In a rented storage unit forty minutes outside Nagoya, Hiroko Tanabe keeps nine thousand painted acetate sheets in archival sleeves, sorted by studio and then by episode. None of them are legally hers, and none of them are legally anybody else's either. That ambiguity is the whole business.
Production cels — the hand-painted transparencies photographed frame by frame before digital compositing took over in the early 2000s — were never treated as assets. Studios binned them, gave them away at wrap parties, or sold them in bulk to shops in Akihabara for the price of the plastic.
Thirty years later the same sheets change hands at auction for the price of a car, and nobody involved can produce a chain of title. The studios that made them have merged, folded, or been bought by holding companies with no records older than 2005.
A market with no paperwork
The dealers are candid about it. Provenance, in practice, means a photograph of the seller's grandfather standing in a studio corridor in 1987, and a story that hangs together. Buyers accept this because the alternative is no market at all.
Tanabe started buying in 2011, after a dealer offered her a sequence from a series her father had worked on as an in-betweener. She paid four times what it was worth and has been paying above market ever since, on the theory that somebody should.
Her catalogue is now consulted by two universities and, awkwardly, by one of the studios whose work she holds. They asked for scans of a sequence they no longer had a record of. She sent them free of charge, and asked for nothing except a line of credit in the restoration.
They threw the artwork away and kept the paperwork. Now the paperwork is gone too.
What restoration actually needs
Restoration teams need the cels less for the paint than for the registration marks: the punched holes and pencil notes that record exactly how a frame was lined up under the camera. Without them, a 4K rescan is guesswork about what the director intended.
This is where the private collections start to matter more than anyone expected. A studio wanting to reissue a 1994 feature may find that the only surviving reference for eleven shots sits in a collector's sleeve in Osaka, and that the collector has strong views about how the reissue should be graded.
Some of those conversations go well. Others have ended in lawyers' letters over material that neither side can prove they own, which is the sort of dispute that produces no case law and a great deal of expense.
What is emerging instead is an informal register — a handful of collectors, two universities and a preservation society, sharing scans and photographs on the understanding that nobody asserts anything. It is not a legal solution. It is the only one anybody has.