This is the lesson the other five were building towards. Up to now you have been describing your garden to the editor. From here the editor starts describing your garden back to you, and the things it says are frequently not what you expected. There are three overlays and one control, and between them they answer the questions that actually decide whether a planting plan works. Where do shadows fall at this hour on this date. Which parts of the plot get enough hours of direct sun across a whole window rather than at one instant. And where does the ground run warm or cold, which matters far more for an early sowing than most people allow for. Everything you did in Lesson one is what makes these three overlays true rather than decorative, and if the compass is wrong then so is all of this.

Three objectives. First, you will be able to turn on and read each of the three overlays: shadows, which show where shade falls at the exact date and time on the sun bar; shade contours, which map how many hours of direct sun each part of the plot gets across a window of days; and heat contours, which show where the ground runs warm and cold. Second, you will be able to widen the question from a moment to a season using the Days control, and explain why sun hours over a window is a more useful number than sun position at an instant. Third, you will be able to act on what you find, by matching a bed against a species' sun need, moving a planting that is in the wrong place, or changing the height of a structure that turns out to be the problem.

The distinction to hold onto is between sun position and sun hours. Sun position is where the sun is at one instant, and the shadows overlay draws exactly that: drag the time and watch shade sweep across the plot. It is vivid and it is useful for catching a specific problem, like the shed that takes the morning off the salad bed. But no plant lives at an instant. What a plant experiences is a total: how many hours of direct light it received across a day, and across the weeks that matter to it. That is what the shade contour overlay computes, by integrating the sun's path over the date window you set with the Days control. A bed can look sunny at noon on the day you happen to be looking and still be a poor bed, and only the hours number will tell you.

Three overlays, three jobs, and they are meant to be used in sequence rather than all at once. Shadows is the diagnostic one: turn it on, drag the time across a day, and you see which structure is responsible for what. It answers the question why. Shade contours is the decision one: set a date window with the Days control and read total direct-sun hours across the plot as a map, then compare that against what your species want. It answers the question whether. Heat contours is the one people forget and then find genuinely useful, because a south-facing wall banks warmth and a low damp corner does not, and that difference decides an early sowing more often than raw light does. Turn them on one at a time. All three together is a pretty picture you cannot read.

Five steps to a light audit of a garden you have planted. First, set the sun bar to a date that actually matters: the week you plan to sow, or the height of summer when everything is competing. Second, turn on shadows and drag the time from morning to evening, watching which structures take which beds and when. Third, turn shadows off and shade contours on, set the Days control to a window rather than a single day, and read the map of total sun hours. Fourth, compare that map against your plantings: the editor knows each species' sun need and will tell you where a plant is sitting somewhere shadier or sunnier than it wants. Fifth, act. Move the planting, lower or move the structure, or change the species to something that suits the light you actually have rather than the light you wanted.

Go back to the three answers you gave in Lesson one, before you had drawn anything. Which way does your garden face, where does shade come from, and what hour does the sun leave the bed you care about. Now put the model next to them. Most people find at least one of the three was wrong, and it is usually the last one, because we remember the hour it got dark rather than the hour the light left that particular corner. Ask yourself what you would have planted based on the guess, and what the model says instead. And then ask the harder question: is there anything you could change about the site itself, a fence panel lowered, a shrub taken out, a bed moved two metres, that would buy back the hours you are short.

Three things. Sun position and sun hours are different questions, and only the second one is about what a plant experiences, so widen the Days control before you conclude anything about a bed. Use the overlays in sequence and one at a time: shadows to find out why, contours to decide whether, heat to work out when it is warm enough to start. And the whole thing rests on the compass bearing and latitude you set in the first ten minutes, so if an overlay tells you something that feels impossible, check the site before you doubt the model. One lesson left. You have a planned, planted, audited garden, and now we get it out of the browser and into a form you can print, share, and keep.