← All Dossiers
Latent Image – Pre-Seed
Long-Term Cultivation StoryHouse Fund II  ·  Pre-Seed  ·  Deeptech / Media  ·  Los Angeles, CA
Dossier generated 2026-07-10 by /deal-dossier  ·  Deal record: recp8XvNFl9VycO6j  ·  Source: Airtable appjxAR3LPe3fkHOp

One-Liner & Thesis

Latent Image is an industrialized AI remastering company: a patent-pending platform that converts stranded legacy 4:3/SD film and TV libraries into modern 16:9 4K at catalog scale, conditioning on each show's own footage rather than generic AI outpainting. Streaming's appetite for library content — which needs 4K/HDR/16:9 upgrades to license — has turned remastering into a high-yield asset-transformation industry, and AI is collapsing per-title cost from six figures over months to low five figures in weeks (Internal/Web).

The founder is a strong domain fit: David Carroll built production GenAI at Spotter and media-systems/workflow at Sony and DirecTV, and has iterated toward this on the same deep-content-analysis core for years. Critically, he has a signed first customer — a 25-year-old, TPN-certified Hollywood post house that put cash down and asked him to raise, with early interest in a studio POC (Internal). The drag is pre-POC technical risk, single-customer concentration, and a solo founder with no institutional syndicate; the raise exists precisely to retire that risk.

Streaming platforms' demand for library content, which requires 4K/HDR (and increasingly 16:9) upgrades to license, has turned film/TV restoration into a high-yield asset-transformation industry (Web). AI has compressed per-title cost from $100-500K over 12-18 months to roughly $8-60K in weeks, expanding the addressable catalog dramatically (Web). The digital film restoration market is ~$1.45B in 2025 growing ~11.7% CAGR, with adjacent AI video-upscaling growing 20-35% CAGR toward $1-1.7B by 2030-33 (Web). The core risk is that incumbents (DNEG/PFT, Deluxe, Technicolor) already offer AI-assisted 4K remastering at scale, so a new entrant must prove its per-show-conditioned approach is meaningfully better than generic outpainting to avoid commoditization (Web).

Investment Score & Recommendation

61/ 100
HOLD

Strong domain-fit founder (Internal) plus a genuine streaming-driven remastering tailwind (Web) are the biggest drivers; pre-POC technical risk on a solo founder with a single signed customer is the biggest drag, holding this at pre-seed watch until the 4K proof-of-concept retires the core risk.

Momentum: SteadyRed flags: 3 / 9Confidence: Low
Market & TAM7/10
25% weight
Team & Founder8/10
25% weight
Product & Traction5/10
20% weight
Deal Terms6/10
20% weight
VC Syndicate2/10
10% weight

Deal Box

Funding Round
Pre-Seed
Use of Funds
4K proof-of-concept
Fund
StoryHouse Fund II
Signed Customer
1 (Hollywood post house, cash down)

Company Snapshot

Sector
Deeptech · Digital Media
Location
Los Angeles, CA
Year Founded
2026
Status
Private

Market Size

$1.45B
Digital film restoration (2025)Web
~11.7% CAGR to ~$3.2B by 2032
$500M
AI video upscaling SW (2025)Web
20-35% CAGR; ~$1.1-1.7B by 2030-33
$8-60K
AI-assisted restoration / titleWeb
vs $100-500K + 12-18mo manual

Streaming's appetite for library titles (4K/HDR/16:9 upgrades to license into modern platforms) is the structural demand driver, and AI is collapsing per-title cost from six figures/months to low-five-figures/weeks (Web). Catalog-scale, per-show-conditioned remastering targets exactly this monetization pressure, but the addressable slice is service-industry sized, not platform-SaaS sized.

Competition

PlayerPositioningFunding / StageEdge vs. them
Latent ImageIndustrialized, per-show-conditioned 4:3/SD to 16:9 4K at catalog scalePre-seed
Prime Focus Technologies / Lowry (DNEG)Incumbent 4K remaster/restore/upconvert; CLEAR AI platformAcq'd by DNEG Jul 2024Studio relationships, scale, in-house AI + DNEG backing
Deluxe / Technicolor / Picture ShopOutsourced restoration infra for studios; built/acquired AI capabilityEstablished / PE-backedEntrenched studio contracts, breadth of post services
The Pixel Farm / Topaz-class upscalersRestoration software tools / generic AI upscalingIndependent / SMBOff-the-shelf tooling, cheap but generic (outpainting artifacts)

Moat: Conditioning the model on each show's own footage rather than generic AI outpainting is the differentiation claim, but it is patent-pending and unproven at scale against well-capitalized incumbents (Web/Internal).

Traction

1
Signed customerInternal
cash down
25 yrs
Customer tenureInternal
TPN-certified post house
~18 mo
Founder known to SHInternal

Exit Potential

Strategic M&A
Likely Path
4-7 yr
Time to Liquidity
Acqui-hire → mid-8-figure strategic
Return Scenario

Consolidation is the clear pattern: DNEG acquired Prime Focus Technologies (Jul 2024) and PFT earlier absorbed Lowry Digital; Deluxe, Technicolor and Picture Shop all built or bought AI restoration capability (Web). Natural acquirers are media-tech/post majors (DNEG/PFT, Deluxe, Technicolor, Picture Shop) or a studio internalizing catalog remastering (Sony, Warner, Disney, Universal, Netflix) once the tech is proven. Base case is a modest strategic acqui-hire; upside is a mid-eight-figure acquisition if per-show conditioning becomes a defensible studio-grade pipeline. At a pre-seed entry into a risk-retiring POC round, a strategic exit in this range can still clear a strong multiple on a small check.

Founders

David Carroll
Founder
Media-technology builder: production GenAI at Spotter and media-systems/workflow work at Sony and DirecTV. Iterated for years on a deep-content-analysis core (WatchMate, TrueFeed, SageID) before a Hollywood post house brought him the remastering problem that became Latent Image. Known to StoryHouse for ~18 months.

Open Questions & Risks

Next Steps

Latest Meeting Notes

2026-06-27 Email update Signed post house, raising for the 4K POC

David Carroll emailed Josh and Miles: after iterating on the same deep-content core for ~18 months, a Hollywood post house brought him the remastering problem, and he now has a signed, cash-down first customer and is raising a focused pre-seed.

Source: Meeting Notes rec9KmJSMmPviA717

Deal Timeline