<\/p>\n
\n
I arrived at the Westin Mission Hills Resort in Rancho Mirage, CA,<\/strong> early Tuesday morning, for the second day of the 2024 HPA (Hollywood Professional Association) Tech Retreat, and the writing was on the wall \u2013 or should I say: the digitally entered text that generated live-action video content loomed large on the venue\u2019s three big display screens.<\/p>\n I\u2019d missed the opening day\u2019s session, kicked off by Seth Hallen, Managing Director of Light Iron and HPA President, along with HPA Board Members (as well as HPA TR-X Co-Chairs) Mark Chiolis (Mobile TV Group) and Craig German, themed around \u201cextreme\u201d workflows. By all accounts, Monday\u2019s session \u2013 which included IMAX\u2019s David and Patricia Keighley talking about setting up Oppenheimer <\/em>for 70-mm celluloid distribution; SMPTE\u2019s Renard Jenkins in conversation with Darryl Jefferson, SVP, Engineering & Technology, NBC Olympics & Sports; and an Extreme Sports panel that included Rob Hammer, SVP Broadcast & Studio Operations, for the World Surf League \u2013 was compelling and filled with industry players at the leading edge of live streaming and broadcast content. But it was mainly a prelude for Tuesday (and most every other session that followed), which dove deep into the hopes, wonders, and fears of AI, which nearly everyone at the Tech Retreat described as \u201cthe single biggest technological disruptor\u201d since the advent of sound a century ago.<\/p>\n \u201cWe\u2019re all intrigued by how these evolving technologies are reshaping our industry, impacting our workflows and processes, and the roles we play in that,\u201d Hallen announced in his opening remarks for Tuesday\u2019s Super Session, which deconstructed the making of the animated short film Ashen<\/em> in the morning and then checked in with a group of artists and filmmakers behind the AI-animated short Trapped in a Dream<\/em> after lunch. The final panel on Tuesday brought together participants from both projects to compare and contrast the leading edge of traditional filmmaking workflows (used for Ashen<\/em>) with the promise (and current reality) of generative AI (used for Trapped in a Dream<\/em>).<\/p>\n \u201cWhen we started to engage with Niko Pueringer (co-founder of Corridor Digital, whose YouTube channel is infused with generative AI content), the learning began,\u201d Hallen added. \u201cAnd when we presented Niko with the script that [Ashen <\/em>Director\/Producer] Ruby Bell\u2019s team was working on, that sparked an insightful revelation. While AI\u2019s potential is undeniable, the narratives best suited to its strengths \u2013 at least as of today \u2013\u00a0require a different approach. AI is a new way of telling stories, so it wasn\u2019t feasible to use the same script for both of today\u2019s case studies.\u201d Hallen noted that \u201cwhat we came to realize [from conversations with Pueringer and other AI-focused filmmakers] set the stage for an adventure that is engaging, surprising and precisely what HPA\u2019s Super Session is all about \u2013\u00a0to show, not just tell, how these new technologies \u2013 including the cloud workflow employed by a global team of filmmakers on Ashen<\/em> \u2013 are impacting the business of storytelling.\u201d<\/p>\n <\/p>\n <\/p>\n Along with Bell, writers Jack Flynn and Charlotte Lobdell, and<\/strong> VFX Producer\/Animator Shaman Marya, Ashen<\/em>\u2019s team included ICG Director of Photography Mandy Walker, ASC, ACS, who came on as Executive Producer. Moderator Joachim \u201cJZ\u201d Zell (ACES Project Vice Chair at the Academy of Motion Picture Arts and Science) introduced Walker as \u201cthe person everyone knew when we went to Panavision, ARRI, Sony, and RED needing camera, lighting, or grip equipment for Tangent <\/em>[the HPA 2021 Super Session short film also directed by Bell]. But now we\u2019re in animation, and no one knows her in this area. And there was a lot of explaining to do. \u2018So, Mandy, why did you get involved with this?\u2019\u201d<\/p>\n \u201cAs I was an executive producer on Tangent<\/em>,\u201d Walker shared, \u201cI was excited to come back on another HPA film. I so appreciate the HPA\u2019s support for filmmakers, and they were incredibly generous in facilitating this project we\u2019re presenting today. It was a labor of love for everybody involved; and as the executive producer, I have to thank all the vendors and various people who helped us out. I\u2019ve not had any experience in animation, so as a cinematographer, it was a learning opportunity. I had worked a little bit in virtual production on a movie with Shaman and [Ashen <\/em>Virtual Camera Supervisor] Jackson [Hayes], and I brought them back.<\/p>\n \u201cIn my job on set,\u201d Walker continued, \u201cI am \u2013 more and more \u2013 interacting with visual effects. So understanding those tools in a cloud-based system, and working with a virtual camera that\u2019s advancing all the time, as well as the Unreal game engine system that\u2019s always getting better, is important to me as a cinematographer. Working all over the world, remotely, has been an exciting journey. A lot of people stepped up to do something they hadn\u2019t done before, or they pushed themselves to use their experience in a new way. That\u2019s been great to see.\u201d<\/p>\n Walker, who earned an ASC Award and a 2023 Oscar nomination for Elvis<\/em>, went on to caution the 1000-plus attendees at HPA that \u201cwhat you are about to see [with Ashen<\/em>] is a work in progress, and we still have a long way to go. So please be generous and open-minded,\u201d she smiled. Zell then asked Walker if she or her agent had ever been approached by major animation producers, like Sony, Pixar, or Disney, to supervise the cinematic look for an animation project. \u201cIs this happening?\u201d he wondered. \u201cIs this a thing?\u201d<\/p>\n The DP replied that she did get asked \u201conce, a long time ago. But, now, I wish it would happen more because I find it all incredibly interesting. The virtual production part of it, especially, is becoming more integrated into what we do as directors of photography. And getting involved, very early on, with the VFX Supervisor [on a new project] is becoming a standard thing. Understanding all of these new tools is important.\u201d<\/p>\n <\/p>\n <\/p>\n To highlight Ashen<\/em>\u2019s complex preproduction process, Bell<\/strong> presented a series of slides that included lookbook imagery and environmental\/character thematics. She explained that \u201ca lot of us come from live action, so we had to lean on the skills we knew. We had to do a lot of rehearsals on Zoom, with our actors, who were all over the world. So, we needed to develop a visual language early on, so everyone had a sense of what a child\u2019s horror film felt like.\u201d<\/p>\n Marya, who\u2019s worked extensively with Unreal Engine in his previsualization work, added that \u201cthe thing I\u2019ve always struggled with is: how do we finish in Unreal Engine? That\u2019s not a new conversation, but when Mandy and Ruby said, \u2018We want to make an animated film and we don\u2019t have any budget,\u201d my response was: \u2018I guess we have to finish in Unreal Engine. That\u2019s the only way it will work.\u2019 It seemed like a crazy idea, and we\u2019re still only about 70 percent there.\u201d<\/p>\n The VFX producer, who joined the project in the summer of 2022, was given character and environmental mock-ups by Bell, which he calls \u201cthe first sprinkling of generative AI to make its way into our workflow. Ruby had these great comps, and we had a concept artist coming in soon after,\u201d Marya continued. \u201cThe idea was: how could we provide material to the concept artist that would elevate [the building of the characters and environments] to another level? I was two days away from canceling my [AI art generator] Midjourney subscription when Ruby asked for our help, and I was like: \u2018Okay, I guess I\u2019ll pay a little more.\u201d<\/p>\n Global Strategy Leader, Content Production, Media and Entertainment for Amazon Web Services Katrina King supplied a key piece of Ashen<\/em>\u2019s workflow puzzle when she spoke to the audience about color finishing in the cloud. \u201cLast year on this stage, Annie Chang said she thought that would be the last thing we addressed in fulfilling MovieLab\u2019s 2030 Vision,\u201d King said. \u201cBut I\u2019m going to challenge that because, without the ability to finish in the cloud, it doesn\u2019t make a whole lot of sense to ingest your assets in the first place. The next time you\u2019re using them would be for VFX pulls, conform, and finishing. So moving all of that to the cloud [without finishing capabilities] would be akin to building a bridge to nowhere.\u201d<\/p>\n King then ran down a visual presentation that revealed how AWS designed a comprehensive production\/postproduction workflow for Ashen <\/em>that included asset management and encoding, transport protocols, color grading, and more. \u201cWe did a Flame implementation with the framebuffer feeding to the Colorfront Screening Server,\u201d she explained. \u201cThat supports 12-bit 444 input, streams over 10-bit over SRT, and encodes on the player side, using a Colorfont Streaming Player that\u2019s routed through an HAA card \u2013 either 10- or 12-bit \u2013 to a reference monitor. The Colorfront Streaming Server also supports Dolby Vision tunneling.\u201d<\/p>\n Pointing to the large room displays, King added, \u201cThis is the implementation we used for a Flame conform in the cloud to connect to a reference monitor. For color grading, we\u2019re using Resolve, which leverages the Nvidia on-board encoder that feeds up to a 10-bit 444 encrypted signal over WebRTC to a client machine, which is then fed to a Blackmagic Design Decklink or Ultrastudio. The final mile is over copper SDI to the monitor itself.\u201d<\/p>\n <\/p>\n <\/p>\n AWS Technologist Jack Wenzinger led off Tuesday\u2019s afternoon session <\/strong>with thoughts about disruptive technology, noting that \u201cJZ came up to me in 2019 to propose the [HPA short film project] Lost Lederhosen. <\/em>Back then, the cloud was considered disruptive, but no one was sure what it did,\u201d he recounted. \u201cGame-engine technology was also coming into our space, and we weren\u2019t sure what that meant for virtual production and VFX. We answered a lot of those questions that year. And since then, [HPA] has continued to bring forth projects that push the needle of innovation. We owe a lot of thanks to people like Mandy Walker, JZ, Ruby Bell, and the many cinematographers who have helped build this out.\u201d<\/p>\n Wenzinger said the key to all of the HPA-featured projects has always been, \u201cHow do we go from disruption to adoption? Everyone\u2019s wondering what generative AI is and how it will impact our industry. But disruption has always been there,\u201d he continued. \u201cAll the innovation happens when we harness the disruptive technologies that are mature enough to create a new workflow. Have you ever once read about a new movie that\u2019s come out where they said, \u2018Yep. We did it exactly the same as the last one!\u2019? That just doesn\u2019t happen.\u201d<\/p>\n Corridor Digital\u2019s Pueringer was next up, explaining that \u201cAI software is changing so fast, the original idea Seth and Jack brought to me, about comparing a film like Ashen<\/em> \u2013 created with leading-edge animation tools like CG, Unreal Engine, cloud management, et cetera<\/em> \u2013 with an AI animation, which, in itself is a super-broad, non-specific designation, was not going to hold up. The AI-animated short [Trapped in a Dream<\/em>] you\u2019re going to see was created with software that came out three weeks ago! With such rapid change and an inherently experimental medium, it didn\u2019t make sense to present a workflow that we couldn\u2019t check back in on. The next best option was to showcase an experimental AI project and include on-stage demos, along with artists committed to the space who have a lot of experience.\u201d<\/p>\n Pueringer said that one of the main goals of his HPA presentation was \u201cto dispel all the hype about AI and show you tools that can be used right now in your pipelines. As I said, AI animation can be many things: a way to generate imagery, a way to generate motion, a way to render existing motion, a fancy filter to use for previsualization, and the list goes on. As long as it\u2019s based on some sort of statistical machine learning model that\u2019s giving you imagery, AI animation can be anything you want it to be.\u201d<\/p>\n Describing AI animation as a \u201cnew software paradigm,\u201d Pueringer added, \u201cYou guys are probably all used to non-linear editing, node-based and layer-based compositing, and various ways of managing your data. AI animation is a new piece of software that\u2019s never existed because you need to have a way to manipulate and adjust these trained statistical models \u2013 to pull something out that is your creative vision. There\u2019s no traditional workflow with AI, and the software is being developed open source by thousands of different users, with thousands of different goals. Which can lead to a few downsides. It\u2019s very fast-moving, and mostly experimental, so things break all the time.<\/p>\n \u201cWhat I\u2019m showcasing today is centered around latent diffusion, a way of generating and understanding images in a machine learning model. This was a breakthrough that basically started with training a computer to de-noise an image. We add more noise, it takes it away until you are giving the computer pure noise \u2013 no image at all \u2013 and, as it turns out, the computer can imagine an image from all that noise.\u201d<\/p>\n <\/p>\n <\/p>\n The AI short Trapped in a Dream<\/em> was a psychedelic, color-saturated<\/strong> video meant to visualize the potential of AI tools in content creation. With a background in VFX, Pueringer detailed how \u201cthere are three ways for us to generate a visual asset at this point. Photograph it with a camera or physically render it, whether that\u2019s digitally or with a paintbrush or a stick in the sand. The third way, which came to light in the 1990s, is computer renderings of geometry, aka CGI, 3D animation, et cetera. That\u2019s pretty much it. Photography, illustration, or computer math. However, with AI, we have a fourth way of creating a visual asset, and that\u2019s with a neural network, which is distinctly different from computational geometry. We\u2019re looking at inference instead of calculations, imagination instead of math.\u201d<\/p>\n Tim Simmons, whose background includes sharing an Oscar with the New Line Cinema team from Lord of the Rings<\/em>, came onstage to supply a recent history of AI content creation. Simmons\u2019 YouTube channel, Theoretically Media, aims to demystify AI in a way that\u2019s accessible to creatives of all skill levels. \u201cBecause I come from the same traditional world of filmmaking that many of you do, I have created all my AI tutorials with that in mind.\u201d Simmons presented various examples of generative AI content over the last two years, including Will Smith eating spaghetti, The Family Guy, and his own fake AI documentary that imagines a \u201clost film\u201d in the Alien <\/em>series directed by John Carpenter using voice cloning of British legend Michael Caine.<\/p>\n \u201cThe world changes when Runway Gen-2 AI is announced, and we all kind of lose our minds and feel, for a moment, that we all just stepped into the future,\u201d Simmons added. \u201cSynthetic Summer<\/em> [created with Runway Gen-2] was an AI-generated beer commercial that went viral in May 2023; and while it was weird, it was also kind of awesome. In July 2023, a mysterious video appeared on the scene from an AI generator called Pica, and that was also equally weird and surreal. People discovered that if you play around with an area the AI model knows, you can get some remarkable results. The Dor Brothers did this with the video Can You Paint a Taste? <\/em>The Dors figured out that Pica knew who Van Gogh was. You could prompt for actions in a Van Gogh style and end up getting something very cool.<\/p>\n \u201cThe Dors video runs at eight frames per second and yet is still very smooth looking. The next big leap, literally two weeks later, was image-to-video, and thus began the era of \u2018parallaxing people staring at nothing,\u2019\u201d Simmons announced to much laughter. \u201cThe natural place people leaned into for this was making trailers, and one of the best early examples was Nicolas Neubert\u2019s trailer for the sci-fi epic Genesis.<\/em> But there started to be a backlash against generative AI as people were mostly doing the same thing and not using the tools to innovate. That\u2019s when people like Sway Molina, using an early version of Kaiber AI, came along to add a lot more stylization to the technology. One of my favorites was a short film called Sh*t!<\/em>, from Abelart, where everything is run through After Effects with tilt shift, out of focus, and camera shake, to create a much more cinematic look with generative AI.\u201d<\/p>\n The last panel of the day, Bringing It All Together<\/em>, was moderated by Hallen and included many of the participants from throughout the day. As Hallen noted: \u201cA lot of people came up to me during the breaks today with questions that I couldn\u2019t answer, so I\u2019d like to open this up to our audience and have the panel help them out.\u201d One attendee announced that \u201cthis is kind of a \u2018bummer\u2019 ethics question that has to be asked. Everyone has talked about how these new AI tools will democratize what it means to be creative. But for those who are trailblazing AI, how do you ensure that the knowledge created by those who pioneered all of these filmmaking techniques AI is building on is carried forward? Eventually, if it\u2019s model on model, how do we source that knowledge back?\u201d<\/p>\n Caleb Ward, who along with his wife, Shelby Ward, runs Curious Refuge, an online platform pioneering the use of AI in filmmaking, said the best AI films they\u2019ve seen through their online school have come from \u201cpeople who have preexisting skillsets in storytelling. We feel these tools will become even more user-facing and easier to use so that people who are classically trained can take these tools, contextualize them, and learn how to supplement their work to get even more content out there.\u201d<\/p>\n <\/p>\n <\/p>\n Notably, just before the start of the 2024 Tech Retreat, <\/strong>OpenAI debuted its new text-to-video generator Sora. A demo of Sora onstage at HPA was impressive, with various environments portrayed \u2013 a woman walking down a rain-washed Tokyo street at night, a spotted Dalmatian dog leaning out a window on a street with richly saturated colored houses, a cat gently nuzzling its sleeping master in bed, and more. While the initial reaction was positive, with many HPA attendees expressing surprise at how realistic the generative AI looked, that was quickly followed by questions of caution, fear, and even trepidation that Sora (and others to follow) may supplant a century of traditional filmmaking. (Not coincidentally, the CEO and COO of OpenAI, Sam Altman and Brad Lightcap, gave demos of Sora to representatives of at least three Hollywood studios a month after HPA, to quash fears that was OpenAI\u2019s intent.)<\/p>\n There\u2019s no doubt Sora\u2019s text-to-video generation was better than anything generative AI has yet produced. But thankfully, the final two days of sessions at HPA \u2013 while still AI-heavy \u2013 included many filmmaking veterans intent on updating attendees on advances in capture and on-set technology that uphold Hollywood\u2019s human-driven production model, one not ready to be taken over by machine learning anytime soon.<\/p>\n That included Mark Harrison, CEO of DPP, an international network for media and technology Harrison founded that counts more than 500 companies as members \u2013 from global tech giants to broadcasters, and from media tech vendors to production companies. Harrison\u2019s annual role at HPA is to report back on the trends he sees at CES (Consumer Electronics Show), many of which are precursors to tech innovations in the professional space.<\/p>\n \u201cWith the lines between business and consumer technology blurred,\u201d Harrison announced during his PowerPoint presentation, \u201cCES is an opportunity to take stock of where we are in the global battle for a finite resource, which is the time, attention, and money of consumers. Where are we on that battlefield?\u201d Harrison turned to a slide representing the last 20 years of original content created in the U.S. alone, \u201cwhich,\u201d he pointed out, \u201cincreased 10-fold, year-over-year, creating an unsustainable bubble. This bubble burst before the SAG and WGA strikes. And when times are tough and revenues are down, what\u2019s the other thing that has to come down? Cost. So after years of seeing companies pitch innovation at CES that encourages us to spend money, we are now seeing innovation that, potentially, encourages us to save it.\u201d<\/p>\n <\/p>\n <\/p>\n Other Wednesday sessions included \u201cNew Developments in<\/strong> Open Source Software for the Motion Picture Industry,\u201d moderated by David Morin (Academy Software Foundation) and featuring professionals from Autodesk, AWS, Netflix, Frame.io and Foundry, and Eluvio\u2019s Michel Munson discussing Content Authenticity and Provenance: Protocol versus Workflow.<\/p>\n Morin opened with a recollection from his breakfast roundtable earlier that morning about workflow. \u201cI was speaking with a post supervisor who told me that when a shot is cut from a movie, you want them to stop working immediately because that saves tons of money. That sounds obvious, but it\u2019s often not the case. Because of the complexity of pipelines, people will keep working on a shot that\u2019s been cut from a movie for quite some time. So we have on this panel people who are close to the ground and working to improve these kinds of workflows, particularly trying to enable the interoperability aspect we\u2019ll be talking about today.\u201d<\/p>\n Munson, co-creator of the content fabric protocol, an open and decentralized Internet protocol for owner-controlled streaming, storage, and monetization of digital content at scale, is an Emmy-award-winning technologist who previously co-founded Aspera in 2004 and led the company as CEO until May 2017, including through acquisition by IBM in 2014. Munson, who co-invented the Aspera FASP\u2122 transport technology, told the room she wanted to \u201ctackle this topic for obvious reasons, given the theme this year of both the threat \u2013 and the opportunity \u2013 of large language models, generative AI, and their impact on the creative industry. I also want to tell you a little bit about the backstory of the community around that of the last year. I want to frame this with a certain point of view, and timeclock, which is end-to-end design that is the foundation and bedrock of how the internet was built.\u201d<\/p>\n Tony Guarino, Head of Paramount Global\u2019s Production and Studio Technology Group, talked about how his team at the studio is onboarding new and emerging technologies in the production space, including generative AI applications for content creation. \u201cRegarding Gen AI,\u201d Guarino announced, \u201c2023 was the beginning of a lot of discovery and some experimentation for us, and 2024 is shaping up to be heavy experimentation and preparing onboarding and tooling for the creative community. AI will have a major impact on our studio as well as all the other major Hollywood studios. I\u2019m here today to talk about responsible AI, from the studio perspective, and how it\u2019s impacting the content creation lifecycle. Used properly, gen AI can be a tool to assist creators by augmenting and enhancing their ideation. Also working in the background, for example, embedded in workflow software, it can speed up creators\u2019 ability to get their work done faster, and, ultimately, reach audiences faster as a result.\u201d<\/p>\n The final day at HPA began with Jim Helman, CTO of the<\/strong> nonprofit MovieLabs (a joint venture of five major Hollywood studios that has put forth a 2030 vision plan for media creation in the cloud), moderating a panel on enabling interoperable workflows. It included technology experts from Universal, The Walt Disney Studios, Adobe, and Autodesk. Following that discussion was a Local 600-sponsored panel titled \u201cStepping Up to Color Managed On-set HDR\/SDR Workflows.\u201d Moderated by ICG Technology Expert Michael Chambliss, the conversation featured Greg Smokler, VP\/GM of Cine Products at Videndum Creative Solutions, whose product teams include Teradek Cine, SmallHD, and Wooden Camera; Rory Gordon, Supervising Colorist for ArsenalFX, who has finished more than 130 episodes of television in HDR10 and Dolby Vision; and ICG Director of Photography Tommy Maddox-Upshaw, ASC, whose Hulu series Snowfall <\/i>won an ASC Award for Best Cinematography for a One-Hour TV Series, making Maddox-Upshaw the first black ASC member ever to win the honor.<\/p>\n<\/div>\n Chambliss got the discussion going by referencing an HDR projection demo from Barco from Monday night that, \u201cportaging an HDR image from a smaller envelope, like a Rec.709 container into a P3\/PQ HDR space, can reveal a host of issues,\u201d he explained. \u201cBut the forward march of intrepid engineers is making on-set HDR monitoring and color-managed HDR workflows an efficient way of preserving creative intent, as well as keeping that intent through the pipeline with both SDR and HDR dailies and on through postproduction. HDR has been around for a while now, so we\u2019ve assembled this panel of really smart folks to tell us why we can do something we couldn\u2019t do four years ago.\u201d<\/p>\n Smokler was first to reply, noting that \u201cwe\u2019ve been capturing in HDR for some time, and finishing in HDR for a decade-plus. But throughout this time, cinematographers never could see full [native] HDR during the capture phase. For us at SmallHD, we\u2019ve always been focused on making small, portable, very rugged monitors that are extremely useful for framing and analyzing exposure, ultra-bright with a rich software feature set, et cetera<\/i>. But we didn\u2019t tackle the \u2018shooting for the stars\u2019 goal of making displays that were reference-quality-ready and worthy of judging a final image on set. We\u2019ve had the capacity and expertise, but we didn\u2019t start until about five years ago. The goal was a 1000-nit HDR monitor that could go anywhere in the world filmmakers need to go, and would have all the requisite software tools.\u201d<\/p>\n Chambliss quizzed Smokler about any feedback he\u2019s gotten, \u201cabout the types of things that need to be built into the software since the monitor has become not just an extension of the viewfinder but much more of a high-end production tool.\u201d Smokler replied that for SmallHD, \u201cWe take a holistic approach to the technology, with custom-designed platforms that make the software user experience in harmony with all of the engineering that goes into the electronics. That enables us to have this ultra-rich operating system that creators and technicians can utilize and customize in an endless way on set. From a tools perspective and software, the things that allow cinematographers to get their lighting levels\/ratios done quickly. These are the same tools you might find in a color bay to assess and dial in things to be incredibly accurate. We want to bring those tools to the fingertips of a cinematographer to allow them to deploy them in the specific, bespoke way they want to for every production.\u201d<\/p>\n Maddox-Upshaw, who just finished a film that used an HDR\/SDR color-managed workflow, spoke about how that changed his process on set. \u201cThe impact with the HDR workflow starts in prep and doing a camera test, where my final colorist is involved as we structure the LUT,\u201d he began. \u201cSeeing how far the roll-off is, and the highlights, and how much more on the bottom end we can see. With the construction of the LUT, I have art department and wardrobe examples in the frame, and in the camera test, I can see the high and low points and the color shifts. Seeing HDR\/SDR simultaneously, and seeing that the HDR is more saturated, I know I can always come down off that HDR peak for the final product if needed. So to see all the HDR and SDR values on a set is a great advantage, but to have it pay off, it all starts in prep.\u201d<\/p>\n Chambliss added that \u201cwhat Tommy is referencing is a conversation we had just before this panel about the challenges cinematographers have when monitoring on set.\u201d \u201cThat\u2019s right,\u201d Maddox-Upshaw picked up. \u201cEvaluating footage on set, we\u2019re not always in the most pleasant viewing environments, in which case you have to fly by the instruments, the values of the monitor. The numbers and values are dictated by the final colorist in terms of the LUT construction, the gamma curve, and all that we established in prep. I\u2019m always metering those values from the waveform [or other features on the monitor] and aiming for the nit values of the high IREs in terms of skin values and highlight blowout points. If I\u2019m out on a bright sunny day, I\u2019m aiming for those value numbers that have been determined by the final colorist and myself for the final image to look a certain way. I can be flying the airplane by the instruments while [the colorist] is in the tower telling me to \u2018pull up,\u2019\u201d he laughs. \u201cIt comes down to the parameters you set up in prep with your color scientist to make it all work.\u201d<\/p>\n Colorist Gordon quickly added, \u201cWe live in a multiple-color-target world now. That\u2019s the big difference now from five years ago. One of the first shows at Arsenal that used on-set monitoring was Raising Dion<\/i>, and that was with [AJA Video Systems\u2019] FS-HDR. Colorfront\u2019s engine was in that rack as well so you had an SDR feed, an HDR feed, and a Rec.2020 container, which was required [by Netflix]. But now I think color management has gotten to the point where, in the LUT design, we can make choices that allow us to have easy parody with the LUTs we send to set. As Tommy was saying, if we know we\u2019re going to want to rescue detail from a specific situation on set, or we\u2019re pretty close to the walls and we don\u2019t want to have detail in this situation. The biggest difference now \u2013 from five years ago \u2013 is that we now live in a world where both an SDR and an HDR color target can coexist.\u201d<\/p>\n<\/div>\n <\/p>\n <\/p>\n![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/53570716273_64b1f1dddc_o.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/53569641142_5288c8b64e_o.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/53571015470_51a9ca86cf_o.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/Jeff-Rosica_AVID-Presentation_JPEG.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/Sora-Demo_JPEG.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/Schubin-Presentation_AI-Spongebob_JPEG.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/IMG_7486-copy.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/53571012880_d57aefd433_o.jpg\")
![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/53570685753_48feec2e19_o.jpg\")
<\/p>\n![\"\"](\"https:\/\/www.icgmagazine.com\/web\/wp-content\/uploads\/2024\/04\/ZEISS-Demo_Innovation-Zone_JPEG.jpg\")