Who Killed the IMAX Projector?

On July 17, Christopher Nolan’s The Odyssey opened to $264 million worldwide, the biggest debut of his career.

Of that total, $6.3 million came from 41 rooms: the only theaters left on Earth that can project IMAX 15/70 film, the format Nolan shot the entire movie on.

Those rooms averaged $153,000 a screen, nearly three times what a digital IMAX screen earned. Showtimes sell out seven weeks in advance. The 2 a.m. screenings are sold out; theaters added 3 a.m. shows, and those sold out, too. In New York, resale tickets have cleared $1,000. Not for a premiere. For a random Tuesday in August.

Everyone treats this frenzy as an act of God: there are only 41 rooms, nothing to be done.

This storyline skips the interesting part: Nobody has built one of these projectors in decades.

The production line is gone.

The suppliers are gone.

Most of the engineers are retired.

And the firm that shut the manufacturing line, deliberately, with its eyes open, was IMAX.

The conventional narrative is a triumph-of-cinema story: make something special enough, and audiences will crawl back at 2 a.m.

Of course, there is truth in that, but for me, as an operations professor, this is the least interesting layer of the case.

The more interesting story is about a machine designed in 1970 becoming the scarcest asset in global entertainment through a sequence of rational operational decisions.

Long-time readers of this newsletter know the story.

Why the Line Died

The machine itself is a piece of Nixon-era engineering. In the late 1960s, a Brisbane machinist named Peter Jones invented the rolling loop, a transport that moves film horizontally in gentle waves rather than yanking it vertically past a lens; a small group of Canadians bought the patent, and one of them, engineer William Shaw, made it work at scale. The frame it projects is roughly ten times the area of 35mm film. The machine weighs 1.8 tonnes, and a single print of a feature arrives on a platter that weighs as much as a grand piano.

These projectors were never mass-produced. IMAX built them by hand, a few units a year, each one sold to a science museum or an aquarium that bolted it into a purpose-built room and ran 40-minute documentaries about coral reefs on it for two decades. The commercial film industry treated 15/70 as a curiosity.

A hand-built machine for a market that small is a craft workshop with a backlog, and a craft workshop survives only as long as the backlog does.

The backlog died in the 2000s, and IMAX killed it from both ends at once.

On the demand side, the institutional market had largely been built out: the museums that wanted a giant-screen theater had one, and the ones replacing equipment increasingly wanted digital, which played more content for less money. On the strategy side, IMAX made the pivot that built the modern company.

In late 2007, it struck a deal with AMC to roll digital IMAX systems into roughly 100 multiplexes, and the business model changed from selling a handful of custom-built machines to institutions into converting commercial screens at an industrial scale.

The digital system fit into an existing booth and played a file.

The film system needed a reinforced floor, a platter rig the size of a dinner table, a 600-pound print, and a trained projectionist.

Every unit of growth the company wanted argued for digital, and it wanted thousands of units. The 15/70 line died of its owner’s success.

If we want to understand the decision on its own terms, it was the right decision: Between 2008 and today, IMAX grew from a few hundred theaters, many of them institutional, into a network of more than 1,800 systems, with nearly all of that growth driven by digital and laser.

Keeping a hand-assembly line alive for a projector nobody was ordering, staffed by specialists with nothing to build, supplied by vendors who were themselves exiting the film business, would make no financial sense.

It’s pretty clear that no spreadsheet had a line item for what it would one day cost to be unable to make this machine.

Ten years later, there is no line to restart, because what you shut down was never just a line. It was an ecosystem, and ecosystems do not restart.

So, what is the Root Cause?

The tempting answer is a forecasting failure: nobody predicted that film would become the premium way to see a movie.

The tempting answer is wrong because IMAX’s forecast was correct: Film projection did shrink to almost nothing, and it is still almost nothing: 41 rooms and 2.4 percent of this film’s opening gross.

But here is the issue: The capital review missed what no average-based review is built to see.

The value of this capacity was never in its mean but in its tail.

In other words, A 15/70 room is worthless most weeks and priceless the week a Nolan film opens, and a review that discounts average utilization will kill every peaker plant in the portfolio.

The demand now crushing the system did not exist to be forecast; it was created afterward, partly by the scarcity itself.

The Dark Knight’s IMAX sequences premiered in 2008, the year the AMC rollout began, and it took 15 years of Nolan releases for shot-on-IMAX film to go from a gimmick to a pilgrimage.

You just cannot project the demand that your own exit helps create.

But even more so, the asset changed categories.

In 2007, a film projector was a delivery technology, valued on cost per showing, a contest digital wins forever.

By 2026, the same machine is an authentic good, valued for being the unreproducible original, the way a vinyl pressing plant or a letterpress shop is valued.

Could IMAX rebuild it now, with demand at an all-time high?

The entire future demand for new 15/70 projectors is a Nolan-scale film event every two or three years, playing in a few dozen rooms.

Standing up a production line from scratch, suppliers, tooling, reverse-engineered drawings, trained assemblers, against a market of perhaps a few dozen lifetime units is a capital project with no plausible return. Rich Gelfond, IMAX’s CEO, has said as much: “There’s certainly more demand. The problem is they haven’t made new IMAX film projectors in about 50 years.” And the fix: “We build new projectors every day, but film projectors using this film, it’s not practical.” What is practical, he says, is to find the old ones, retrofit them, rebuild them.

So that is what the company does.

It spent more than a year before this release hunting projectors in decommissioned museum theaters, cannibalizing some for parts and rebuilding the rest, which is how the network grew from 30 rooms to 41 for The Odyssey.

That is capacity recovery, the operational equivalent of archaeology.

The supply curve for the critical asset is not steep… It is vertical, at any price… Trapped capacity, literally.

And the projector is only the last link.

Follow a single frame of The Odyssey from set to screen, and you pass through five links, every one of them the last of its kind:

This is what the phrase industrial base flattens.

A format is a chain in series: it lives only if every link lives, and every link here is a monopoly, which means the whole format is one bankruptcy, one lease dispute, one retirement away from extinction.

And it nearly happened.

FotoKem is the last lab on Earth that strikes 70mm prints; every frame of this movie passed through one building in Burbank, where technicians hand-spliced thousands of cuts and timed color on a light table.

Kodak, the only company still coating motion-picture film, had one profitable year between 2004 and 2012, then went into bankruptcy after its motion-picture film sales fell 96 percent in a decade.

What saved that link was not a market recovery but buyers constructing a floor: in 2015, the six major studios, lobbied by Nolan, Tarantino, and Scorsese, signed purchase commitments to keep the coating line running.

The cameras tell the other half of the story. For decades, IMAX maintained a small fleet of film cameras rather than manufacturing new ones. Productions rented them, insured them, occasionally wrecked them.

IMAX had been developing a replacement for two years when, in the spring of 2024, Nolan set a deadline: shooting The Odyssey entirely on film required a quieter, lighter camera.

A year later, the company had built four, the first new IMAX film cameras in a generation: carbon-fiber bodies, redesigned movement, quiet enough that the cinematographer compared the old model to an unhinged lawn mower and the new one to a sewing machine.

The last link in this chain is the one with a pulse. Literally, a 15/70 projector is not a device you switch on. It needs a projectionist who can thread a horizontal rolling loop, service a xenon lamp that technicians handle in protective gear because it can explode, and break down a 250-kilogram platter of film without scratching a $36,000 print.

Multiplexes eliminated the projectionist as a craft two decades ago, and digital projection runs on a playlist.

So every retrofitted projector also requires IMAX to reconstitute a labor pool the industry deliberately dismantled, either by pulling veterans out of retirement or by apprenticing new people on machines with no manuals.

The equipment was durable, but, as usual, the knowledge wrapped around it decays much faster.

The defense world, which lives with this problem permanently, has a name for it: diminishing manufacturing sources, the slow-motion loss of the ability to buy what you once bought routinely.

In 2010, the option to keep any of these links alive looked worthless, and nobody prices a worthless option.

In 2026, the expired options are the scarcest assets in the exhibition business.

Seven Weeks of Line, Stored in Fandango

Let’s run the arithmetic: 41 rooms, call it 400 seats each, five or six shows a day if you program one at 2 a.m. On the order of 100,000 seats a day worldwide, against $50 million of advance sales, the largest presale in IMAX history, sold starting a full year before release.

Demand is a multiple of the capacity. And the rooms cannot cover for each other: 25 sit in the United States, the rest are scattered from London to Melbourne to Prague, so for a fan in Miami, the relevant number of rooms is zero.

A theater has three ways to absorb a demand shock: more capacity, inventory, or time.

IMAX cannot add capacity for all the reasons above. A movie seat cannot be inventoried; the empty seat at the 11 p.m. show is worthless by midnight, which leaves time, and time alone, to absorb the entire mismatch. That is what a seven-week sold-out horizon actually is. The line for this movie is seven weeks long. It is stored in Fandango rather than on the sidewalk.

And because tickets stayed at roughly $30 instead of what the market would bear, the queue charges its members in currencies other than money.

The committed pay in planning, booking a Tuesday seven weeks out.

The desperate pay in sleep, which is why the 2 a.m. slot, capacity worth nothing in any normal week, now sells out; when even 2 a.m. clears, the system has no slack dimension left.

The vigilant pay attention, camping on refresh buttons for dropped seats.

And the rich pay scalpers, because a $1,000 resale against a $30 face price measures everything the posted price refuses to collect, rent flowing to arbitrageurs instead of to the people who made the movie and rebuilt the projectors.

The studios could auction those seats and keep the rent themselves; they will not touch it, because the $30 ticket to the impossible screening is part of the story audiences are telling themselves.

A sold-out Saturday matinee tells you a movie is a hit, a sold-out 2 a.m. Tuesday tells you the system has failed to clear and cannot be cleared.

What the Dead Line Is Worth

The Odyssey opened on 885 IMAX screens worldwide; 41 of them are film rooms, and the other 844 are digital and laser:

The film network produced 2.4 percent of the movie’s opening gross. As a revenue line, 15/70 is a rounding error on a $250 million production that out-grossed its budget in three days.

But now look at what those 41 rooms generated.

A year of coverage about tickets selling out in minutes, reporters filing dispatches from 2 a.m. screenings, $1,000 resale headlines in the New York tabloids, and a $50 million presale record built substantially on the fear of missing the version Nolan intended.

The 41 rooms are the marketing engine for the 844.

Every person who turned away from a film screening bought a laser screening instead, pre-sold on the premise that this movie is an event. The company’s own statement had an edge of disbelief: “We’ve never seen demand for IMAX Film on this scale.”

Scarcity converts a screening into a pilgrimage, and a pilgrimage does work for a brand that no marketing budget can buy.

The luxury industry has run this play for decades; Hermes could clear the Birkin waitlist tomorrow and never will.

The difference is that Hermes chose its constraint.

IMAX inherited one from its own graveyard.

Nobody at IMAX in 2007 killed the film line to manufacture mystique in 2026; they killed it because digital was better business, and it was.

It is a simple supply chain failure that aged into a luxury strategy, and the company now has to decide, by design, what it once decided by default: how scarce this format should be.

That is the trap hiding inside the retrofit program.

Some of the demand for these 41 rooms is driven by image quality, and some by difficulty.

I don’t think anyone knows the split.

Every projector IMAX hauls out of a museum basement and rebuilds adds measurable revenue, while subtracting an immeasurable amount of specialness.

There is some number of film rooms, maybe 60, maybe 200, at which the 2 a.m. sellout stops happening, the resale premium collapses, and 15/70 becomes what it was in 2005: a format with a small loyal audience and no queue.

Final Words

So should IMAX rebuild the machines?

The answer is no, at least not past the point where the queue disappears.

The rule flows directly from the case. Capacity expansion is usually the safest investment in operations.

Here, it is the only investment that can destroy the asset it expands, because some unknowable share of this demand is demand for the difficulty.

Mystique is the input that the other 844 screens are monetizing.

The right number of film rooms is the largest number at which the 2 a.m. shows still sell out, and my guess is that number is closer to 60 than to 200.

Rebuild to the edge of the sellout, not one room past it, and put the real energy into the thing that actually leaks value: pricing. Auction in the first month. The scalpers are already running the auction; the only open question is who keeps the proceeds.

The main lesson here is, yet again, about how capabilities die: quietly, rationally, one sensible capital review at a time.

You know my mantra: Tell me what constrains you, and I’ll tell you where your strategy should focus.

But this time we can also run the sentence backward: tell me what you are shutting down, and I’ll tell you what will constrain you in twenty years.