Contacts are ranked, not identified. Nothing in an echo names a fish. The rail lists the family first because that is the answer worth trusting — length and depth are shared across a whole family, so picking one species out of 39 is mostly luck. Off the echo, length is inferred from echo strength corrected for range, which is a weak proxy; a contact from the volume says modelled instead, because there the length is known rather than guessed. The season sets the water temperature, which is the strongest clue there is. Seasons & limits are Michigan orientation figures only — check the DNR guide before anything goes in the cooler.

Three pictures share the glass. 3D volume is the two-node scene from the 3D lab, looking down into the cone. Side look is the same scene from the beam, so you see the water column the way the lab's side view does. Both are the real volume, not a drawing of it, which is why the depth tape agrees: the scene is rescaled to whatever the sounder reports. Pinch, wheel, or the Zoom slider move the camera in; the boat stays framed. Fish are modelled, so the strip says VOLUME SIM or SIDE LOOK, and on a live feed they never replace a real echo. With either 3D look up the glass drops the compass and the speed block: the scene is the picture. Drone lowers the towed second sounder; it starts stowed.

Plan 2D is the cone from above, straight off the pings at /api/frame. Rings are depth. Zoom on the plan moves in on the strongest contact rather than shortening the range.

Bed and depth come off the echo either way. Black is transparent on AR glasses, so move Sun glare up to see why brightness matters on open water.