Abstract
Accelerated by the COVID-19 pandemic, recent discourse has positioned VP as a radical approach capable of reimagining conventional filmmaking workflows through real-time technologies. However, such accounts often privilege high-profile case studies and promotional narratives, obscuring the practical constraints and realities, organisational demands, and workflow implications that shape VP filmmaking. This article examines VP through a workflow-centred perspective, focussing on how VP is planned, coordinated, and implemented across different production contexts, rather than treating it solely as a technological endeavour. Drawing on interviews with industry professionals working in China, the UK, and the USA, the study identifies a critical ‘pre-development’ phase in which stakeholders assess creative intentions, technical feasibility, and resource allocation prior to formal pre-production. The findings unveil that VP workflows are highly project-specific rather than standardised, and that early decision-making plays a decisive role in mitigating technical pitfalls, budget escalation, and scheduling delays. Ultimately, the article proposes a workflow framework that emphasises the central coordinating role of the Studio Manager and the importance of structured planning before formal production begins. In doing so, the study contributes a practice-based account of VP that foregrounds workflow organisation, coordination, and decision-making as critical factors in contemporary VP filmmaking.
Keywords
Introduction
The global film industry has seen interest pivoted in Virtual Production (VP) filmmaking, which has largely been driven by the 2020 global pandemic. Much of the recent discourse surrounding VP, however, has been shaped by marketing narratives and high-profile case studies, which tend to frame VP as a transformative solution capable of collapsing the traditional boundaries between pre-production, production, and post-production (Bickerton, 2022; Chanpun, 2023). As Felix Jorge asserts, ‘People call it Virtual Production, but in the future, it’s just going to be called filmmaking’ (Stolz et al., 2021: p. 4). While such claims point to the potential of VP, they often obscure the practical constraints that shape how, when, and whether VP can be effectively deployed.
Despite the excitement surrounding VP filmmaking, it remains a relative immature and unevenly standardised production approach, presenting significant challenges for both the film industry and academic institutions offering filmmaking-related courses. VP is not a neutral or universally applicable solution, and in practice it introduces new technical dependencies, workflows, and organisational demands. Without sufficient planning and coordination, VP can increase production risk rather than reduce it, leading to technical delays, budgetary pressure, and comprised creative outcomes (Kadner, 2019; Lam et al., 2025).
Additionally, the steep learning curves associated with acquiring new skills and processes – such as previs, techvis, and virtual asset development and integration – are often underestimated when compared to traditional production workflows. While these tools can enable earlier creative decision-making, they also require substantial investment in time, skills, and cross-departmental collaboration, which can act as a barrier to adoption for practitioners and educators alike (An 2022; Bédard, 2022; Swords and Willment, 2024; Zappe, 2025).
Moreover, the interdisciplinary nature of VP filmmaking means that professionals from different backgrounds have differing viewpoints on what VP entails, and how it should function within a production pipeline. These competing perspectives result in inconsistent definitions and fragmented workflows, suggesting multiple, and sometimes competing approaches to VP practice (Chanpun, 2023; Thacker, 2012). Establishing a streamlined or transferable workflow is therefore more complex than it may initially appear, particularly when projects differ widely in scale, budget, and available expertise.
This article investigates the workflows that underpin VP filmmaking, bringing together professionals from diverse disciplines at this intersection. Drawing on interviews with industry practitioners from China, the UK and the US – including producers, studios managers, gaffers, and directors of photography – the study examines how VP workflows are understood and implemented in practice. The thematic analysis of data suggests that effective workflows indicate an early ‘pre-development’ phase, in which stakeholders assess project requirements and critically evaluate the suitability of VP in relation to creative intent, budget, and technical infrastructure.
Ultimately, this research proposes a workflow framework that reflects the realities of VP practice rather than its promotional rhetoric. The framework emphasises the importance of detailed planning prior to formal pre-production and highlights the central coordinating role of the Studio Manager within VP environments. By making these processes explicit, the article aims to support both industry professionals and educators in articulating clearer, more realistic VP workflows, thereby helping to lower barriers to adoption while avoiding inflated expectations about what VP can deliver.
Literature review
Virtual production as a shift in production workflow
Conventional filmmaking relies heavily on the construction of physical sets, location-based shooting, and large crews, all of which present logistical, financial, and environmental challenges. VP has emerged in response to these pressures – particularly following the COVID-19 global pandemic – offering a hybrid production approach that integrates digital environments with live-action filmmaking. Rather than functioning solely as a technological innovation, VP fundamentally reorganises production workflows by enabling creative and technical decisions to be made earlier and iteratively across pre-production, production, and post-production (Chanpun, 2023; Bickerton, 2022).
While contemporary VP workflows draw upon a convergence of technologies – such as real-time game engines, camera tracking, motion capture, and dynamic lighting – the significance of VP lies less in any individual tool and more in the way these technologies are coordinated within real-time environments (Kavakli and Cremona, 2022; Swords and Willment, 2024). Historical antecedents such as rear and front projection provide useful context, but contemporary VP differs in its ability to support live interaction between performers, physical sets, virtual environments, and cinematography during principal photography (An, 2022). A defining key characteristic of contemporary VP practice is the use of LED wall technology.
VP LED walls are typically installed within volumes – sound stages that combine linked LED panels with physical props, set elements and computer – synchronised camera systems. Notably, Industrial Light & Magic (ILM) introduced the LED volume technique under the StageCraft brand during the production of Rogue One: A Star Wars Story (Edwards, 2016). This system enabled CGI scenery to be presented directly within the studio environment, allowing digital environments to function as part of the physical set. VP LED Walls were further developed in the Disney+ series The Mandalorian (2016–). ILM describes the production workflow as follows: Actors in The Mandalorian performed in an immersive and massive 20′ high by 270-degree semicircular LED video wall and ceiling with a 75′-diameter performance space, where the practical set pieces were combined with digital extensions on the screens. Digital 3D environments created by ILM played back interactively on the LED walls, edited in real-time during the shoot, which allowed for pixel-accurate tracking and perspective-correct 3D imagery rendered at high resolution via systems powered by NVIDIA GPUs. The environments were lit and rendered from the perspective of the camera to provide parallax in real-time, as if the camera were really capturing the physical environment with accurate interactive light on the actors and practical sets. (Anon, 2020)
Following the success of The Mandalorian, LED volume workflows became widely perceived as a new paradigm in global film production. Yet, while these technological advancements demonstrate clear practical benefits, scholarly understanding – and to some extent industrial – of how VP functions as a production workflow remains comparatively underdeveloped. Moreover, as Zappe (2025, 101) observes, while LED volumes are often positioned as the future of filmmaking, ‘finding ways to eliminate the technical drawbacks to increase creative freedom becomes of great importance in the industry’. Addressing these tensions forms one of the key motivations of the present study by delineating a VP workflow.
At this point, it may also be questioned how a unified workflow can be established given the differences in virtual production scale between major studio works such as The Batman (Reeves, 2022) and The Mandalorian. The latter filmed ‘in a 270-degree oval shaped studio with curved LED-walls and LED ceiling’, which contrasts with smaller studios that, in some instances, operate using a single flat wall (Zappe, 2025, 105).
Such variation can be observed in Lam et al.’s (2025) examination of the University of Nottingham Ningbo China’s (UNNC) facility – the first ARRI Virtual Production and Motion Capture Studio in Asia. Although modest in scale, this facility operates using comparable technical principles and software applications, which the article will ultimately examine. Crucially, regardless of VP stage size, Zappe (2025, 105) emphasises that ‘[n]o matter the size of the Studio, technical components involved fundamentally stay the same’. What differs, therefore, is not the underlying technology but how VP workflows are configured, scaled, and combined in practice.
LED wall, greenscreen and hybrids
Variations in workflow configuration became particularly evident when comparing LED wall, greenscreen, and hybrid VP approaches. The transition from traditional blue and green screen workflows to LED wall systems represents a key operational distinction within VP practice. As Swords and Willment (2024) argue, chroma key techniques require performers to imagine virtual environments during filming, whereas LED walls provide immediate visual context. As already noted, this distinction became particularly pronounced during the COVID-19 pandemic, when location shooting restrictions and social distancing measures compelled production companies to reassess conventional filmmaking practices. And although VP techniques had been evolving prior to this period – evident in productions such as Avatar (Cameron, 2009) – the pandemic marked a turning point in the scale of industrial adoption. LED wall systems provide especially valuable in enabling location-independent production, allowing crews to create visually complex environments within controlled studio settings when on-location filmmaking was not possible (Lam et al., 2025; Swords and Willment, 2024).
However, LED volumes, greenscreen, and hybrid approaches should be understood as distinct VP workflows rather than interchangeable solutions. Greenscreen-based virtual production typically retains a more traditional production structure, in which virtual environments are composited primarily during post-production. While previsualisation and digital assets may inform production decisions, lighting, performance, and cinematography are often executed without direct visual feedback of the final environment on set. As a result, greenscreen workflows tend to preserve a clearer separation between production and post-production phases (Kadner, 2019; Wall, 2022).
By contrast, the distinction between LED volume workflows and greenscreen-based approaches lies not only in visual presentation, but in the reordering of production phases. As Zappe (2025, 125) explains, shooting within an LED volume fundamentally changes the order of pre-production and post-production, as digital worlds and visual effects that would traditionally be created after principal photography must instead be developed in advance to be projected live on set. This inversion of workflow places increased emphasis on early asset development, technical testing, and interdepartmental coordination. In contrast, greenscreen workflows typically defer environment integration to later stages, relying more heavily on post-production processes (Kadner, 2019; Wall, 2022).
Hybrid VP workflows occupy a middle ground between these two approaches. They typically combine LED walls with greenscreen extensions or conventional compositing techniques as a pragmatic response to spatial, budgetary, or technical constraints. While hybrid approaches can offer flexibility, they also introduce additional workflow complexity, as productions must plan simultaneously for real-time rendering and post-production pipelines (Kadner, 2019).
This study focuses specifically on LED volume workflows, as these most clearly expose the organisation and procedural shifts that underpin contemporary VP practice. As Kadner (2019, 7) notes, shooting in an LED volume requires that digital environments and effects – normally created after principal photography – be prepared in advance for live projection on set. This reorganisation of labour foregrounds the importance of pre-production planning, technical testing, and cross-departmental communication, making LED volume workflows a particularly useful lens through which to examine VP as an industrial process rather than a purely technological one. One of the earliest and most consequential sites at which this reorganisation becomes visible is previsualisation, where creative, technical, and logistical decisions are increasingly consolidated prior to principal photography.
Previsualisation versus storyboarding versus pre-production
While previsualisation forms as a key pillar of VP, the practice of visualising scenes before production has deep historical roots. Storyboarding was pioneered at Walt Disney Studios in the 1930s for animation production and later became essential to live-action filmmaking, enabling scripts to be translated into frame-by-frame visual representations that established cinematographic composition and mise-en-scène in advance of shooting.
This approach shares similarities with the animatic methodologies used in advertising, where storyboards combined with voiceovers to test the visual messaging (Reynolds and Gengsler, 1991; Jiménez et al., 2020). Digital previsualisation evolved through experimental practices in late twentieth century cinema. Films such as Star Wars: The Return to the Jedi (Manquand, 1983) employed miniature sets for test shooting, while Francis Ford Coppola’s One from the Heart (Coppola, 1982) made use of an ‘electronic cinema’ approach to explore staging, lighting, and camera movement prior to production. Given this history, a question that may arise as to why previsualisation should be considered distinctive within the context of VP.
Contemporary VP-based previsualisation extends these historical practices by embedding them directly within real-time production environments. Rather than function solely as a planning tool, previsusalisation increasingly informs on-set decision-making, shifting creative labour earlier in the production process and requiring closer collaboration between departments traditionally separated by linear workflows (Bevan and Mackness, 2025; Kadner, 2019).
Central to this change is the adoption of real-time game engines, most prominently Unreal Engine. First introduced by Epic Games in the late 1990s as part a first-person shooter (Unreal), Unreal Engine was conceived as a modular and extensible system capable of evolving across success technological generations (Coleman, 2025). Since then, it has progressively advanced beyond its origins in game development, becoming a key platform through which practices from gaming, animation, and filmmaking increasingly meet at the confluence. A particularly significant moment for virtual production was the release of Unreal Engine 4 (UE4) in 2014, which broadened access through free educational licencing and the introduction of the Unreal Marketplace, thereby lowering barriers for digital asset creation. At the same time, the engine’s cinematic toolset matured, enabling the use of virtual cameras, real-time shot sequencing, and in-engine capture and rendering workflows that more closely resemble established film production practices (Coleman, 2025).
Within VP contexts, these developments have important workflow implications. Game engines enable previsualisation to function less as a representational exercise and more as a simulation of production conditions, allowing camera movement, lighting behaviour, and virtual set design to be tested interactively. Bevan and Mackness (2025) note, this continuity support more extensive test shooting and improved coordination across departments. By contrast, in traditional filmmaking, previsualisation typically remains limited to storyboards or animatics, which do not directly translate into the footage captured during principal photography. With the release of Unreal Engine 5 (UE5), these elements broadened, particularly in relation to cinematic sequencing. As such, the engine functions less as a tool for visual output and more as an infrastructural component within VP workflows, enabling earlier feasibility assessment and front-loaded decision-making across pre-production phases (Coleman, 2025).
As Noah Kadner argues in The Virtual Production Field guide, real-time engines such as Unreal Engine can eliminate many of the bottlenecks associated with budgeting, scheduling and development time, precisely because the assets and materials created during previsualisation can be carried through into the final shoot (Kadner, 2019). In this sense, VP-based previsualisation offers a level of accuracy, continuity, and potential cost efficiency that is difficult to achieve within traditional filmmaking workflows.
However, the adoption of such workflow also introduces new challenges. The integration of Unreal Engine brings with it a set of specialised technical skills that are not widely embedded within traditional production roles (Swords and Willment, 2024). Given the engine’s historical development within the game industry, many Unreal Engine operators and technicians entering VP environments come from game development backgrounds (Ibid). While technically proficient, these practitioners may not always be fluent in the established language and grammar of filmmaking, such as exposure expressed in stops, lens-based approaches to camera movement, colour, or conventional lighting strategies grounded in physical causality.
Conversely, for filmmakers trained within conventional production paradigms, VP-based previsualisation can present its own conceptual and practical challenges. Lighting logic, in particularly, operates differently within LED volume environments, where the LED wall functions simultaneously as background, light source, and reflective surface. Established approaches – such as motivating sunlight through a window so that it falls naturally onto a subject – must be rethought when that window is virtual and its luminance, colour, and spatial behaviour are governed by a real-time engine rather than physical light. Without sufficient familiarity with these constraints during previsualisation, creative intentions may be limited rather than strengthened (Rosnell, 2024).
As a result, while the VP previsualisation has the potential to accelerate decision-making and reduce uncertainty, it can also have the opposite effect if skills gap, mismatched disciplinary languages, or incomplete conceptual alignment are not addressed early in the process. As noted earlier in this literature review, VP requires the coordination of professionals from disciplines not traditionally associated with on-set filmmaking (Swords and Willment, 2024). Consequently, challenges emerge not only in terms of technical integration, but also in relation to communication, role definition, and workflow alignments, as practitioners of film production, visual effects, and game development converge within a shared production environment.
These tensions highlight the importance of understanding previsualisation in VP not as merely as a technical tool, but as a critical cite of negotiation between creative intention, technical constraint, and disciplinary knowledge. Perhaps, they also help to explain why virtual production remains conceptually contested, as the instability of roles, responsibilities, and workflows complicates efforts to establish a stable and universally applicable definition of VP practice (Bickerton, 2022; Chanpun, 2023; Zappe, 2025).
Conceptualising virtual production: Evolving definitions
Virtual production remains conceptually contested within academic discourse. While VP has attracted increasingly scholarly attention, it is often characterised as a nascent field of enquiry due to the absence of a stable or universally accepted definition (Bodini et al., 2024; Chanpun, 2023; Thacker, 2012). Sebastian Sylwan, the former chief technology officer at Weta Digital and member of the VES Founding Board notes: ‘There’s no checklist of things you can go through and say, if you have all of these, this is virtual production’. Further, he explains that the industry is still in a process of learning and exploration. Consequently, ‘[…] it’s hard to come up with a definition that serves all the possible aspects of virtual production […]’ (quoted by Thacker, 2012).
This conceptual uncertainty is echoed by Zappe (2025, 125), who emphasises that VP reshapes not only technical workflows but also communication and responsibility structures on set. As Zappe explains: The Brainbar is typically a central point of communication since the content on the wall ties together the responsibilities of Director, DP, Lighting department and the Production Design. In interest of the DP, color, contrast, and luminance of the scene of the wall, hence also the ambient light casted by the wall, can be adjusted and the lighting can be set up accordingly or vice versa.
Such observations illustrate how VP workflows are organised less around discrete departments and more around shared points of technical and creative convergence. For this reason, much of the existing scholarship approaches virtual production through organisational or process-oriented definitions, rather than treating it solely as a technological development.
The first organisational definition of VP emerged from a joint initiative assembled by the American Society of Cinematographers, the Art Directors Guild, the Visual Effects Society, the Previsualization Society, and the Producers Guild of America. Published in 2012, this definition describes virtual production as ‘collaborative and interactive digital filmmaking process which begins with virtual design and digital asset development and continues in an interactive, non-linear process throughout the production’ (Sergeant et al., 2014, 444). This organisational definition resonates with Alex McDowell’s notion of creating a ‘virtual space’ in filmmaking integrating different types of technologies (Beck, 2014, 74), and Dunlop’s (2014, 304) conception of VP as ‘the application of real-time techniques within the scope of filmmaking’. One markedly explicit definition of VP can be found in Helzle et al. (2015, 347), who describes it as a concept combining ‘key aspects of media production in a real-time, or close to real-time, environment where creative decisions can be taken in direct consultation with other members of the team’.
Taken together, these definitions emphasise collaboration, real-time processes, and the reorganisation of production workflows, rather than on any single technology or toolset. As such, much of the literature positions VP as a procedural and organisation shift within contemporary filmmaking practice.
Benefits, challenges and identified gaps
Building on these conceptualisations, existing scholarship has sought to evaluate the implications of VP filmmaking in practice. Despite its conceptual elusiveness, scholars largely agree on the significance of VP filmmaking (Kavakli and Cremona, 2022; Mitchell et al., 2025; Swords and Willment, 2024). Several key advantages emerge from the literature. Firstly, VP can be cost effective, reducing the necessity for elaborate physical sets, locations, and props, resulting in substantial savings. Secondly, it offers flexibility in terms of adjustments to virtual environments and scenes, affording filmmakers greater control over their work. Thirdly, it enables real-time visualisation, allowing adaptation and adjustment of scenes during shooting, which can significantly reduce post-production (Entertainment Technologists Inc, 2023). Finally, virtual production does not just enhance workflow efficiency through more organised production processes, but it also provides safer working conditions than traditional filmmaking, which sometimes requires crew members to operate in risky environments (Reinhard et al., 2012, 707).
However, beyond these established benefits, VP also presents pressing challenges in terms of workflow and the extent to which the technology can be fully exploited, as identified in this literature review. Moreover, the industry faces a skills shortage of professionals proficient with Unreal Engine and related real-time systems needed to manipulate virtual environments, virtual cameras, and virtual lighting systems. This is observed by Swords and Willment (2024, 12): ‘[T]he novelty of virtual production, both in combination of technologies and workflows, has created knowledge, skills and labour gaps in the production network’. Indeed, there are challenges with creating virtual environments, particularly pertaining to the considerable time and costs required to develop photorealistic digital assets. In addition, VP cannot be used to achieve all forms of filmmaking, contrary to some of its early marketing claims, and may therefore not be suitable for all types of productions.
For instance, Zappe (2025, 126) discusses limitations associated with LED walls, noting that they typically operate at a maximum of 60 frames per second, ‘which makes shooting super slow-motion with 100 fps or higher impossible without having flicker artifacts in the wall’. As a result, while traditional filmmaking can accommodate a degree of spontaneity, VP workflows require a more deliberate and controlled approach. Extensive test shooting, alongside advanced conceptualisation of shots during pre-production, becomes essential in order to avoid unforeseen technical and creative complications during principal photography. This is supported by Swords and Willment (2024), who report insights from a Virtual Art Department supervisor at a global VFX company, revealing that while a scene may take only a couple of days to film, it can require several days of preparation and testing in advance to ensure that everything runs smoothly on the day.
Furthermore, as Rosnell (2024, 4) outlines in his discussion of the ARRI Virtual Production Colour Management Pipeline, ‘most LED-screens reproduce colors with RGB diodes, [and] the spectrum is riddled with missing hues’. In short, the way a camera interprets RGB colours is unique to a given camera system and manufacturer. Without a clear colour pipeline – and without LED calibration – the combination of real-world foreground elements and virtual backgrounds captured by a camera sensor can result in visible colour mismatches between the two (Ibid, 4).
This is further supported by Gudemann et al. (2022, 109), who note that ‘[d]igital cameras employ a color correction stage within the image processing pipeline so that the color of common objects under common illumination spectra can be reproduced in an accurate and aesthetically pleasing manner’. LED screens, however, are not designed to function as illumination sources for physical objects. Their spectral distribution differs significantly from the standard illuminants for which camera colour correction is optimised, containing substantial gaps – particularly in the cyan and yellow wavelength ranges – where relative intensity is low. As a result, without careful colour management, LED-driven illumination can undermine colour fidelity and visual continuity within VP environments.
In short, this literature review reveals two key gaps. First, while there is general agreement that VP enhances workflow efficiency, there is little empirical evidence examining how these workflows function across different production contexts and scales. As Barnett et al., 2025, 1) note, ‘[w]e don’t yet know what ways of thinking and doing will stick (…)’. Second, despite rapid industrial uptake – evidenced by surveys indicating that 40 percent of UK and US executives were using virtual production tools in 2022 (Swords and Willment 2024) – there remains insufficient scholarly attention to how VP workflows are structured, coordinated, and evaluated in practice. Given that the ‘VP industry is predicted to have an annual growth rate of 18.2% globally and the market is predicted to be worth $6.78 billion by 2023’ (Willment et al., 2025, 7), these gaps highlight the need for workflow-focused research which can support widespread industry adoption. In summary, these gaps point to the need for research that not only considers technological considerations, but also examines VP as a managed production process, attentive to coordination, decision-making, and workflow organisation. It is within this context that the article positions its contribution.
Conceptual framework: Workflow management support system
To address the first gap regarding VP workflow practices, this study draws upon the conceptual framework of workflow management support system, which provides a structured means of analysing how complex, interdependent tasks are planned, coordinated, and executed across multiple stakeholders. This framework is particularly suited to VP filmmaking, where creative, technical, and organisational decisions are distributed across disciplines and must be aligned early in the production process to mitigate risk and inefficiency.
Over the years, organisations have introduced workflow management systems to better organise their operations (Van der Aalst et al., 2003). Organisations with large human-resources and administrative structures use workflow management systems to define and monitor diverse business activities, streamline work using digital tools, and reduce processing times (Van der Aalst and Van Hee, 2002). Thus, ‘a workflow management support system is a system that defines, manages and executes workflows through the execution of software whose order of execution is driven by a computer representation of the business process’ (Joosten, 1995: p. 4).
The use of technology plays as crucial to organising, managing, evaluating and analysing business processes with an ultimate goal of procedure improvement. These improvements can be short-term (e.g. redelegating tasks to balance workload) or long-term (e.g. redefining workflow components to avoid future blockages). Regardless of organisation type and industry, workflow management system is important as they ensure: • Better organisation of work by identifying who does what, when, how, and in which order; • Efficiency by evaluating roadblocks and removing unnecessary steps; • Transparency through constant monitoring and identification of issues; • Adaptability to meet market and client demands; • Quality control through regulatory compliance; • Cost savings by reducing errors and conserving resources (Georgakopoulos et al., 1995; Van der Aalst et al., 2000).
Effective workflow management systems require four key components: (i) input (raw materials, information, or resources), (ii) transformation (activities converting inputs into outputs), (iii) output (achieved tangible and intangible goals), and (iv) stakeholders (individuals or groups who have an interest or influence in the workflow).
Workflow management system can take various forms – manual, semi-automated, or fully automated workflow (Van der Aalst et al., 2003). Organisations can implement several types of workflows based on their specific needs: 1. 2. 3. 4. 5. 6. 7.
Three key operational components support these workflow systems: interface processors (connecting applications and organising data), event managers (which maintain task lists and deadlines), and workflow managers (facilitating and coordinating across workflows) (Ouyang et al., 2015).
Taken together, these components illustrate how workflow systems function not merely as technical infrastructures, but as mechanisms for organising labour, coordinating decision-making, and sequencing processes across complex production environments. Therefore, as a conceptual framework, workflow management systems (WMS) have been widely used in organisational and management studies to analyse how complex processes are structured, coordinated, and optimised. In the context of this study, WMS provides a useful lens through which to examine VP filmmaking, where workflows are increasingly non-linear, interdisciplinary, and technically interdependent. Rather than treating VP solely as a technological development, this framework allows production to be understood as a managed process shaped by decision-making, coordination, and sequencing across multiple departments.
This perspective is particularly relevant when considering the practical conditions under which contemporary film production operates. In the fast-paced world of filmmaking, production efficiency and organisational clarity are critical, particularly when working within constrained schedules and budgets. VP workflows further amplify these demands, as creative, technical, and logistical decisions are often required much earlier in the production process than in traditional filmmaking. From pre-production to post-production, each phase of a film relies on structured workflows that connect creative teams, technical departments, and materials resources, ensuring continuity across the production pipeline (Hogg, 2021).
A conventional filmmaking workflow typically follows a relatively linear progression, moving from ideation and script development into shot design, principal photography, and finally post-production and delivery. As An (2022) outlines, filmmakers begin by establishing narrative intent and visual composition and subsequently transform these into moving images. However, as highlighted in the literature review, VP disrupts this linearity by collapsing and reordering stages of production, particularly through practices such as previsualisation, real-time rendering, and early asset development.
This disruption reflects broader changes within contemporary film production, which has increasingly shifted from linear to non-linear workflow structures, combining traditional sequential stages with parallel workflows to manage uncertainty, compressed timelines, and technical complexity (Priadko and Sirenko, 2021). VP filmmaking both reflects and accelerates this shift, as it integrates digital and physical environments and requires continuous coordination between departments throughout the production lifecycle (Bickerton, 2022; Chanpun, 2023).
When applied to VP, the workflow management system framework offers a new analytical perspective, as it foregrounds the coordination of people, technologies, and processes – the three dimensions repeatedly identified in the literature review as central to effective VP implementation. Rather than focussing on individual tools or software platforms, the framework enables analysis of how workflows are organised, scaled, and managed in practice, including how responsibilities are distributed and how decisions are made under real-time production conditions.
Accordingly, this study adopts a workflow management system for several reasons. First, the workflow management theory explicitly addresses the coordination of people, technology, and processes – the three key dimensions of virtual production. Second, the framework’s differentiation between workflow types (sequential, case-based, project-based, parallel, collaborative, etc.) provides analytical categories for investigating how traditional film production processes are being remodelled. Third, the framework’s emphasis on key components (inputs, transformation, outputs, and stakeholders) aligns with our research objective to identify critical roles and dependencies in the VP ecosystem.
For these reasons, the workflow management system framework is well suited to the aims of this study, which seeks to understand how VP workflows operate across diverse production contexts in China, the UK and the USA. By applying this framework, the study systematically examines how virtual production reshapes established filmmaking workflows, offering insights into coordination, decision-making and organisational practice that are relevant to both industry professionals and educators.
Methodology
To achieve our research aim of identifying the innovations and disruptions brought by VP technologies in the traditional filmmaking workflow, this study draws on interviews with 10 virtual production professionals from China, the UK, and the USA. While the number of countries represented may prompt questions regarding the use of the term global, it is important to clarify the international scope and professional reach of the interviewees.
The study was conducted with the support of ARRI, a global leader in film technology, which has been directly involved in the development of major infrastructures worldwide, including the establishment of Tencent’s Virtual Production Studio in Shenzhen, Mainland China (ARRI, 2022) and participation in large-scale international productions such as 1899 (2022) for Netflix (Mcknight, 2023). ARRI also operates dedicated virtual production facilities in Munich, London and New York, reinforcing its transnational presence.
Through this collaboration, the authors were able to engage with industry professionals whose work routinely spans multiple countries and production contexts. Although some interviewees are currently based in China, their professional experience, collaborations, and project histories extend across international markets, enabling them to offer insights that are not confined to a single national context. As such, the perspectives gathered reflect globally distributed VP practices, rather than regionally isolated ones.
The interviewees occupy a range of senior technical, creative, and managerial roles within the international VP ecosystem, including: • Virtual Production technology consultant and technical science engineer, specialising in VP solutions, based in Mainland China; • Studio Manager of a major Virtual Production facility in the United States; • Virtual Production Colour Management Specialist, based in the United Kingdom; • Virtual Production solution architect, trained in civil engineering with experience consulting in Europe and currently residing in Mainland China; • Head of Sales for Virtual Production solutions in Mainland China; • Gaffer and Unreal Engine operator, specialising in Virtual Production, based in Mainland China; • Virtual Production Unreal Engine Operator and technician working at a major VP studio in the USA; • Senior technical specialist at a global virtual production solutions company, with a focus on camera technology, based in Mainland China; • Established film director with extensive experience working in Virtual Production in Mainland China; • Camera operator with background in camera stabilisation systems, experienced in virtual production filmmaking, currently based in South Korea; • Former production professional at Bodi Film and Television company, with responsibilities spanning film production, film culture, urban tourism, and urban co-production initiatives.
Interviews took place between August 2024 and April 2025. Interviews followed a semi-structured format, and VP professionals were asked the key questions below (though not limited to these), aimed at providing a general understanding of the sub-questions that were developed for the semi-structured interviews for data collection: • How do professionals (i.e. producers) describe their use of VP and the changes it has brought to their practice? • How can professionals integrate VP system with legacy systems? • What is the nature of established workflow systems in current production environments? • Are there any automated workflow systems being used by specific industry stakeholders? If yes, how do professionals integrate it into their practice?
Interviewees were also given the opportunity to add anything they thought was important to for the interviewers to know. Interviews were carried out by two of the authors of this paper in English and in Chinese and then transcribed and, when needed, translated into English.
Interview content was coded using a two-step process. Open coding was adopted to identify concepts related to VP workflow, and its challenges. A second round of coding was conducted based on these concepts and categories formed using axial coding. Interview coding resulted in the following themes: Processes, Technical Features, (Hardware/Software/Digital Assets/etc…), Skill-Sets/Roles, Challenges, and Opportunities.
Kappa and frequencies of theme identified by coders (N = 10).
The average Kappa is 0.408 indicating a fair agreement between the two coders.
Once the coding was completed, a mind map was created (Figure 1) to help develop a logical plan for the analysis.
Coding mind map.
Institutional ethics approval was obtained to conduct the research. Interviewees will be referred to according to assigned numbers to maintain anonymity.
Analysis
Overview
This section will address key interconnected areas: the variable nature of Virtual Production workflows across production contexts and different regions; the necessary preparation phases that occur even before pre-production begins; a detailed framework concerning pre-production planning that can help ensure a VP project’s success. Rather than unveiling a prescriptive model, the analysis works to offer insights and questions to guide practitioners in determining their approach based on their specific production needs.
Ironically, the research initially set out to define a standardised framework for VP implementation, but interviews with industry stakeholders and the authors’ own experience with a high-end virtual production facility revealed that the marketed promise and practical realities of VP filmmaking are far more complicated than anticipated. As will be demonstrated through our interviews with leading virtual production practitioners globally, there is clear consensus that no unified workflow exists – and for good reason. The goals, budget constraints, regional resources, and creative intentions of each project necessitate bespoke approaches.
Current state of virtual production workflows: A global perspective
Each participant, whether explicitly or implicitly, echoed the sentiment that different solutions are dependent upon the specific project. This was supported by Interviewee 3, who commented that ‘A has a workflow for A, B has a workflow for B, C has a workflow for C. But then anyway, they all come together in pre-production and then there’s a production’ (Interview 3). This is corroborated by what we observe as the most mature application of Virtual Production filmmaking from a studio manager based in New York City, USA, who maintained that ‘[i]t’s almost always different project to project. Of course, there’s consistency in some of the steps or workflow taken, but it’s largely driven by project goals, etcetera’ (Interview 2). From a workflow management perspective, this suggests that VP filmmaking operates as a project-based workflow system, in which processes are configured around case-specific objectives through collaborative design across pre-production and production phases (Ouyang et al., 2015).
Furthermore, what Interviewee 3 is alluding to are pipelines; that is, different roles and production elements operate through role-specific within virtual production. Thus, while a production designer or Unreal Engine operator may follow distinct pipelines, particularly during the planning stage, their workflows ultimately converge prior to shooting. This point of convergence can be understood as coordinated workflow integration phase, in which parallel task streams are aligned to enable execution (Jablonski and Bussler, 1996). This observation also corroborates Zappe’s (2025, 125) argument that Virtual Production ‘changes the workflow on set, the communication and the responsibilities of the departments involved’.
However, this convergence is not always underpinned by fully stabilised or mature workflows. Interviewee 3 noted that it would be valuable to speak with practitioners working in high-end studio environments, as such contexts are more likely to have established and advanced VP pipelines, including efficiencies and shortcuts developed through sustained practice – forms of institutional knowledge and infrastructural maturity that smaller or emerging teams may not yet possess. This observation resonates with Lam et al.’s (2025) findings, which reflects on the challenges encountered when producing a VP short trailer within a small-scale studio context at the University of Nottingham Ningbo China (UNNC). In workflow management perspective, this reflects uneven process maturity, where optimisation practices accumulate over time but remain unevenly distributed across organisations (Van der Aalst et al., 2003).
This unevenness reinforces the view that virtual production remains an emergent practice, as already highlighted in the literature review, and suggests that practitioners operating outside of high-end industry studio environments may experience forms of knowledge inequality, often shaped by cost constraints and limited access to established infrastructures (a point returned to later). This interpretation aligns with Barnett et al.’s (2025, 1) argument that Virtual Production practice has yet to stabilise into a clear set of industrial norms, as the field remains in an experimental phase.
This lack of standardisations confirms Sebastian Sylwan assertion that there is no definitive checklist by which one can declare a production to be ‘virtual production’. From a workflow systems perspective, this absence of standardisation reinforces the view of VP as a case-driven and adaptive workflow environment rather than a codified procedural model (Ouyang et al., 2015). Taken together, these perspectives suggest that the uneven maturity of VP workflows is not simply a matter of scale, but it reflects broader condition of emergent practice within the industry.
Linking this back to Zappe’s (2025) discussion of shifting communication workflows, one explanation for this instability lies in the increased emphasis placed on pre-production within VP contexts. As Swords and Willment (2024, 6) observe, Virtual Production has significantly expanded the scope and concentration of pre-production labour, with a greater number of practitioners involved earlier in the process. This redistribution of labour reconfigures workflow ownership and decision-making authority, moving critical transformation activities upstream within the production lifecycle. As Lam et al. (2025) confirmed based upon their own practical experience, VP differs significantly compared to traditional filmmaking due to the amount of time allocated and spent in pre-production.
In practical terms, this structural shift places greater pressure on early alignment around creative and technical objectives. What was clear though, as a starting point according to Interviewee 6, is that there should be an established and clear goal that the ‘director and the producer agree on what they might want to see in the end’. This importance aligns with many of the participants who, as Interviewee 2 notes, recognise that ‘when you bring 3D into the conversation and Unreal, budgets can balloon very quickly […] I feel like a lot of that sort of what the workflows and the process that’s going to come as far as the technical and creative often is budget driven from the beginning’. Interviewee 5 further alludes to the challenges of not having a clear goal and the importance of working backwards to understand if it is viable or whether certain modifications need to be made in terms of director and producer expectations. They note that a huge LED screen wall may facilitate more adventurous projects, but it requires far more manpower (and thus budget) in terms of technical competency to operate Unreal Engine in particular (Interviewee 5). This may explain why Executive Producer Connie Kennedy of Profile Studios, advises that producers must ‘do a lot of listening’ when engaging with studio managers and technical VP specialists, in order to assess whether creative ambitions are technically and financially feasible (Kadner, 2019, 26). In this sense, the importance of clearly defining project goals at an early stage becomes evident, as prompt alignment can help to circumvent technical, logistical, and budgetary pitfalls later in the production process.
In workflow management terms, this reflects a case-based workflow logic, where goals are iteratively revised and resources reallocated as constraints become visible (Van der Aalst et al., 2003). Unlike traditional workflows that follow fixed sequence, VP workflow necessitates having a flexible approach to execute tasks especially when there is a lack of clarity for the desired goals. A case-based workflow is therefore effective in managing uncertainty by allowing manpower, time, and technical resources to be adjusted in response to evolving requirements (Van der Aalst et al., 2003). This flexibility is particularly relevant within Virtual Production contexts, where workflows are highly dependent on specialist knowledge and uneven skills distribution (Lam et al., 2025).
In this regard, the relevance of a case-based workflow is corroborated by Willment et al. (2025, 7), who note that a challenge in VP filmmaking arises from skills shortages. Where appropriately skilled personnel are unavailable or unevenly distributed across a production, this can directly affect a project’s ability to maintain schedule discipline, with consequent implications for budgets and the achievement of intended project outcomes. These constraints do not only shape how a project is executed, but also what can realistically be attempted in the first place (Lam et al., 2025).
From this perspective, the earlier definition of creative goals between director and producer in particular becomes critical. As Interviewee 6 remarked, you cannot do everything in virtual production; some projects will not be appropriate, despite virtual production often being branded as a creatively unlimited solution. They go so far as to add that VP has hard limitations when one tries to recreate reality outside of science fiction or fantasy. This is corroborated by Interviewee 7 who posits that ‘if you were trying to reproduce an environment that was entirely realistic, the question is like, why don’t you just shoot it in reality?’ In part, this observation is particularly revealing when considered alongside the literature review. Drawing upon Zappe (2025, 105), it was noted that, regardless of LED wall size ‘the technical components involved fundamental stay the same’, implying that knowledge and best practices developed within VP filmmaking can, in principle, be transferred across projects of difference scales, from high-end studio productions to educational or experimental contexts. However, Lam et al.’s (2025) account of producing a VP trailer film within a small-scale studio complicates this assumption.
Their study demonstrates that while software infrastructures and system configurations may remain consistent, the practical and aesthetic demands of realism become more challenging in constrained environments (see Figure 2). As they observe, ‘[t]he use of virtual production in small areas presents challenges that differ significantly from those encountered in large-scale environments’ (Ibid, 4). In larger LED volumes, background elements positioned at greater distances from the camera can tolerate lower texture resolution without comprising perceived realism. In contrast, within small-scale VP settings, even seemingly simple objects – such as a bench – require substantially higher levels of geometric and textural details to maintain visual credibility (see Figure 3). Example of a small-scale virtual production stage located at the University of Nottingham Ningbo China. Source: Lam et al. (2025, 2). Example of a large-scale virtual production stage featuring a landscape as the background. Source: Lam et al. (2025, 5).
In this sense, while the underlying technical systems may not differ between large and small VP stages significantly, the production implications do. Technical camera and lighting placement, and digital asset fidelity become more critical and more costly (potentially) in smaller environments, where proximity to the LED wall exposes limitations more readily. Consequently, small-scale VP productions may paradoxically require greater investment in asset quality to achieve realism, challenging the assumption that reduced scale necessarily equates to reduced technical or financial demand (Lam et al., 2025).
From a workflow management perspective, this highlights the importance of early feasibility assessment as a decision-making mechanism within VP workflows, particularly for small-scale studio environments. As Interviewee 4 alludes, as part of the initial preparation for VP, it is important to understand in what ways virtual production can help. In fact, Interviewee 5 argued that: Virtual production studios are not being purposed to typically shoot an entire film there. It's there to first select scenes that will cut time. Absolutely. And that's the best way to approach it from a production perspective. Unless maybe you have a short film and it's in one location that works nicely.
The point here is twofold: explicitly, it has become evident that virtual production cannot offer solutions to every filmmaking scenario and there are inherent technical limitations, as already highlighted. Implicitly, this suggests that VP functions most effectively when deployed selectively within a broader production workflow, rather than as an end-to-end replacement for traditional production methods.
This also aligns with Zappe’s (2025) observation that LED wall systems impose constraints that may not be immediately apparent, such as maximum frame rates that typically operate at around 60 frames per second. As a result, scenes requiring high-speed action combined with slow-motion cinematography – particularly when using lower shutter angles – may not be artistically or technically feasible within a VP environment, even if such an approach initially appears convenient. These constraints act as hard boundaries within the workflow system, shaping what tasks can be assigned to VP and which must remain within conventional production pipelines.
Still, from the workflow management perspective, one possible solution to cope with the limitations of VP is to apply a case-based workflow, which allows for greater flexibility and adaptability to unstructured tasks and to deal with unpredictable situations (Chen et al., 2025). Rather than enforcing a rigid sequence of activities, a case workflow enables stakeholders to respond to technical constraints, creative revisions, and emerging risks.
The case-based workflow can be challenging despite offering solutions to uncertainties of virtual production processes. Key stakeholders, such as studio managers, directors, DoPs, VP supervisors etc., as workflow leaders, need a sound level of judgement and decision-making to determine the best ways to apply VP techniques. This places increased responsibility on early-stage coordination and decision ownership, reinforcing the centrality of pre-production within VP workflow system.
Yet, to the original point, understanding the goal is fundamental from the outset, and from then onwards pre-production is fundamental. Before exploring insights concerning pre-production in detail, it is therefore necessary to understand how key stakeholders are supported in defining their objectives, constraints, and decision thresholds when pursuing virtual production.
The planning paradox: Preparing to prepare in virtual production
It was clear from the interviews that substantial groundwork must be completed prior to pre-production formally unfolding, a finding that aligns with concurrent scholarship (see Barnett et al., 2025; Bevan and Mackness 2025; Bevin et al., 2025; Kadner, 2019; Lam et al., 2025; Mitchell et al., 2025; Silva Jasaui et al., 2024; Swords and Willment, 2024; Willment et al., 2025). This begins with understanding the goal but also navigating and understanding if it is achievable, how it can be accomplished, defining the creative intention, along with budgets, manpower, et cetera – all of which is fundamental and necessary.
From a workflow management perspective, this stage functions as a pre-workflow configuration phase, in which objectives, constraints, and dependencies are defined before formal task sequencing begins. This creates somewhat of a paradox, something that takes place under the guidance of a studio manager (or stakeholder) of a VP stage (Interviewee 2). Interviewee 6 provided a detailed account, which can be summarised as follows, offering guidance that those both experience and new to virtual production can employ when speaking with key stakeholders at a studio: • The studio team facilitates on-site demonstrations using the LED wall, as experiential understanding proves more effective than verbal explanations alone. • Clients receive a structured tour including virtual scouting, previsualisation examples, and technical demonstrations of studio capabilities. • Discussions focus on specific technical needs – camera tracking systems, Unreal Engine employment, LED wall configuration, and lighting requirements – based on creative objectives. • The manager establishes parameters for digital assets, including file formats, camera image profiles, frame rate, shutter angle/speed, lens choice, T or F – stops, colour management, quality specifications, and Unreal Engine integration requirements. • Required plugins and coordination between production teams and virtual production specialists are considered and documented. • Using the accumulated technical and creative specifications, the studio manager develops a project-specific workflow designed to achieve (typically) the director’s and cinematographer’s intended visual outcomes.
Taken together, these steps illustrate how early VP planning operates as a workflow design activity, in which inputs, responsibilities, and technical interfaces are defined before execution begins. Basing an initial consultation on the above is important, particularly for those with limited experience.
One of the complexities of virtual production filmmaking is that it converges the real and virtual. There exists a physical camera and lighting system, along with a virtual camera and lighting system, so if a director or cinematographer/gaffer wishes to use a certain camera system to achieve a desired look, it may have implications on costing and feasibility given that studios are structured in a way where all physical and virtual systems must communicate with one another. As one studio outfit boasts, ‘since the entire setup is from ARRI, everything is compatible, and you don’t have to worry about any compatibility issues’. In addition, as interviewee 7 notes: Lighting is very different in virtual production versus reality. The physicality of having a volume in the background means that you can’t place lights in that position because one is the physical element, you can’t get that light there. There’s also the consideration that you also can’t [have] hard light from a certain direction because hard light will spill onto the screen, the volume and it’ll ruin the shot.
The above findings corroborate Interviewee 3’s observation that cinematographers must adapt to virtual production technology but also to new shooting techniques, as VP ‘significantly [affects] how they are shooting things’. This reflects a transformation stage within the workflow, in which established cinematographic practices are reconfigured in response to technical constraints.
This aligns with scholarship by Gudemann et al. (2022), Rosnell (2024) and Zappe (2025), all of whom identify challenges relating to light and colour in virtual production filmmaking. Specifically, as outlined in the literature review, LED walls reproduce light using RGB diodes, resulting in spectral gaps and missing hues, which can lead to mismatches in colour reproduction and overall visual consistency (Ibid).
In discussion with Interviewee 3, it emerged that ARRI, as a major film technology manufacturer and virtual production specialist, has developed dedicated plugins and colour management workflows designed to address these issues within its VP systems. However, from a workflow management standpoint, these solutions represent highly specialised resources embedded within particular technological ecosystems. As a result, it can be inferred that a knowledge and access divide exists for productions operating at lower entry points into VP, reinforcing earlier observations regarding inequality in skills, resources, and workflow maturity.
Moreover, applications such as Unreal Engine (which is a game engine), requires certain plugins to streamline the workflow and circumvent technical issues arising (or at least minimise them). In addition, if realism is important, then the digital assets will need to be developed accordingly with high resolution, which is both costly and time-consuming. As Interviewee 7 remarks, ‘[g]enerating or preparing those virtual content is very time consuming and sometimes can be very expensive’.
From a workflow optimisation perspective, this necessitates early task prioritisation and resource allocation. Understanding this early allows directors, cinematographers and producers to strategically plan their shots before pre-production begins. By focussing close-ups on specific sections of the LED wall, they enable digital asset teams to concentrate quality improvements in those areas only, significantly reducing costs compared to uniformly high-quality development across the entire environment. This sentiment is echoed by Interviewee 4, who states ‘we have some composition. It’s like a mid or wide shot. So the assets, you want to spend extra money to make it very detailed’. More specifically, as Interviewee 3 observes, ‘if a good producer fully understands virtual production, they would be very careful about their virtual content at the very beginning. Because the virtual content, that’s something that can directly affect the budget’.
This illustrates how early workflow decisions directly shape downstream costs and production efficiency, particularly in addressing the technical challenges with small-scale projects identified in the previous subsection, drawing on Lam et al.’s (2025) account of producing a VP trailer film within a constrained studio environment. This observation aligns with Kadner (2019) and Bevan and Mackness (2025, 13), who argue that when producers possess a clear understanding of VP workflows, virtual production can function as a cost-saving strategy rather than a source of unforeseen expense.
Interestingly, a key stakeholder from Mainland China discusses how in this region ‘virtual production is not being as popular or as well adopted’ and they suspect ‘one of the important reasons would be people, including directors or producers, underestimating the pre-production part a little bit’. This further reinforces the argument that VP success is less dependent on technology alone and more on the effective management of workflows, knowledge, and decision-making structures at the earliest stages of production.
The pre-production paradox: Spend more time to save more resources
All participants agree that virtual production filmmaking in terms of workflow relies heavily on careful planning and collaboration of many key stakeholders from the outset at the pre-production stage (Interviewee 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10). As stated by Swords and Willment (2024), 8), virtual production: Has increased the importance of the pre-production phase of the production process. This is a major change to the production network with greater time spend in pre-production, more people employed in this phase and a larger proportion of the budget being spent there.
This observation situates pre-production as a critical workflow phase in which coordination sequencing, and decision-making are deepened. Moreover, Interviewee 1 discusses the context of a one-or-two-day shoot, noting that test shoots often take place one or 2 weeks in advance. This allows sufficient time for modification – both technically and creatively – thereby reducing downstream risks to schedules and budgets.
As Interviewee 8 remarks, ‘I was to definitely say virtual production would require the director to have more pre-planning in their work process’. As interviewee 3 believes, ‘I think [when] people start underestimating the difficulties or the precision you have to achieve during pre-production, that’s when you get a bad start’. Together, these insights emphasise pre-production as decisive workflow stage rather than a preparatory formality.
In taking into consideration the previous point about budget and the creation of the digital assets, Interviewees 4 and 9 highlight the importance of a detailed storyboard (and/or shot list). After liaising with the studio stakeholders during the initial consultation, key creative stakeholders should take into account the parameters they are working (Interviewee 8). These parameters are typically defined and managed by a department specific to virtual production: the virtual art department. This department is coordinated by the virtual production supervisor, who collaborates with the physical (traditional) art department and the director to ensure the virtual and real elements converge accordingly to realise the creative intent (Swords and Willment, 2024).
It goes without saying that the initial pre-production phase involves the virtual art department providing numerous images of potential virtual environments, props, and more, while the physical art department provides corresponding references for the director (Interviewee 5). As Interviewee 4 notes, ‘we have the [art] department and virtual art department [who] should be involved together. They should be discussing together’. This highlights pre-production as a collaborative workflow environment in which parallel departments operate in continuous alignment.
Simultaneously, as the director develops the shot list and storyboard – potentially with input from the cinematographer – the virtual production supervisor coordinates between the virtual art department, the physical art department, and the lighting and camera teams. This collaboration is particularly critical as the lighting design and camera movements are defined through detailed planning documentation, such as lighting plans for both the real and virtual worlds. From a workflow management perspective, this stage represents a transformation phase in which creative intent is translated into executable technical processes.
This process requires a distinct approach to ensure a collective and streamlined result. This will ultimately lead to the final asset creation before moving onto the next aspect of pre-production, which is previs, enabling the director and relevant stakeholders to view ‘the whole virtual world in the Unreal Engine virtual environment’ (Interviewee 5). As Kadner (2019, 12) remarks, this allows the creative team to explore virtual camera angles, lighting options, and the environment itself in real time. Importantly, previs functions as a decision-support mechanism within the workflow, providing immediate opportunities for experimentation, and enabling the adaptation of storyboards, adjustments to shot composition, transitions, and pacing. Kadner (2019, 12) argues that, in contrast to traditional filmmaking – where storyboards and shot lists may themselves be understood as forms of previs – within VP these materials operate instead as ‘precursors’ to previs. In VP workflows, previs therefore represents a distinct and more operational phase, rather than a purely conceptual one. This distinction reinforces the shift identified in the literature review, which frames previsualisation – particularly through tools such as Unreal Engine – less as a representational exercise and more as a simulation of production conditions, within which testing and iterative decision-making can occur (Bevan and Mackness, 2025; Coleman, 2025).
Subsequently, techvis emerges as the practical execution plan for shots and scenes on the VP stage. This phase allows stakeholders to ensure virtual environments and physical props converge effectively with the virtual background, whilst optimising placement of physical cameras, lighting systems and practical effects (Kadner, 2019). The importance of techvis is emphasises by Kadner (2019, 12), who notes that ‘[i]t is also the area where camera moves, camera placement, and lens choices can be validated, mitigating the risk of physically implausible virtual choices’. In contrast to previs, techvis is less concerned with visual fidelity or editorial exploration and instead focuses on ‘the physics and specific camera data for use of the effects artists, and less so with visual fidelity or usage in an evolving edit’ (Kadner, 2019, 13). Within this workflow framework, techvis formalises task sequencing and resource coordination, enabling technical teams to document camera angles, focal lengths, and tracking movements to ensure alignment between physical and virtual elements.
Moreover, throughout pre-production, it is important to add that the SFX Department are collaborated throughout because, as Interviewee 6, adds, ‘there are also practical elements that we need to introduce to the virtual environment. So if we’re outside [for example], the wind needs to blow, people need to get wet when it rains’. This fixes a spotlight on the interdependency of workflows in VP, where physical effects must be synchronised with virtual environments.
Once pre-production has been concluded, production takes place, where virtual production provides real-time visualisation and iteration capabilities that traditional production workflow cannot match. In post-production, the integration of virtual elements with live footage is significantly streamlined, as many compositing decisions have already been made on set.
These findings suggest that virtual production involves all core key components of traditional workflow management systems, including: (i) input (i.e. digital assets, physical set elements, and creative decision-making), (ii) transformation (i.e. activities and processes that help to converting input into outputs using all resources and technology), (iii) output (i.e. finalised footage which fuses and composes virtual and physical elements), and (iv) stakeholders (v) (i.e. individuals or groups who have an interest or influence in the workflow such as directors, VFX supervisors, traditional film crew and Unreal Engine operators).
By mapping VP practices onto these workflow components, the analysis demonstrates how virtual production reorganises – not replaces – existing production workflows, while pronouncing the importance of early coordination, decision-making, and process management.
Conclusion
This article addresses a key gap identified in the literature: the lack of empirically grounded accounts of how virtual production workflows operate in practice. Drawing on interviews with international VP professionals and supported by workflow management theory, we propose a flexible workflow framework (see Figure 4) that enables practitioners to exploit VP’s creative potential while anticipating its technical, organisational, and budgetary constraints. Rather than offering a prescriptive model, the framework is intended as a decision-support structure through which stakeholders can align expectations, identify risks, and mitigate issues that might otherwise result in budget pitfalls or scheduling delays. Virtual production workflow.
A central contribution of the framework is the identification of a pre-development phase, which functions as a critical bridge between traditional and virtual production processes. This phase foregrounds early technical evaluation, asset feasibility, and resource planning – ideally overseen by studio manager/s or VP supervisor/s – before a project formally enters pre-production. In contrast to existing conceptualisations of VP workflows (see Figure 5), this explicitly recognises that many of the most consequential VP decisions occur prior to conventional pre-production, and that failure to address them early can destabilise downstream stages. Virtual production workflow by fox VFX lab (Deguzman, 2023).
While VP has frequently been framed as revolutionary or the inevitable future of filmmaking, our findings suggest that it is successful integration depends less on technology itself and more on careful workflow design, communication, and skills alignment. The regional disparities we observed – particularly between more established markets such as the UK and the USA versus emerging VP outfits in Mainland China – highlight the need for greater knowledge exchange, training, and transparency.
Importantly, this study confirms that there can be no single, unified VP workflow. Instead, workflows must remain project-specific, shaped by creative intent, scale, budget, and available expertise. The value of the proposed framework therefore lies not in standardisation, but in offering a shared roadmap for negotiation and planning during the earliest stages of VP decision-making. As VP continues to evolve, such frameworks will remain essential in supporting practitioners as the technology – and its practices – continue to mature.
Finally, given that VP remains in its infancy, further research is required to support the maturation of VP workflows, particularly in relation to digital asset creation and the mitigation of technical constraints (Lam et al., 2025; Swords and Willment, 2024). Several interviewees alluded to the potential role of artificial intelligence (AI) in streamlining VP processes – especially in asset development, optimisation, and pipeline efficiency (see Azzarelli et al., 2025; Bennett et al., 2021; Clavadetscher, 2014; Dhahir et al., 2022; Wang et al., 2025) – although such applications were not yet sufficiently established to be examined in detail within this study. Similarly, colour management emerged as a recurrent concern across two interviews and the literature review, emphasising the need for more focused research into colour pipelines and their integration with VP workflows.
Footnotes
Aknowledgments
Thanks to Dr Daniele Garrisi for his support with Kappa statistic.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: AHRC IAA; PR10947
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.