VII. The Layers, Use, and Direction-Taking of the Compass (Luopan)

1. The Luopan Is Not a Fortune Detector

The basic function of the luopan is to fix a bearing and a position. The face of the plate has many rings — commonly the heaven pool (tianchi), the pre- and post-heaven trigrams, the Twenty-Four Mountains, rings related to the Three Cycles or San He, degree divisions (fenjin), and lunar lodges (xiudu), among others. Different schools' compasses carry different contents, and more rings does not by itself mean greater accuracy. In the basic survey of a commercial shop, steadying the needle, reading the bearing accurately, and fixing the Twenty-Four Mountains correctly matters more than interpreting dozens of rings at once.

The needle at the center of the heaven pool is affected by the earth's magnetic field and by nearby metal. Before use, check that the plate body is level and the needle turns freely without sticking, and that the red line is aligned; while measuring, keep the plate horizontal and let the crosshairs run parallel to the building's reference line. If the reader is wearing a magnetic watch strap, carrying an earbud case, or a large bunch of keys, these should be set aside for the moment. A shop's roller shutter, steel structure, freezer compressor, and electrical distribution box are all common sources of interference.

2. The Three-Stage Method of Taking a Direction

The first stage measures the building's or shopfront's reference line from a relatively open spot outside the shop, as far as possible from the door frame and any vehicles, taking three consecutive readings. The second stage measures the actual direction of the doorway near the threshold, but takes readings separately from inside the door, in the middle of the door, and outside the door, observing any magnetic interference. The third stage moves inside the shop and re-measures along the main axis or against the parallel side walls. If the three stages agree, confidence rises; if they differ noticeably, record both the difference and the metal sources found on site.

Never record just a single whole number each time. It is advisable to record "the range of readings, the average trend, the instrument, the position, the time, and the environment." For example, if the exterior reading is 181 to 183 degrees, the reading by the door jumps between 176 and 188 degrees, and the interior axis reads 182 degrees, one can judge that the reading by the door is disturbed, and take the exterior and interior-axis readings as primary. If the true north given on the building drawings differs from the magnetic compass, one must also state whether magnetic north or true north is being used. The traditional luopan generally works from magnetic north, while modern drafting commonly uses true north, and the two must never be mixed without a label.

3. Measuring the Door's Direction, the Building's Direction, and the Shop's Direction

For a freestanding street shop, one can record both the direction of the whole building and the direction of this shop's own door at the same time. If the two agree, the facing is relatively clear; if the building's sitting-facing and the shop's door open at an angle to each other, one should build a two-level model of "the wider surroundings — the shop's qi mouth" separately. A mall unit further requires recording the mall's main entrance, the incoming route of the escalator on that floor, the shop's own opening, and the shop's interior main axis. Practitioners differ on whether a Xuankong house chart should be plotted from the building's direction or the shop's own direction, and this should be stated in the report.

This article recommends: use the building's direction when studying the long-term house period of the entire commercial complex; when studying the door, path, and register inside a leased unit, take the unit's actual qi-receiving direction as the primary reference, while still keeping the building-level chart's outer influence on record. Do not present the same star chart as the sole chart for both the entire building and the small unit at once. Keeping the levels separate is what allows the reasoning to be checked.

4. A Checklist of Common Compass Errors

The power supply inside an illuminated shop sign, a hidden steel beam, an automatic door's motor, a metal counter, shelving, a phone's magnetic case, an electric scooter, and underground pipelines can all deflect the needle. The steel-reinforced environment of a high-rise building can also make indoor readings unstable. When something unusual appears, do not keep turning the compass chasing the needle — instead, move away from the interference source, look for a parallel reference line, and judge from the shared trend across multiple points.

The surveyor's own stance can also introduce error. Failing to align the crosshairs with the wall line, failing to keep the plate level, entering a sitting-degree reading into software as a facing-degree reading, and adding or omitting the 180-degree offset for the opposite direction are all common problems. When a report includes site photographs, the photos should show the measurement point, the reference line, and the compass's orientation — while still respecting privacy and any photography rules of the mall.

5. Establishing the Pivot: Where the Nine Palaces Unfold From

Establishing the pivot (liji) means fixing the central base point of the space, from which directions are laid onto the floor plan. For a regular rectangular shop, the geometric center can be taken directly; for a unit with a missing corner or a protrusion, one should first restore the main enclosing outline, then calculate the center of gravity and judge whether the protrusion or missing corner counts as effective interior space. A narrow, elongated attached storeroom, an outdoor platform, or a public corridor should generally not be weighted equally with the main sales floor in this calculation, but if they take up a large proportion of the space and are used for business over the long term, they need to be addressed separately.

A duplex shop should first have its overall pivot fixed for the whole shop, and only then have sub-pivots set for each floor. The overall pivot governs the overall palace positions; the sub-pivots govern the local palace placement of the door, stairs, stove, and counter on each floor. One must not directly project the ground floor's center straight up onto a second floor of an entirely different shape, nor should every small room be given its own independent house chart while ignoring the whole. A common practice is "one house, one master chart, with each floor and room given an auxiliary set of Nine Palaces."

Once the pivot is established, the door, secondary door, stairs, elevator, escalator, stove, toilet, register, main counter, storeroom, and spirit tablet should all be marked accurately onto the plan before the flying stars are even discussed. If the floor plan's proportions are distorted, any object near a palace boundary should be marked as straddling two palaces, rather than forced into one palace for convenience. One may judge which palace an item belongs to by the center point of a door, or observe a wide door by the proportion it covers in each palace — what matters is that the convention used stays consistent throughout.

6. Data for Determining the Period, and Levels of Evidence

Plotting a Sanyuan Xuankong house-period chart requires clarity on when the building or shop was completed, when it underwent major renovation, and when it was reopened for use. Tradition itself is not unanimous on what counts as "changing period": some weight the topping-out of construction, some weight reoccupation after a long vacancy, some weight major renovation such as changing a door, replacing a roof, or breaking ground. A commercial tenant often knows only the date of signing the lease, not the building's history, so evidence should be graded.

First-tier evidence includes completion records, property-management archives, and approval documents; second-tier evidence includes the owner's contract and clear drawings of past renovations; third-tier evidence is oral accounts and inferences drawn from photographs. Where evidence is insufficient, one may tentatively plot two candidate period charts and use documented door-and-path changes and observable events over the years as a traditional cross-check — but one must never fabricate events to "verify the chart." Where it cannot be determined, state this plainly as an item of uncertainty, rather than presenting the chart as absolute.

7. The Format of a Direction-Measurement Record Sheet

A raw record sheet that can be independently recomputed should contain at least the following fields: project number, date of measurement, weather, surveyor, instrument type, note on magnetic declination, exterior reference point, point outside the door, point in the middle of the door, point inside the door, interior-axis point, three readings at each point, sources of interference, the final degree adopted, the sitting mountain, the facing mountain, the judgment on slant facing, the method used to establish the pivot, the floor-plan version, the evidence for the determined period, and remarks.

The record sheet is not a mere formality. Months later, when the shop has replaced a metal door, moved the register, or sealed off a side door, the original data can help determine whether the configuration has actually changed or whether it is only a lapse of memory. If a traditional survey cannot be recomputed, it easily degenerates into on-the-spot patter.