programs, to adjust the brightness and the sphere to adjust the colour of the image. Film-style purists try to work with exposure as much as possible, as this simulates an analogue light setting and the image remains very natural. The contrast in all colour channels and the scene linearity are retained when changing the exposure, for example. Shadows, midtones and highlights do not correspond to lift, gamma and gain, as the individual areas are limited by pivot points. If, for example, the shadows are lifted or coloured using Shadows, this is only done up to a defined point in the curve.

With Lift, on the other hand, the entire image is processed, but the shadows are the strongest. In film workflows, the LAD grey test field approximately in the middle of the curve is the anchor point both for calibration and for the FilmGrade colourist. It was defined by Kodak and is a little darker than 18% medium grey. Full black and white are difficult to determine on film because the curve is very flat. This is why a film-style colourist, unlike a video-style colourist, is not so concerned with achieving 100% white or black. With this concept, a soft clip is already active via the simulation of the print material. The visual impression takes centre stage. This grading concept is therefore more natural than VideoGrade.

There are also fundamental problems with this concept. Although controlling the brightness via Exposure works quite naturally, it is not exactly the same as changing the camera aperture or the ISO value. This is due to the log coding used, which does not define the black level exactly to zero, for example. The detailed reasons are beyond the scope of this article.

FilmGrade is now used not only on Cineon log data, but also on any type of log coding such as LogC from Arri. On the output side, not only film emulations but also modern approaches such as ACES are used. This brings us to future-proofing: log coding can only store a limited dynamic range. The Cineon curve, for example, was no longer sufficient for the high contrast range of the Alexa camera. This is why Arri developed the LogC curve. Future camera generations and HDR displays will require further adjustments, which may involve compromises. This also applies to the VideoGrade operator, which was originally only designed for video signals with a standard dynamic range.

BaseGrade – under the bonnet

The time is ripe for a next-generation colour grading operator. Filmlight uses neither a video signal nor film material as its foundation, but algorithms that are modelled on human perception. The dynamic range is not limited by a technical format such as Rec. 709 or Cineon coding, but is ready for the future characterised by HDR.

BaseGrade works identically in every working colour space and therefore always feels the same to the colourist, regardless of the camera used. However, in addition to a sensible colour setup in the scene settings, this also requires the correct keywording of the material, which is usually done automatically. BaseGrade autonomously converts the image into a linear colour space in which the original brightness ratios of the scene on set prevail, as in linear compositing. The user is not aware of the colour space conversion; the next operator in the stack receives the image in the defined working colour space again.

BaseGrade uses an internal colour model based on Lab, which consists of a pure lightness component L and two colour components a and b. The colour plane spanned by a and b has been distorted by the developers with regard to colour correction so that colour and saturation changes have the same visual effect in all colour areas and work feels more intuitive. Four parameters affect the entire image: Flare, Balance, Contrast and Saturation. In addition to these global parameters, BaseGrade divides the image into brightness zones. This is immediately reminiscent of Ansel Adams’ legendary zone system, which, according to the developers, also served as a source of inspiration.

At first glance, there are few parallels to existing tools in Baselight or comparable video grading software such as Resolve. Raw development in Adobe Lightroom comes closest to working with BaseGrade. However, BaseGrade not only offers brightness controls for each zone, but also colour and saturation. In addition, each zone area can be fine-tuned via pivot and falloff. All exposure settings and pivot points are specified in f-stops. This also comes close to human perception and helps photographers and cameramen to understand them. Stops are also a widely recognised and established unit.

Baselight 5.0 

Users are provided with new builds every few weeks, but now version 5 is finally in the starting blocks. The significance of this release becomes clear when you consider that the last full version jump from 3.O to 4.O was more than seven years ago (2OO9).

Although BaseGrade is the most significant innovation in the upcoming version, it is of course not the only one. For example, the software will be greatly expanded in the finishing area with a dedicated blue/green screen keyer, perspective tracker and warper as well as a grid warper and a paint tool. In the plug-in area, support for the Autodesk Flame matchbox shader format and GPU acceleration for OFX will be added. 

CGI renderings with WSP and normal maps can be relit in Baselight 5 or individual objects can be graded separately. And the powerful colour management has been further developed with a special focus on HDR and has been structured more clearly for the user. DP will be looking at further new features in a future issue. 

The colourist can now give the camera operator direct, meaningful feedback, such as: “I have made the entire image half a stop brighter” or “I have lowered the highlights by one stop”. If a Dailies colourist works with BaseGrade, it is even conceivable that such feedback – like the copy light report in the past – will help the cameraman when working on set. In any case, communication is simplified, especially when the cameraman cannot be physically present during grading.
BaseGrade’s reference point is medium grey, as found on 18% grey cards. The brightness zones are defined from there in f-stops. A correction of three f-stops up or down is the maximum in standard mode and is sufficient. For extreme cases, up to six f-stops can be corrected in “Extended mode”.

User interface

The user interface is based on VideoGrade and FilmGrade. Three main parameters at the top, which are mapped to the three sphere-ring combinations on the blackboard panel. Below this is a visualisation of the current grade as a gradation curve, framed by other parameters such as pivot points. The developers have divided BaseGrade into two tabs. The first page is called Dim/Balance/Light and the second Dark/Balance/Bright. The most important parameter, Balance, is permanently visible and mapped to the centre sphere/ring combination on the blackboard. Flare, Contrast and Saturation are also visible on both sides and can be adjusted via potentiometers. As with all other tools, the colourist can of course adjust everything in the user interface and on the blackboard to suit their individual requirements.

The developers have come up with something special for visualisation using a gradation curve. They superimpose a luma waveform display of the current image over the curve. This means that the colourist can always see which parts of the image he is currently working on and how he should readjust the pivot points if necessary
Pivot points should be readjusted.

Practical test

A little familiarisation time is needed to find your way around. However, the curiosity of having a revolutionary tool at your fingertips makes it easy to get started, after which the results motivate you to continue.
Firstly, the promise of the exact aperture scale is checked. And indeed, increasing the balance by one f-stop is exactly the same as doubling the ISO value in the raw developer. And this applies to all tested cameras from Arri, RED and Sony. The best thing about this is that no raw material is required. For example, if you are working with a mixed ProRes and ArriRAW timeline and all raw shots have been pushed by one f-stop, you can now apply the exact same correction to all shots. A cumbersome switch to the De-Bayer settings is no longer necessary and does not bring any qualitative advantages compared to working with BaseGrade.

Adjusting from shot to shot seems to be another ideal task for the tool. Many colourists divide their grade into a base correction per shot and the creative look. Used as the first layer in the stack, BaseGrade’s scene-linear functionality makes it easier to compensate for exposure differences. Shots with diffuse lens flares, which appear in the image as a raised black, are a common problem when equalising. Even technically high-quality lenses such as the master primes show this effect, especially in scenes with light-coloured backgrounds. If you tried to remove the flare using Lift, for example, the entire image would change and the grade would no longer fit correctly. The flare parameter helps in these cases as it corresponds to optical stray light. If the exposure is basically correct, the differences in the black level can be corrected well using flare.

When copying corrections from shot to shot, the colourist should also pay attention to the flare value and readjust it if necessary. For example, if you copy from a scene with strong stray light, the black in the new shot may be completely washed out. With none of the other controls except Flare can you get it back to a decent level. However, if Flare is set correctly, the black can practically never be crushed. In BaseGrade, the shadows are pressed into a pleasant-looking compression and not clipped hard.

The Saturation slider is also pleasant to use. It is amazing that something as widespread as colour saturation could be improved even further. In direct comparison, BaseGrade behaves visually more evenly than existing implementations. Previously, primary colours such as red, for example, quickly became overweight in the image when saturation was increased. Not so with BaseGrade, the strength of the effect is distributed more evenly across the colour wheel. And the best thing about it is that it works in the same way when desaturating images.

BaseGrade shows its particular strengths through the zone model. The gradation of an image can now be modelled in great detail with just one operator. Previously, this required detailed and sometimes quite fiddly tweaking of the curves in CurveGrade or Keying, but now there is a more intuitive alternative. Extracting detail from a sky normally requires a luminance key on the highlights. With BaseGrade you can get surprisingly far without any secondary correction. Bright is lowered and by raising Light you can tease out the last details until just before clipping. If you initially set the correction too high, it is easier to find the right pivot points. You can then reduce to a realistic level.

If the colourist opts for extremely strong corrections, there is a risk of unsightly effects, a preliminary stage of solarisation, so to speak. Although BaseGrade prevents true solarisation, i.e. negative gradients in the gradation curve, the colourist can flatten the curve so much that the drawing is ruined in certain brightness zones. In these cases, a larger falloff, the transition area of the zone, usually provides a remedy, but at the same time reduces the effect slightly.

BaseGrade can also be convincing with a common technique in colour correction. Cross process toning is a popular stylistic device that makes images more interesting in terms of colour. This involves colouring the highlights and shadows in different shades. Complementary colours are often chosen for this: the shadows, for example, in cold turquoise blue and the highlights in warm orange. Colourists often use HLS keys so that they have better control over the effect. With BaseGrade and its zone model, you can precisely control the strength and colour tone of the effect. This will probably save a few layers in the timeline in the future.

On the subject of VFX pregrading: Flare is an artefact that occurs in the camera optics and distorts the scene linearity. With the help of BaseGrade, the real brightness conditions on set can be reconstructed using a correctly set flare value. Afterwards, Balance is used for complete scene-linear colour correction. The flare correction makes BaseGrade more suitable for scene-linear pre-grading than Exposure/Printer-Lights in FilmGrade.

Conclusion

BaseGrade is powerful, but also complex. A properly set up project is the be-all and end-all for correct functioning. If the colour spaces are not correct, it will not work properly. It is advisable to familiarise yourself with the tool slowly at first and then integrate it into your daily work bit by bit. A good way to start, for example, would be to make all saturation corrections with BaseGrade. The next step would be to use it for basic corrections from shot to shot, and so on.

The new grading concept could also be a good introduction to moving image colour correction for photographers with Lightroom experience and, thanks to the intuitive and aperture-based approach, also for cameramen.
The first few days with BaseGrade were very promising: the tool has the potential to create both more natural and possibly completely new looks. It also makes day-to-day work easier. The decisive factor will be how well it is accepted by users.

The post BaseGrade and the evolution of colour grading first appeared on DIGITAL PRODUCTION and was written by Bela Beier.