The catalogue of unheard sounds

The mug began to shiver before Lin saw the meter move.

It sat on a felt pad at the far end of the archive table, safely away from the sample drawers. A pale ring of tea climbed one side of the cup and fell. On the wall display, a caption appeared: [LOW PULSE, 4.2 SECONDS]. Lin felt the pulse through her fingertips where they rested on the table.

Across the room, Dr. Saye said something. Lin turned the caption camera toward him.

“I said, is that the last one?”

“Last of the first twelve,” she told him.

The display replaced the waveform with a magnified image of the probe’s microphone membrane. It looked like paper, though its fibers carried a fine conductive mesh. A split ran inward from the rim. At its tip, the fibers crossed in a thin, irregular line.

Lin opened drawer twelve. The raised caption on its lip read SAMPLE 12: RESTORED RESPONSE. Inside lay a paper facsimile of the membrane image. She set it beside the image for sample four, whose split was wider.

Saye came closer. He wore his coat over his evening clothes. He had been called back to the institute for a finding he could announce before morning.

“The interval repeats,” he said, pointing at the waveform. “The music group has named it the return.”

Lin watched the caption settle, then looked again at the two membranes. “What returned?”

“The opening phrase. You saw their report.”

“I saw the report.”

Nine months earlier, the probe Wren had settled beside a hollow structure on a moon with a dense, restless atmosphere. Its first transmissions had been ordinary: wind, grit striking its housing, the faint rattle of its own cooling vanes. Then came twelve sounds from somewhere close to its open microphone. Each had left a different shape in the membrane. The probe sent a pressure trace and an image after every sound.

By the third arrival, people were sharing melodies made from the traces. By the eighth, a chamber orchestra had asked for the complete sequence.

Saye slid a thin authorization sheet onto the table. The proposed public statement filled its upper half. Lin knew where her name would go. As chief archivist, she would certify that the sequence and its order were authentic. A second signature, from the interpretation group, would appear below the words an intentional musical greeting.

“The minister wants to speak at eight,” Saye said. “We have an hour.”

Lin lined up samples four, seven, and twelve. The split had changed shape between them. On seven, its edges curled toward each other. On twelve, they lay together, one edge tucked beneath the other. A narrow line continued past the original tip, as if the membrane had flexed there long enough to remember it.

“Look at seven.”

Saye bent over the prints. “I am.”

“The mark isn’t part of the pressure trace. It’s a seam.”

“It resembles one.”

“The split disappears after it forms. And the probe’s calibration click comes back in sample nine.”

“The instrument may have stabilized.”

“That’s what I’m saying.”

Saye straightened. His caption lingered on the wall a moment after he spoke. “You think someone repaired our microphone.”

“I think we should test whether it was repaired.”

He looked at the authorization sheet. “This is your appointment, Lin. Records director. I put your name forward because you don’t let a beautiful result outrun the material.”

She ran a finger across the raised number on drawer seven. She wanted the job. The archive had drawers whose captions she could read by touch, and three newer rooms where labels were still printed flat behind glass. She had a plan for them, costed and ready for whoever took over.

“Then give me the hour,” she said.

“You have forty minutes. After that I need your decision.”

Saye left the sheet on the table.

Lin called the test workshop. While she waited, she turned the sound on again at the level she used for work. A rough flutter reached her left ear. The rest came through the tabletop and the meter’s bright rising bars. She stopped playback at the instant the split first began to close.

---

Tess arrived carrying a case of membranes and the expression she brought to any request marked urgent by someone outside the workshop.

“You need a singer or a torn microphone?” she asked.

“Torn microphone.”

“Good. We’re short on singers.”

Lin showed her the images. Tess spread them beneath the work light without touching the waveform charts.

“Could be material creep,” Tess said. “Could be a bad image from a sensor that was shaking. You’re calling it a repair because a seam looks like one.”

“I’m calling it a possible repair because the calibration click returns.”

“Better.” Tess opened her case. “What do you want to put at risk?”

“One witness membrane.”

“They’re cheaper than the probe.”

They worked in a room lined with felt pads to keep loose equipment from rattling into the measurements. Tess fitted a new membrane into the bench microphone and made a split matching the one in sample four. Under the magnifier, its edges stood apart. Lin checked the vibration meter while Tess spread conductive paste on the contacts that held the test frame. A gray smear remained on her thumb.

“What happens if we play the recorded sound at it?” Lin asked.

“We find out what that recording does here. That’s all.”

Tess set Lin’s mug on the test table, beyond the frame. “Your early warning device.”

“It warned you that you were leaning on the table.”

“It’s versatile.”

Lin started the playback. The mug shivered. The membrane snapped inward, then sprang back with a larger split. The meter flashed red. Tess stopped the driver and lifted the frame toward the light.

“There’s your concert,” she said. “It made a hole.”

Lin compared it with the image from Wren. In the probe image, the split’s tip had hardly moved. The edges had changed position along its length.

“The recording is from the damaged microphone,” she said.

“Of course it is.”

“The probe corrected its gain as the split closed. We played that correction back as if it were pressure.”

Tess set down the ruined frame. “Maybe. What do the images tell us that the trace doesn’t?”

They laid all twelve facsimiles in order across the table. Lin took them from their drawers one at a time, reading each raised caption before passing it over. Tess measured the changing split with a transparent grid. At sample five, one edge had lifted. At six, a crease ran from the rim toward the tip. At seven, the lifted edge had begun to settle beneath the other. The last four images showed the seam holding while new pulses struck it.

Tess tapped sample six. “This bit opens the split.”

“Yes.”

“If you wanted it shut, why open it?”

“To move the edge without buckling it?”

“That’s your romantic answer.” Tess angled the image toward the light. “My answer is that the first edge was stuck. Whatever made this sound had to free it.”

She adjusted the driver to follow the membrane’s measured movement rather than the recorded volume. On the second witness membrane, the lifted edge folded over itself. The split closed, but the fold made a ridge that caught the next pulse. It opened again.

Tess swore softly. Lin’s captions supplied [INDECIPHERABLE SPEECH].

“I caught that much,” Lin said.

“Then you know it failed.”

Tess reached for a third membrane. Lin stopped her. On the probe image for sample four, a tiny dark fleck sat beside the split. It remained there through five and vanished beneath the edge in six.

“The fleck moves first,” Lin said.

Tess peered at it. “That could be dust.”

“Dust would do.”

They had treated the split as a clean cut. On Wren, grit had held one side away from the other. Tess placed a grain from the workshop’s calibration jar near the third split. This time the first movement shifted it. The edge rose, passed over it, and settled. The membrane held through the next pulse.

Tess let the driver run the probe’s calibration click. The meter showed a clean, narrow response. Lin heard a faint tick and saw Tess glance up at the same moment.

“Again,” Tess said.

The click returned at nearly the same strength. Tess made it a third time, then switched off the driver. Their mug sat still beside the frame.

“Something moved those edges into a working position,” Tess said. “That much we can say.”

Lin looked at the authorization sheet she had brought from the archive. Its bright white corner stuck out beneath the sample prints.

“Can we send something back?”

Tess’s hand stayed on the frame. “A sound?”

“A repair. Wren has a spare calibration strip in its maintenance cassette. It has a factory slot. We can demonstrate what we’ve learned on that.”

“We’ve learned how to copy an effect. One successful strip on a bench.”

“We can show our attempt.”

Tess considered the probe images again. “We can also show where our attempt fails.”

On the bench, their seam held the calibration click, but a slower pulse made one end lift. Tess marked that end on the frame with a small dot of conductive paste. Lin photographed the strip before it was disturbed.

Saye called while they were doing it. His face appeared beside the live meter.

“I need the authorization.”

Lin placed the bench images where his camera could see them. “Our first replication enlarged the tear. The next two taught us what moved. This one restored the calibration response. It still has a weak end.”

His eyes went to the clock. “You can certify the catalogue without certifying the interpretation.”

“Yes. I’m doing that.”

“The statement goes out without the greeting.”

“The greeting is unsupported.”

For a moment he said nothing. Behind his shoulder, people moved through the press room, setting water on a lectern.

“I’ll amend the release,” Saye said. “The appointment board meets Friday. They will ask why you stopped this.”

Lin looked at Tess’s marked membrane. “Give them the images.”

When the call ended, Tess said, “You’re allowed to be unhappy about that.”

“I am.”

“Good. I’d hate to do this with a saint.”

---

Signals took forty-three minutes each way between the institute and Wren. Its maintenance cassette held strips made with known slots so the microphone could check itself after a dust storm. Ground control had never needed to open it.

Lin and Tess watched the cassette slide out on the probe’s camera. The hollow structure stood beyond it, dark against a sky full of suspended grit. Its nearest opening was turned toward Wren. Nothing moved there that the camera could resolve.

They sent the bench sequence through the probe’s small calibration driver. They also sent the images of the torn strip, the closing seam, and the end that lifted again. Wren could display the images on a panel beside the microphone. Lin did not know whether anything near the structure could see the panel. She knew the earlier sounds had arrived at the membrane.

The command log changed to TRANSMITTED. Eighty-six minutes later the playback acknowledgment arrived. Its image showed the factory slot closed along the middle, with one end folded up like their bench attempt.

Tess drank the cold tea from Lin’s mug without asking. Lin let her. They watched a thin line on the meter show the probe’s ordinary tremor. Once, wind pushed grit across its housing, and the caption read [SCATTERED IMPACTS]. The strip in the cassette remained unchanged.

Saye sent a copy of the revised release. It called the catalogue a sequence of sounds and said its meaning was unknown. It mentioned damage to the microphone and an experimental repair at the institute. Lin read it twice. No melody had been given a title.

An hour after the playback acknowledgment, a new pressure trace arrived. It began with a broad, slow movement unlike any in the catalogue. Then came the short motion that had freed the edge on the bench. The image after it showed Wren’s spare strip with its weak end lifted clear of the seam.

“Did they open it?” Lin asked.

Tess was already enlarging the image. “Wait.”

Two more traces arrived. In the first, the lifted end lay flat without the small fold their bench strip had formed. In the second, the calibration click ran through it. Wren sent another image. The seam held from rim to tip.

Tess put their bench photograph beside the probe’s. The difference was narrow but visible. Their strip had closed from the middle outward, trapping a fold at one end. Wren’s had first loosened the full length of the edge.

“They corrected the order,” Tess said. She rubbed the paste from her thumb onto a rag. “Or something did.”

Lin handed her a fresh membrane. Tess made another split. They followed the new movements in the order Wren had received them: the broad release, the brief lift, the closing pulse. The edge settled flat. The calibration click returned. When Tess sent the slower pulse that had defeated their earlier seam, it stayed closed.

They ran it once more on another strip. Lin watched the meter and felt the slow pulse through the table. Neither seam lifted.

Tess sat back. “Now we have a correction.”

Lin sent Wren a short confirmation: the calibration click, then a fresh image of the repaired bench strip. The probe could show it beside its own. She did not add a greeting. There was no agreed sound for one.

By morning, the institute had a second release to prepare. Saye wanted the probe images checked against the archive copies before anyone saw them. Lin carried the new sample sheets back to the drawer room herself. Her mug, rinsed by Tess, waited on a felt pad beside the table.

She made two raised captions. The first read OUR REPAIR ATTEMPT. The second read RECEIVED CORRECTION. She slid each into its drawer, then ran her thumb over the letters to check that none had caught beneath the rim.