They spent a week preparing and none of it felt like enough.
Kael knew this was a rational assessment rather than anxiety. The Bleed was a different category of environment from the Echo and the difference was not one of degree but of kind.
The Echo was dangerous the way a dark road was dangerous: mostly empty, occasionally not, and survivable if you moved carefully and paid attention. The Bleed was dangerous the way a pressurized system was dangerous: stable until it was not, and when it stopped being stable the problem arrived faster than most people could respond to.
He had four days before Mara would agree to enter.
He used them.
Day one: he went back into the Echo alone and spent three hours running Resonance at the Bleed boundary from multiple positions, building a detailed frequency map of the differential. The map was not the same from every angle. The Bleed pressed against the Echo boundary unevenly, stronger in some places and thinner in others, and the pattern of the unevenness was information. He catalogued the positions where the differential was thinnest.
Those were the places where a Bleed-native entity could most easily sense activity on the Echo side without being in the Echo itself.
He also catalogued the positions where the differential was thickest. Those were the approach angles least likely to trigger immediate awareness in whatever was active on the other side.
He mapped the optimal entry vector.
Day two: he cross-referenced the Veil Office's partial Bleed data against what he had mapped himself. Cross's archive had fourteen documented Bleed-level encounter events over twenty-three years. All fourteen had occurred within the first thirty feet of the Bleed boundary. None of Cross's teams had penetrated further than sixty feet into the layer before extracting under pressure or sustaining casualties.
Fourteen events. Sixty feet of penetration maximum. He and Mara had been to the boundary on the third entry and had read presences much further in than sixty feet.
He noted the gap between what the Veil Office's instrumentation could reach and what Resonance and Threading together could reach, and he noted that the gap was significant and that Cross did not appear to fully understand how significant it was.
That asymmetry was a resource.
Day three: compound preparation. He and Mara worked in the pathology chemistry station for six hours producing enough of the modified compound for four entries each with two doses per entry. Eight vials apiece. He also produced a batch of the original formulation for Gordo, for the post-Bleed entry he had promised.
Day four: he read Aiken's full account again. Not for the Ledger material this time. For the Bleed specifically. Aiken had been in the Bleed eleven times across his eleven years.
His descriptions were careful and specific and organized with the methodical attention of someone who had understood from early on that what he was experiencing needed to be recorded accurately because it might matter to someone later.
Aiken described the Bleed's residuals in detail. The former humans absorbed into the layer. He noted that they did not react to the presence of marked individuals unless the marked individual ran an active ability at close range to one of them. The residuals tracked frequency, not form. If you moved through the Bleed without broadcasting a detectable signal, the residuals did not orient on you.
He noted this carefully. He noted Aiken's description of the Bleed-native entities. He noted the distinction between entities that moved on instinct and entities that moved on decision. The instinct-driven ones responded to frequency amplitude. The decision-driven ones responded to convergence patterns.
He thought about the entity in the Bleed that had let him read its marks on the third entry.
Decision-driven. Convergence-aware.
It had been waiting. He closed the notebook and called Mara.
"Tomorrow night," he said. "I have been ready since day two," she said. "I was giving you the preparation time you needed." He looked at the ceiling of his apartment. "Thank you," he said.
A pause.
"The optimal entry vector is seventeen degrees east of the standard boundary approach," he said. "The differential is thirty percent thicker there. It will take longer to push through but we will be less visible on the other side."
"I mapped the same vector this afternoon using Threading from the Echo side," she said.
He should not have been surprised. He was not, particularly.
"Tomorrow at eleven," he said. "Eleven," she confirmed, and ended the call. He sat in his apartment in the quiet and thought about the Bleed and what was in it and what it would cost to go through and what it would cost not to.
The mathematics consistently pointed in the same direction. He went to bed.