Plants, stages, harvest window and DLI

How a strain profile becomes Want bands, how expected stage and the harvest window are computed, what the DLI estimate is made of, and the plant · probe · assignment model.

A plant is a roster row with a strain profile, sprout date, medium and recipe. A probe is hardware with a slot number and an idle home. Assignment links the two, and is separate from the idle home: detaching a probe does not move it, and a soil test uses the home without changing the assignment.

The strain profile (online API › on-Pi corpus › curated YAML, each tagged with its source) carries climate and root targets per stage family (seedling, veg, flower). The brain picks the family from the plant's expected stage and writes the tent's Want bands from it when you Apply to tent. The 2×4 in Follow plants mode re-derives its targets from its plants roughly every 12 hours and on roster changes.

Days since sprout, plus the flip date for photoperiod strains, mapped onto the stage table. With Catalog-scaled stage model on, the flower sub-stages stretch or shrink to the strain's catalog flowering time; without it, the fixed table applies. Autoflowers count from sprout. The flip date is stamped when the plant crosses veg → flower and can be edited on the plant panel.

Per plant: expected finish, days to finish and flower day, from the flip date (operator-entered › journal flip entry › calendar stand-in, always labelled which) plus the catalog flowering span. Per tent: a harvest-window sensor from its plants. These are expectations for planning; the plant decides when it is done.

Running and projected daily light integral per tent from: the bound fixture's PPFD layers (maker data) × the canopy distance (Settings › Integrations; left empty, the maker's recommended hang height for the stage is used and labelled an estimate) × the dim percent × the photoperiod hours. A tent with more than one lamp estimates each lamp at its own dimmer and hours and combines them over the lamps' map footprints; spill between lamps is not modelled. A generic stage band gives a verdict (under, in band, over). It is labelled maker map everywhere because it is not a PAR measurement; a light-curve calibration (Kit › Calibrate) refines the brightness response but not the map.

The generic flower DLI band depends on whether the room gets added CO₂ (Settings › Light › CO₂ supplemented, off by default). Off: 30–40 mol/m²/d through bulk flower and 40–50 in late flower — what a canopy can use at ambient CO₂. On: 35–55, which needs about 1270 PPFD over 12 h and only pays with added CO₂. The Light desk verdict, the reach-target line and the run report use the band that applies; seedling and veg bands are unchanged.

Guide contents  ·  How it works  ·  Troubleshooting  ·  Digital Stealth Care

DSC-Brain user guide · Plants, stages, harvest window and DLI · this page is edited on the site; the Notion original was retired after the mirror