Inside AMN Studio: staging a results night in augmented reality
How the studio team rehearses, cues and renders a results wall that has to be correct on the first take.
- The results wall is driven by the same feed that powers the website widgets.
- Rehearsals run against a replay of the previous cycle's live data.
- Presenter control allows any race to be pulled to air in under four seconds.
Machine-generated from the reporting on this page and reviewed by an AMN editor before publication.
The results wall an audience sees on election night is driven by exactly the same feed as the widgets on this website. There is no separate broadcast dataset, and that is the entire point of the architecture.
Rehearsals run against a full replay of the previous cycle's live data, at real speed, so that the team practises the actual sequence of calls rather than an idealised one.
Presenter control is a single touch surface. Any race, any comparison, any historical overlay can be pulled to air in under four seconds, which is roughly the length of a sentence.
The augmented reality layer is rendered in the same engine that produces the on-air lower thirds, which keeps typography and colour identical across every surface the network publishes to.
None of this is unusual in isolation. Running it from one feed, for both web and broadcast, still is.