Every figure we publish in this section is one of the measurements defined below. Where a number is not something PulseMesh measures — audience size, for instance — we say where it came from instead.
Event audience
The number of people physically attending an event.
PulseMesh does not measure this. It comes from the event organizers, who count it the way they always have: ticketing, registration, gate counts, or their own attendance estimate. When we publish an audience figure we attribute it to the organizer and, where the figure is an estimate, we say "approximately."
Peak simultaneous listening devices
The maximum number of PulseMesh endpoints simultaneously active during a stated window — usually the show itself.
This is a device count, measured by the platform. Each endpoint is one app instance or one browser session on one device. If a device drops off the network and returns, it is the same endpoint; it does not raise the peak twice.
Unique devices
The number of distinct endpoints seen over a stated period, deduplicated, so a device that connects on three different nights counts once.
Unique devices is always reported with the period it covers. Without a period the number means nothing — a season figure and a single-night figure are not the same measurement.
Listener session
One connection that meets the service's qualification threshold: sustained playback, not a bounce.
A device that opens the experience and leaves a few seconds later is not a session. The threshold exists so that session counts describe people who actually listened to a show, and it is applied consistently across every figure we publish, including historical ones.
Active display
A display that recorded at least one qualified listener session within the measurement window.
A display that exists in an account but never had a qualified session in that window is not counted as active for it. As with unique devices, the window is always stated alongside the number.
We never infer people from devices
Devices are not people, and we do not convert between them.
Group listening is common and, at large events, encouraged: attendees share a handset, set one phone down for a circle of friends, or pair to a Bluetooth speaker so a whole camp hears the same soundtrack. That makes any people-per-device ratio a guess, and multiplying a device count by an assumed ratio would produce a number nobody measured.
So where group listening occurs, we report both figures independently — the audience count from the organizer, and the device count from the platform — and we leave them un-multiplied. If a page here shows two numbers side by side, they are two separate measurements, not a fraction waiting to be resolved.
When numbers change
Definitions can be refined. When one changes in a way that affects a published figure, we restate the figure and note that it was restated, rather than quietly leaving two incompatible numbers on the site.