The boundary the ceiling states and the floor does not
The ceiling is the one surface any two zones share, unbroken and visible from anywhere in the room, which gives it the deciding vote. A cove running continuously across it erases every boundary underneath, however carefully the lower layers were zoned.
There are two honest options. Keep the ceiling continuous and zone the lower layers, where work surfaces and furniture already draw the lines. Or break the ceiling physically, with a drop, a beam or a change of plane the cove can stop against.
The floor almost never does this job: it is covered by furniture, seen at a shallow angle, and a material change there rarely registers as a boundary.
How the transition reads when you walk across it
A boundary is either a hard edge or a gradient, and the two need different geometry. A hard cutoff at a beam reads as deliberate, since there is a physical reason for it; the same cutoff mid-ceiling reads as a run that failed.
Where there is no physical break, overlap the zones. A transition two to three feet wide, with both contributing across it, reads as a gradient rather than a seam. Walk the line before signing off; a boundary that looks right in plan often lands wrong at head height.
Two zones sharing one sightline in open-plan lighting
When both zones sit in one view, the ratio between them decides whether the far one is legible. The eye adapts to the brighter zone and the darker one falls away faster than the numbers suggest.
Ratios worth mocking up rather than trusting on paper: adjacent zones in one view tend to sit around two to one or three to one, and once the near zone is ten times the far one, the far zone reads close to black. Check it from the seat that looks that way.
The boundary that moves when the levels move
A zone boundary is only fixed while the levels are. Drop the kitchen to a fifth of full and the dining table becomes the bright zone, so the boundary jumps toward the kitchen though nothing physical moved.
So every zone needs its own level rather than a shared dimmer, and the boundaries have to be checked at each setting, not just at full output. Zone planning as a process is covered in how zones get set before anything is ordered.
Wiring and control across the boundaries
Separate zones need separate control channels, so read channel count, current per channel and dimming behavior straight off published data for the exact units. Two dimmable labels do not amount to a verified pairing. One multi-channel option to check against its published data is the Hub5 controller from EnoSync; the wider controllers grouped by channel count let you compare.
Zones do not get their own voltages. If the tape reads 24V then the supply feeding it and any controller between them read 24V too, and a 12V zone stays 12V right to its last emitter. Constant-voltage runs come back to the driver in parallel and are never chained one into the next, and where a single supply feeds several zones the arithmetic sits in running several lengths from a single driver. Add the feet, multiply by the published watts per foot, hold at least 20 percent of current headroom above that total, and confirm the guidance and the supply's stated output. Connector fit, cut interval and beam angle are per-product answers, not category ones.
Warmth at the tape and at the driver is normal in service. An aluminum channel helps dissipate heat and gives a zone a defined end, and fitting one is never required. A channel reduces but does not guarantee that dotting disappears, an outcome LED density, diffuser type and channel or profile depth also govern.
A licensed electrician handles every line-voltage connection here, meaning the box behind a wall dimmer or switch, a driver that is hardwired rather than plugged in, the junction box serving it, and any cable inside a wall. That work follows NEC requirements where those govern, local electrical code, and the printed manufacturer instructions, and begins with the breaker shut off and the circuit tested dead. Low-voltage guidance is not permission for it.
