Simsense

Why a master is worth stealing, why a copy isn't, and how a thirty-second chip names the body that made it.

Simsense is not video. Video records what a camera pointed at. Simsense records what a nervous system did — the muscle firing, the balance, the pressure on the skin, the adrenaline, the intention arriving a fraction of a second before the movement. Play it back and you don't watch somebody run down a corridor. You run down the corridor, in their body, with their fear.

That is the whole reason it was worth inventing, it was invented here in Hollywood by people trying to make a better picture, and it is still made the way film was made: on lots, with crews, with a performer wearing a rig.

Two things govern half the jobs in this city, and it's worth being blunt about which is which. Copying a sim is trivial. Forging one is impossible.

Copying is trivial

It's data. A master duplicates bit for bit, as many times as anybody likes, and so does everything derived from it. There is no version of this where a recording is a unique physical object.

Which means nobody has ever technically held the only copy of anything. Exclusivity in this trade is a promise, and promises are what actually get bought and sold — worth remembering every single time somebody pays a great deal of money for the only copy of something.

Forging is impossible

Nerve conduction isn't uniform between people. Signal speed varies, so does the distance it travels, so the delay between intention and sensation is individual and consistent, limb by limb. A recording carries that pattern at full resolution. It's called the print, and street usage is exactly that: does it carry a print?

You can edit content. You cannot manufacture a print, and every attempt fails the same way: it is too consistent. Real nerves are noisy, the noise drifts as the body tires, and the drift across a recording correlates with what the body was doing in it. A fake either has no drift or has drift that doesn't match the content. It takes a lab an afternoon.

So a recording proves three things nothing else in this world proves together: that a specific body made it, that the body was alive and working while it was made, and that nobody has touched it since.

Raw and mapped

The other distinction, and the one the industry runs on. It isn't about copying — both of these copy freely.

Raw is what the rig wrote, at full bandwidth, exactly as one nervous system produced it. It's close to unplayable by anyone else: your neurology isn't the performer's, and raw fed straight into a stranger's head is overwhelming, wrong, and in the wrong circumstances medically dangerous. Master means raw.

Mapped is the playable edition — raw put onto a standard response curve, compressed, normalised and made safe for anybody's head. The mapping is one-way. Everything that made the recording that person's is precisely what gets averaged off to make it wearable by everyone.

Raw master Mapped edition
Playable by a stranger No Yes
Duplicates freely Yes Yes
Carries the full print Yes Reduced — enough to identify the wearer, not enough to authenticate the recording
Proves it hasn't been edited Yes No
Can be turned into the other Yes, by mapping No. You cannot un-map

So raw is the only thing that is evidence, and it's also the only thing that is proof against you. That's why masters get stolen, why they get destroyed, and why a studio will spend more defending a vault than a set.

Then why steal a master at all?

Because possession here is an inventory problem rather than an information one, and that catches people out.

A production's raw masters live on logged cartridges — one per shooting day, checked in and out of a vault, initialled. Getting the data is easy. Getting it so that nobody knows you have it means the slot in that vault cannot be empty, and what goes in it has to carry a plausible label and play as something unremarkable.

Which is a forgery, and forgeries fail. Not immediately, and not to anybody reviewing dailies — but the substitute's print won't match the performer named on the call sheet, and a lab finds that in an afternoon. So the real question on a job like that is never whether the swap works. It's how long before anybody has a reason to check.

How you identify the person

Two layers, and they cost wildly different amounts.

One: the body, free, instantly. Anyone who plays a sim learns things about the wearer without trying, because they are briefly inside them. Height and reach, from where the world sits. Handedness. Old joints or young ones. Corrected vision. Whether they were in pain, and where. Whether they were cybered — a chromed limb is conspicuous by its silence, because it reports no proprioceptive noise at all. None of this needs equipment. It's why performers are protective and why sim actors are strange people.

Two: the print, which needs something to match against. A print is an identity, but it is not a name. It's a match, and a match needs a reference. Those exist in more places than people expect:

No reference on file and the print names nobody. This is the limit that matters: it's why identifying a stranger from a chip is an information-brokerage problem rather than a technical one, and why the answer usually costs money rather than time.

And it can't tell you when. A print says whose; only the content says when. A chip is proof that somebody was alive when it was recorded, which is not the same as proof that they are alive now — and a great many desperate people have paid a great deal of money to avoid noticing the difference.

What the crew knows