What the Mesh Is Made Of
The instrument was sitting in the corner of the workroom when Gu-ship-pal went to get it. He had been working in the building for eleven years and had brought four instruments to it in that time. This one was the fourth. The others were dismantled when they proved insufficient — not failed, just not the right thing — and the wood from the first two had been reused in the workshop's window frame, which still stood.
He picked up the instrument and set it on the worktable.
The mesh had taken him six months to think about and three weeks to build.
The thinking had begun after Zone 2. Zone 2 had produced a clear sympathetic response in the southeast corner of the zone — the mesh had vibrated at a specific frequency — but he had not been able to determine whether the response was to the zone's actual depth-field structure or to a resonance interaction between the mesh and the building's ambient infrastructure. The building's acoustic infrastructure was managed by an old Sentinel system, one of the early autonomous building managers, and Sentinel had a known behavior: it made micro-adjustments to corridor air circulation every forty seconds on the quarter-hour. These adjustments were invisible to occupants but measurable if you were running sensitive equipment. He had reviewed the Zone 2 data three times and had not been able to rule out Sentinel's circulation cycle as a confound.
The mesh he had been using was a single-gauge wire, which had the advantage of uniform response across the surface: you always knew what you were measuring. But single-gauge was also vulnerable to environmental noise because there was no differential to compare against. Every vibration looked the same.
He had spent six months thinking about this.
The solution — three gauge wires woven into the same mesh, with each gauge producing a different resonance frequency — came from a technique he had learned from an instrument builder in the city's old industrial quarter who made percussion instruments from reclaimed materials. She had told him that the way to hear what a space was doing was to give it more than one thing to respond to and see which one it chose. He had understood this in the context of percussion. He had not thought to apply it to zone mapping until the Zone 2 problem made the limitation of single-gauge measurement clear.
The three-gauge approach also solved the Sentinel problem. Sentinel's circulation adjustments affected all gauges equally — a uniform pressure shift with no differential. If a zone-junction produced sympathetic resonance, it would preferentially activate the gauge whose frequency matched the junction's depth-field frequency. The differential between the three gauges would separate zone response from infrastructure noise. He had verified this with a test against a corridor near Sentinel's primary air distribution point, watching the three gauges respond identically to the circulation pulse, then diverge when he moved to a zone area with known depth-field activity.
He spent two weeks working out the weaving pattern. The three gauges needed to be distributed across the mesh surface in a way that prevented any one gauge from clustering in the same area — if the heavy gauge was always in the center, the center readings would be biased. He used a diagonal interleaving pattern: each row advanced one step in the gauge sequence. The result, when he held the finished mesh up to the light, looked irregular at first and became obviously systematic when you understood the diagonal.
Three weeks of building. His hands remembered the work.
He ran one finger across the mesh surface. The wires were cold from the workshop air. They would warm from the zone's depth-field when he had been standing in the stairwell long enough.
Before Zone 2, he had described the work, when asked, as zone-documentation methodology. After Zone 2, he had a better answer: it was a way of asking whether a zone had a preference. Not what a zone contained. What it chose to respond to. Single-gauge mesh could not ask this question because it gave the zone only one thing to respond to. Three-gauge mesh gave the zone three things to respond to, at three different frequencies, and the differential between the responses told you not just whether something was happening but what kind of thing.
He was not certain this was correct theory. He was certain the distinction mattered to the design, which was enough.
The logging sheet was in the instrument case. Thirty-five cells: seven horizontal positions across the stairwell, five vertical heights. Each cell would hold a duration value — how long he held the instrument in position — and a response notation: which gauge vibrated, at what amplitude, whether the vibration was continuous or intermittent.
The logging sheet was his own design. The standard zone-mapping format used a grid of single response values, which was appropriate for single-gauge measurement. For three-gauge measurement, he needed a notation that could capture differential response across the three gauges simultaneously. He had developed a notation during Zone 2 and refined it for Zone 3: three characters per cell, one per gauge, with amplitude levels encoded as characters rather than numbers. It fit in the cell without crowding.
He had shown the notation to Nalgeot-Chae, who had said it looked like it was encoding more than it was recording. He had thought about this for a week before deciding she was right, and that this was not a problem — sometimes a notation needed to reserve space for information that didn't exist yet. The three blank characters in the low-response cells were holding room for a refinement he hadn't reached.
He closed the case.
The Zone 3 test was tomorrow. He had done everything he could do tonight. The instrument was correct. The logging sheet was correct. The zone would be there, and Sentinel would be running its circulation cycle, and the three-gauge mesh would separate the two.
He had thought, at the beginning of the Zone 3 work, that the stairwell would be the simplest of the three zones — the most structurally regular, a consistent shaft rising through four floors, no corners, no adjacent rooms with their own depth-field complications. He had changed his mind after the pre-test. The junction identification had confirmed sympathetic resonance at 1.7 meters in the southeast quadrant. The stairwell was not simple. It was concentrated.
He turned off the workroom light and locked the door.
He walked home through the Lend District.
The district was doing its early evening behavior — the practitioner traffic had eased, the residents who used the clinics during clinic hours were already inside, and the streets had the quality of a place that has transitioned from one kind of use to another without quite settling into the new one. He had lived in the Lend District for fourteen years and knew this quality well.
He thought about the pre-test result.
Shoulder height, 1.7 meters. Southeast quadrant. The mesh had responded with the middle gauge — the gauge he had selected for its sensitivity to frequencies in the 180-220 Hz range, which was the range he associated with the Lend District's stairwell depth-fields based on Zone 1 and Zone 2 experience. The response had been continuous, not intermittent, which meant the junction was stable. Not a passing resonance but a feature of the space.
The relay at 2.1 meters had surprised him more than the junction itself. A relay meant the junction was part of a larger depth-field structure that extended above the primary resonance point. He had not found a relay in Zone 1. He had found one in Zone 2, in the upper section, which he had noted but not fully mapped because the Zone 2 work had been primarily about the lower zone. Zone 3 had a relay he had identified before the main test.
He made a mental note to add a row to the logging sheet for the relay.
His apartment was three streets from the building. He arrived at it, unlocked the door, went inside. Made dinner. Did not think about the instrument or the logging sheet or Zone 3 while eating — a practice he had developed because thinking about the work during meals made the meals poor and the thinking no better. After dinner, he thought about the mesh for another hour.
Then he got out of bed — he had been lying there for fifteen minutes without sleeping — and added a note to his device: extend grid to 2.1m, add relay row before entering zone.
He thought about whether Sentinel's circulation record from the Zone 2 date was accessible. Probably; the building management system logged everything to its own archive and old Sentinel systems were not known for data deletion. If he could pull Sentinel's Zone 2 circulation timestamps and compare them to his own Zone 2 data, he might be able to map exactly which cells in the Zone 2 record were contaminated by the circulation pulse and which were clean zone-junction response. He added a second note: pull Sentinel logs Zone 2 if accessible, compare timestamps.
He had been preparing for nine months, counting from when he built the mesh in March. The mesh had been in the workshop corner for nine months, waiting. Tomorrow was the full use.
He did not think it would tell him everything. He thought it would tell him the next thing.
This was, in eleven years of working in buildings with old autonomous infrastructure, the most he expected from any instrument: not the answer, but the question that came after the one he was asking.
The zone was unchanged by any of this.
The mesh had been built in March.
He remembered the March work clearly because March in the Lend District was when the district's acoustic character shifted — the building heating ran less, the thermal expansion of the walls was smaller, and the ambient sound profile of the clinic floors changed in ways that were subtle but measurable. He had been aware, while building, that the instrument he was making would be tested in a different acoustic season than the one in which it was being made. He had decided this was acceptable because the mesh's purpose was not to respond to thermal acoustics but to zone depth-fields — a junction that existed would exist in all thermal conditions, even if its amplitude varied.
He had finished the mesh on a Thursday in late March. Had held it in the workshop and run the basic calibration tests. Had taken it home to check it against his apartment's acoustic signature, which he knew well, and had confirmed that the three-gauge differential was producing distinct response profiles for each gauge rather than converging toward a single response. He had also tested it against his apartment's building management system — a newer Arbiter unit, not a Sentinel — and had confirmed that Arbiter's circulation adjustments were invisible to the mesh's differential: all three gauges moved equally, exactly as expected.
The instrument had been ready by the end of March. Zone 3 had begun in April, with a preliminary survey. The survey had taken three visits. The pre-test had been this week.
The test was tomorrow.
He had been living with the Zone 3 question for nine months — not anxiously, but steadily, the way you live with a problem you trust yourself to eventually answer. The mesh had been in the corner of the workshop for nine months, and he had thought about it during that time: refined the logging sheet notation, reviewed the Zone 2 data three times looking for the Sentinel contamination, talked with two colleagues who had done junction mapping in the Lend District's older buildings and heard nothing that changed his approach.
He was ready. The mesh was ready. The zone had been doing what it was doing for forty years.
Tomorrow he would find out what it preferred.
He fell asleep before midnight. The alarm was set for 7:30. The instrument was in the case, packed and correct, waiting in the corner of the workshop three streets away. .