Postproduction Services and Other Motion Picture and Video Industries (U.S.)

NAICS 2022 code 51219 — a Histometrics rollup primer for public-market and private investors. (NAICS = North American Industry Classification System.)

1. Overview

When a film, series, streaming show, or commercial finishes shooting, it still has to be finished: edited, color-corrected, filled with visual effects (VFX), titled, captioned, translated into other languages, mastered for release, and — for older material — developed, scanned, restored, or licensed from a footage library. NAICS 51219 gathers the "everything after the shoot, minus distribution and the theater" service businesses into one industry group, made up of two very different children [1]:

• 512191 — Teleproduction and Other Postproduction Services: the large one. Editing, color, VFX, animation, subtitling and captioning, localization, and digital mastering — the finishing work done for hire.
• 512199 — Other Motion Picture and Video Industries: the small one. A specialized craft-and-licensing niche: motion-picture film laboratories, stock-footage libraries, and film preservation/restoration.

The honest headline is that this level is 512191 wearing a slightly larger hat. On our federal figures, postproduction is roughly 96% of the group's revenue and employment; the "other" bucket is about 3–4% [2][3]. So most of what follows is postproduction economics — a project-based, labor-intensive, thin-margin services business that levers studio and streamer content budgets — with a distinct, structurally different tail (labs, libraries, archives) bolted on.

For investors the practical takeaways are the same across both children: there is no clean pure-play U.S.-listed stock for either industry, the operating companies live mostly in private and private-equity (PE) hands, and the two forces reshaping media at large — content-budget discipline and generative artificial intelligence (AI) — are reshaping this group too. Sections 4 and 10 lay out the real routes in.

2. What's inside — the two children and how they differ

The distinctive thing about this level is the contrast between its two children. They share a supply chain (both serve the content factory) but differ in size by more than an order of magnitude, in market structure, in economics, and in direction of travel.

Three contrasts worth holding onto:

1. Size is lopsided. Postproduction out-earns the "other" bucket by roughly 27 to 1 [2]. Any statement about "the industry group" is, numerically, mostly a statement about 512191.
2. Concentration runs the opposite way from size. The tiny child is the more concentrated one (HHI ~709 vs. ~184; CR4 ~47% vs. ~22%) [2]. A handful of surviving film labs and footage libraries dominate their small pond, while postproduction is a genuinely fragmented field of boutiques beneath a few global groups. That is why the whole level's HHI (~176) is even lower than postproduction's own — blending in a small, unrelated segment dilutes rather than raises measured concentration [2].
3. Their futures diverge. Postproduction demand rises and falls with how much content gets made, so it is cyclical around a broadly durable trend. The "other" bucket is structurally bifurcated — photochemical processing keeps shrinking as digital and AI automate it, while preservation and archive-licensing (including selling clean footage catalogs to AI model developers) give the winners a defensible floor.

Also note the establishment/firm counts: 512199 holds ~8% of the group's establishments but only ~4% of its revenue [2][3], i.e. its shops are even smaller on average (~$1.3M receipts each) than postproduction's (~$2.1M each) [2]. Both children are, by federal standards, small-business industries end to end.

3. How big it is

Our ground-truth federal figures for the whole group, NAICS 51219 (2022 Economic Census for receipts/firms/concentration; 2023 County Business Patterns (CBP) for establishments, employment, and payroll — note the two are different survey years and count different units):

What the numbers say. This is a fragmented, competitive, labor-intensive industry group. An HHI of 176 sits far below the 1,500 line economists treat as even "moderately concentrated," and the four largest firms hold barely a fifth of revenue [2]. Payroll alone — $2.51 billion against $6.66 billion of receipts, about 38% — before adding freelance and contractor spend, shows how much of every revenue dollar goes to skilled people [2][3]. The average establishment is small: roughly $2 million of receipts and about seven employees [2][3].

Our federal file for this level carries no figure for operating margin, capital spending, utilization, debt, or industry growth, and does not break out ownership (public vs. private vs. PE vs. individual). We do not infer those; nothing here is a government forecast.

The undercount caveat — it matters, and it runs in several directions:

• In-house post is invisible. When a studio or streamer edits and finishes a show with its own staff, that spend is booked under production (512110), not here [1].
• Sound post lives elsewhere. Dialogue re-recording, sound design, and mixing sit in Sound Recording Studios (512240), so a whole "post" discipline is excluded by definition [1].
• The big commercial cousins are classified elsewhere. Large-scale stock-photo licensing and much of what people picture as "footage" (Getty, Shutterstock) is placed in the internet-publishing/data families, not here — so the household names dwarf the code that literally names the activity [1].
• Runaway work leaves the country. A large share of post and VFX for U.S. content is performed in Canada, the U.K., India, and Australia to chase tax incentives and lower labor cost — none of it in U.S. business statistics [6][15].
• Freelancers, nonemployers, and public archives are thinly captured. CBP and the Economic Census primarily cover employer businesses; gig editors and sole-proprietor scanning/restoration work are largely omitted [5]. Film preservation is dominated by government and non-profit institutions (the National Archives' Motion Picture Preservation Lab, the Library of Congress, the Academy Film Archive) that sit outside business statistics entirely [5][26]. Because small, individual, and institutional ownership dominates the fringes of both children, treat the $6.66 billion as the for-hire, private, employer slice of a materially larger activity.

Private market-research estimates using broader definitions put U.S. postproduction closer to $7–9 billion, with U.S. VFX alone near $3.3 billion [27]. Treat those as approximations, not official counts.

4. The investable universe — where value concentrates across the children

No pure-play, either side. Neither child has a clean U.S.-listed stock. The VFX house that came closest to becoming one — DNEG — terminated its 2022 special-purpose-acquisition-company (SPAC) listing plan [7]. So public exposure to this whole group is indirect or embedded, and the operating companies live mainly in private and PE hands. Where value concentrates differs by child.

In 512191 (postproduction — ~96% of the level), value sits in three public rings plus a deep private core:

• Near-direct listed operators (mostly foreign): Prime Focus (India; NSE: PFOCUS / BSE: 532748), which owns Oscar-winning VFX house DNEG, and IMAGICA Group (Tokyo: 6879), parent of localization/post platform Pixelogic — both thinly traded for U.S. investors [7][8]. IMAX (NYSE: IMAX) does genuine film-remastering work but is primarily an exhibition-and-technology company [9].
• Content buyers (the demand side / vertical integration): Disney (DIS, owns Industrial Light & Magic), Netflix (NFLX, owns Scanline, now Eyeline Studios), Sony (SONY, owns Pixomondo), Warner Bros. Discovery (WBD), Comcast/NBCUniversal (CMCSA), Paramount Skydance (PSKY), Amazon (AMZN), Apple (AAPL) [10]. Post is a cost line inside these giants, not a segment.
• Picks-and-shovels toolmakers: Adobe (ADBE), Autodesk (ADSK), Dolby (DLB), Nvidia (NVDA), Apple (AAPL) — they capture post activity regardless of which facility wins the job, but post is a sliver of each.
• Private core (where the operators actually live): Deluxe (Platinum Equity) [11]; FC3 / Framestore–Company 3 (Aleph Capital, Crestview Partners) [12]; Streamland Media [13]; DNEG (via Prime Focus); Wētā FX; localization scale players Iyuno and TransPerfect Media [14]; and hundreds of independent editorial and boutique VFX shops. Two once-public names went private recently — Keywords Studios (to EQT) and Avid Technology (to Symphony Technology Group) [20].

In 512199 (labs, libraries, preservation — ~4% of the level), value sits in a few proxies plus small private shops:

• Film-lab / film-stock proxy: Eastman Kodak (NYSE: KODK) — the closest listed operating exposure, making motion-picture film stock and running a film laboratory, though a small part of a diversified company [21].
• Stock-footage + AI-licensing proxies: Getty Images (NYSE: GETY) and Shutterstock (NYSE: SSTK) — the cleanest listed footage-library exposure and the emerging AI data-licensing line, but both are stills-primary and statistically classified outside 512199 [22][23].
• Private / other owners: family-owned film labs (FotoKem, Cinelab, Colorlab, Kodak's lab); footage libraries (Pond5, Envato, Storyblocks, Filmsupply); preservation-for-hire (Prasad North America, Pacific Title Archives); and non-investable public archives (National Archives, Library of Congress) [26].

The practical read across both children: public investors buy the ecosystem — content buyers, toolmakers, a few foreign-listed operators, and library proxies — while private investors buy the operators. There is no way to own "NAICS 51219" as such.

5. How the money works

The two children run on genuinely different engines — one of the clearest contrasts in the group.

512191 — capacity economics. Postproduction is a project-based professional-services business, closer to a law firm than a factory. Revenue = billable hours (or VFX shots) × rate, and the master metric is utilization — the share of artists, edit suites, storage, and compute that is billed rather than idle. Costs are dominated by skilled labor plus real estate, render farms, storage, cloud compute, security, and software. Cash flow is feast-or-famine: work arrives on production calendars the facility does not control, so when greenlights slow the pipeline empties while the fixed-cost base stays put. The VFX corner adds the fixed-bid trap — studios demand fixed quotes then request changes, reshoots, and "one more version," and the vendor eats the overruns [6]. Scale, in practice, has not reliably produced margin.

512199 — three engines under one roof. (A) Fee-for-service labs and preservation earn on billable utilization of technicians and capacity utilization of expensive machines (processors, scanners, restoration suites); when photochemical volume falls, fixed costs do not, and margins collapse — survivors earn pricing power by being among the last able to do the work. (B) Licensing libraries treat the catalog as a capital asset: once a clip is shot or acquired, each additional license sells at near-zero marginal cost, so incremental margins are very high; the metrics are library size, download/license volume, revenue per download, and retention. (C) AI data-licensing — bulk-licensing clean archives to AI model developers — is the new high-margin line, but only the few players with catalogs large and rights-clean enough can sell it (Shutterstock reported over $100M of AI data-licensing revenue in 2024) [23].

What to watch in an operator, either side: utilization, revenue per billable hour or per download, gross margin by service line, rework and turnaround, recurring/managed-service revenue, backlog vs. available capacity, work-in-process and unbilled receivables, customer concentration, royalty/content-acquisition cost, and — for labs and archives — equipment condition, security, and insurance. In both children the strongest operators sell a defensible workflow, rights certainty, and specialized know-how — not raw processing capacity sold at volatile prices.

6. What drives demand

• Studio and streamer content budgets — the master variable for the whole group. Both post volume and footage/lab demand track how much content gets made. The 2010s "peak TV" boom was a tailwind; the 2023–2025 pivot to profitability cut greenlights and squeezed budgets [6].
• Production shutdowns upstream. The 2023 dual strikes — the Writers Guild of America (WGA) and the Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) — halted shooting, and because post sits downstream, the work dried up months later; series in production in early 2024 were down roughly 30% versus 2022, with a slow recovery since [16].
• Rising production values. 4K/8K, high-dynamic-range (HDR), VFX-heavy franchises, and virtual production raise the amount of post per project even when project counts flatten.
• Global streaming and localization. International distribution needs dubbing, subtitling, and accessibility versions in dozens of languages — the fastest-growing slice of 512191 demand [14].
• Digitization and re-monetization of legacy film (512199). Aging collections drive scanning and restoration; studios reissuing back catalogs in 4K/HDR drive film-to-digital transfer — a relatively defensive demand pocket.
• State and national tax incentives. Programs that cover qualifying VFX/post spend decide where geographically mobile work lands [15].
• AI — both directions. It compresses hours in commodity tasks (rotoscoping, cleanup, captioning, versioning) and creates a new demand line (permissioned training-data licensing), while threatening to commoditize generic footage and undercut prices [23][28].

7. Regulation

Neither child is licensed or rate-regulated; the rules that move the economics are spread across tax incentives, labor, copyright/AI, accessibility, and content security.

• Tax incentives (the biggest lever). Because post and VFX are mobile, state/national film incentives function as industrial policy over where the work is done. California more than doubled its Film & Television Tax Credit to $750 million a year (Program 4.0, 35–40% refundable, explicitly covering qualifying VFX and virtual production) to fight runaway production; dozens of states run competing programs [15].
• Labor. Federal baselines run through wage/overtime and organizing law. Many post workers are represented by IATSE (the International Alliance of Theatrical Stage Employees); VFX crews began unionizing with IATSE in 2023–2024, and recent Hollywood agreements add wage, benefit, and AI provisions. Upstream union actions can stop the whole pipeline [16][17].
• Copyright and AI. Copyright governs ownership, chain of title, and AI-generated material — the core legal risk for labs and libraries alike. The unresolved question of whether AI developers must pay to train on footage is consequential: in Getty Images v. Stability AI, the England & Wales High Court in November 2025 largely rejected Getty's copyright claims, a warning that the data-licensing upside is not legally guaranteed [18][24].
• Accessibility. Federal Communications Commission (FCC) rules require captioning and audio description for covered programming; the Twenty-First Century Communications and Video Accessibility Act (CVAA) and the Americans with Disabilities Act (ADA) push captioning demand into web, education, and government. (Captioning is 512191 work; it bears on 512199 mostly through clients.)
• Content security. Handling pre-release content requires meeting the Motion Picture Association's (MPA) standards, audited through the Trusted Partner Network (TPN) — a real cost and a barrier for small shops [19]. Photochemical labs additionally carry environmental and chemical-handling duties, and archives carry data-security obligations.

8. Consolidation and competitive dynamics

The federal data describe a fragmented group (HHI ~176; CR4 ~21.5%) [2], but the two children consolidate for different reasons.

• 512191 — PE roll-ups and cost competition. For a decade PE firms have assembled multi-discipline platforms (Deluxe under Platinum Equity; FC3 under Aleph/Crestview; Keywords under EQT) betting on one-stop-shop scale, global time-zone coverage, and procurement leverage with big studios [11][12][20]. U.S. facilities also compete against subsidized lower-cost hubs abroad. But scale has not conferred safety: Technicolor — owner of MPC, The Mill, and Mikros — collapsed into insolvency in 2025, scattering its units to buyers and sending shockwaves through VFX; high fixed cost, fixed-bid risk, a post-strike demand air-pocket, and heavy debt proved fatal even for a global leader [6].
• 512199 — library deals and lab attrition. Consolidation here runs through catalog acquisitions rather than greenfield build-out. Getty and Shutterstock agreed to merge in January 2025 but terminated the deal in July 2026 after the U.K. Competition and Markets Authority (CMA) demanded an editorial divestiture — a reminder that content-library tie-ups attract scrutiny [25]. Film labs, meanwhile, have consolidated to a handful of survivors after digital cinematography gutted volume.

Judgment across the group: scale and a global footprint do not guarantee sustainable margins; the durable operators own defensible workflows, irreplaceable archives, or trusted client relationships rather than undifferentiated capacity.

9. Risks

• Utilization / content-cycle dependence. A weak content slate leaves expensive labor and machines idle; the post-"peak-TV" pullback shrinks the addressable pool of projects for both children [6].
• Client concentration and pricing power. A handful of studios and streamers are the buyers; mergers (WBD, Paramount–Skydance) mean fewer greenlights and more leverage over vendor pricing and payment terms.
• Fixed-bid / execution risk (512191). The VFX bidding model transfers overruns to the vendor; scope creep can turn profitable projects into losses [6].
• Structural decline of physical film (512199). The film-lab addressable market keeps shrinking; survivor economics are volume-dependent and fragile.
• Runaway production / location risk. Mobile work migrates to wherever incentives, wages, and currency are most favorable — a permanent competitive drag [15].
• Labor and strike risk. Upstream WGA/SAG-AFTRA actions can freeze the pipeline; unionizing the post workforce raises the cost base [16][17].
• AI substitution — double-edged. AI compresses hours (a productivity gain) while threatening headcount and per-unit prices; it can simultaneously lower a facility's costs and erode its pricing [23][28].
• IP/legal uncertainty. If courts let AI developers train on footage without paying, the data-licensing upside erodes and libraries risk being scraped for free [24].
• Thin balance sheets and private-company opacity. High fixed cost, lumpy cash flow, and thin margins leave little room for error (Technicolor is the proof), and most operators disclose no audited financials [6].

10. How to invest, and the outlook

Public routes (all indirect or embedded):

• Near-direct operators (postproduction): Prime Focus (owns DNEG) and IMAGICA (owns Pixelogic) — foreign-listed, thinly traded [7][8]; IMAX carries smaller embedded exposure [9].
• Demand-side / vertical integration: the content buyers whose budgets drive volume — DIS, NFLX, SONY, WBD, CMCSA, PSKY, AMZN, AAPL. A bet on content spend broadly, not a post segment [10].
• Picks-and-shovels: ADBE, ADSK, DLB, NVDA, AAPL — post activity regardless of which facility wins, but a sliver of each [—].
• Labs/footage proxies (512199): Kodak (KODK) for the film-lab/stock niche; Getty (GETY) and Shutterstock (SSTK) for footage libraries and AI data-licensing optionality — read segment disclosures closely, since each tracks much broader businesses [21][22][23].

Private routes (where the operators actually live):

• PE-backed platforms and founder-owned houses — Deluxe, FC3, Streamland, DNEG, Wētā FX, Iyuno, TransPerfect (post); and niche film labs, footage catalogs, and archive/preservation businesses (other) — accessed through PE funds, secondaries, private credit, or direct investment [11][12][13][14].
• Distressed and roll-up opportunities — Technicolor's collapse and the ongoing film-lab shakeout create asset sales; the diligence question is whether a target owns a defensible workflow and archive, or merely sells labor and capacity at volatile prices [6].

Because there are no clean public comparables, value operators on normalized project economics — utilization, gross margin by service line, rework, backlog quality, recurring revenue, customer concentration, unbilled receivables, archive/rights quality, incentive dependence, and leverage — not headline multiples.

Outlook (forward-looking judgment, not settled fact). The group's dominant child, postproduction, faces a slow production recovery below the peak-TV high, with the pace of studio greenlights the swing factor; incentive programs like California's Program 4.0 should repatriate some work, partly offsetting offshoring [15][16]. The small child bifurcates further: legacy photochemical and manual-transfer work keeps compressing as AI automates it, while preservation, differentiated footage, and AI data-licensing give the winners a floor [23][28]. The shared wild card is AI, which could relieve the group's chronic margin problem by cutting labor hours or commoditize the work and compress prices faster than costs fall — most likely thinning routine roles while raising the premium on senior creative and archival expertise. Bottom line: a fragmented, labor-intensive services group that levers content spending, earns thin margins in good times, and fails hard in bad ones; the honest ways in are to own the content buyers, toolmakers, and a few foreign-listed operators and library proxies on the public side, and the private operators on the other.

---

Sources

1. U.S. Census Bureau, 2022 NAICS Definitions — 51219 and its children 512191 (Teleproduction and Other Postproduction Services) and 512199 (Other Motion Picture and Video Industries), with cross-references to 512110, 512120, 512131/512132, 512240, and 325992. https://www.census.gov/naics/
2. U.S. Census Bureau, 2022 Economic Census — receipts, firm counts, and concentration ratios (CR4/CR8/CR20/CR50, HHI) for NAICS 51219 and children (Histometrics ground-truth federal file for this level). https://www.census.gov/programs-surveys/economic-census.html
3. U.S. Census Bureau, County Business Patterns (CBP) 2023 — establishments, paid employees, annual and Q1 payroll for NAICS 51219 and children. https://www.census.gov/programs-surveys/cbp.html
4. U.S. Small Business Administration, Table of Small Business Size Standards (512191 = $39.0M; 512199 = $28.5M average annual receipts), 2023. https://www.sba.gov/document/support-table-size-standards
5. U.S. Census Bureau, 2022 Economic Census Methodology / Understanding NAICS — employer vs. nonemployer coverage and government/non-profit exclusion. https://www.census.gov/programs-surveys/economic-census/year/2022/technical-documentation/methodology.html
6. Variety, "Behind the 'Sickening' Collapse of Technicolor and Why Its Failure Is Sending Shockwaves Through the VFX Industry," 2025. https://variety.com/2025/artisans/global/technicolor-collapse-shockwaves-vfx-1236326607/
7. Variety, "DNEG to Go Public in $1.7 Billion SPAC Deal" (and subsequent termination, SEC Form 8-K), 2022. https://variety.com/2022/biz/news/dneg-public-1-7-billion-spac-1235162649/
8. Prime Focus Limited, "Investor Centre" (parent of DNEG; NSE/BSE listing), 2026; IMAGICA Group, "Stock Information / Pixelogic." https://www.primefocus.com/investor-centre/resources/; https://www.imagicagroup.co.jp/en/ir/stock/situation.html
9. IMAX Corporation, 2025 Form 10-K (film-remastering / Digital Media Remastering revenue), 2026. https://www.sec.gov/Archives/edgar/data/921582/000162828026011770/imax-20251231.htm
10. Netflix, "Bringing the Best in VFX and Virtual Production Together as Eyeline" (Scanline VFX), 2025. https://about.netflix.com/en/news/bringing-the-best-in-vfx-and-virtual-production-together-as-eyeline
11. Platinum Equity, "Deluxe," 2026. https://www.platinumequity.com/our-company/deluxe/
12. FC3, "About," and Crestview Partners, "Crestview and Aleph Capital Back Framestore's Acquisition of Company 3/Method," 2020. https://www.fc3group.com/about
13. Streamland Media, "Postproduction Companies," 2026. https://streamlandmedia.com/
14. TransPerfect, "TransPerfect Acquires Blu Digital Group," 2025; Iyuno (media localization). https://www.transperfect.com/about/press/transperfect-acquires-blu-digital-group; https://en.wikipedia.org/wiki/Iyuno
15. Entertainment Partners, "California Expands Film & TV Tax Credit to $750M (Program 4.0, 35–40% refundable)," 2025; National Conference of State Legislatures, "State Film and Television Incentive Programs," 2025. https://www.ep.com/blog/california-expands-film-tax-incentive-to-$750m-in-2025-to-keep-productions-at-home-in-hollywood/; https://www.ncsl.org/fiscal/state-film-and-television-incentive-programs
16. Wikipedia, "2023 Writers Guild of America strike" and "2023 SAG-AFTRA strike"; VFXwire, "How the VFX Industry Is Recovering From Last Year's Strikes," 2023–2024. https://en.wikipedia.org/wiki/2023_SAG-AFTRA_strike
17. The Wrap / Variety / IATSE, "Marvel Studios VFX Workers Vote to Unionize with IATSE" (2023) and IATSE Hollywood Basic/Area Standards ratification with AI provisions (2024). https://iatse.net/iatse-members-overwhelmingly-ratify-hollywood-basic-and-area-standards-agreements/
18. U.S. Copyright Office, "Copyright and Artificial Intelligence," 2025. https://www.copyright.gov/ai/
19. Trusted Partner Network (Motion Picture Association), "MPA Content Security Best Practices and TPN Assessment," 2025. https://www.ttpn.org/
20. Keywords Studios / EQT acquisition completion (2024); NewscastStudio / SEC (Avid Technology 8-K), "STG Completes Acquisition of Avid for $1.4B," 2023. https://www.keywordsstudios.com/en/about-us/news-events/news/keywords-studios-eqt-acquisition/; https://www.newscaststudio.com/2023/11/07/private-equity-firm-stg-completes-acquisition-of-avid-for-1-4b/
21. Eastman Kodak Company, Annual Report on Form 10-K for 2025 (motion-picture film products and processing laboratory), 2026. https://www.sec.gov/Archives/edgar/data/31235/000119312526104214/kodk-20251231.htm
22. Getty Images Holdings, Inc., FY2025 Form 10-K (revenue ~$981M; net loss ~$206M; growing video demand), 2026. https://www.sec.gov/Archives/edgar/data/1898496/000162828026018160/gety-20251231.htm
23. Shutterstock, Inc., Full-Year 2025 Results / Form 10-K (revenue ~$990M; adjusted EBITDA ~$272M; owns Pond5, Envato) and Reuters/MarketBeat on AI data-licensing (>$100M in 2024; OpenAI deal up to $250M by 2027). https://investor.shutterstock.com/news-releases
24. Latham & Watkins / Mayer Brown, "Getty Images v. Stability AI — England & Wales High Court judgment (Nov 2025), copyright claims largely rejected," 2025. https://www.lw.com/en/insights/getty-images-v-stability-ai-english-high-court-rejects-secondary-copyright-claim
25. Getty Images / Shutterstock, "Termination of Merger Agreement" (Form 8-K; terminated July 7, 2026 over U.K. CMA editorial-divestiture remedy; U.S. clearance unconditional), 2026. https://www.sec.gov/Archives/edgar/data/1549346/000114036126028035/ef20077612_8k.htm
26. U.S. National Archives, "Motion Picture Preservation Lab," 2023; Library of Congress "National Film Registry"; Academy Film Archive. https://www.archives.gov/preservation/products/definitions/mopix-lab
27. IBISWorld, "Video Postproduction Services in the US," 2026; Verified Market Research / Precedence Research, "Post-Production and VFX Market Size" estimates, 2024–2025. https://www.ibisworld.com/united-states/industry/video-postproduction-services/1247/
28. Genra.ai, "Generative AI video (Sora, Runway, Veo) and the displacement of generic stock B-roll," 2025–2026. https://genra.ai/blog/sora2-runway-gen4-veo3-ai-video-licensing-guide