Written by Aramanae Talasea -- Azur Bureau of Geometry Drive Research.

In the early years of the interstellar age, there was the belief that we could not enhance the geometry drive, so to speak. That it was such a strange and incredible design that we, poor human minds, had no chance of upgrading it. That the original drive, as it had been discovered by Rani, was all that we would get. For a moment this idea held true. For the better part of three decades we had to make do with simple copies of the drive found in the Needle.

The first crack in the idea that the drive was a “finished” object came with the Starmoth Initiative and what they like to call “Butterfly drive” — a rather bad choice of words, I think, in the sense that the device that equips their long-range exploration vessels is a run-of-the-mill geometry drive. Its secret doesn’t hide in the crystal itself but within the complex arcanes of geometry jump calculations. Through complex mathematical tricks, the Starmoth Initiative managed to drastically improve the long-range accuracy of their drives, all the while decreasing their requirements in terms of computing power. Quite a feat, despite the shortcomings of the Butterfly drive — namely, that its vastly improved long-range performance goes alongside a massive downgrade in short-range capacity.

The crystalline cube itself is as mysterious as it’s ever been and we can barely understand, let alone modify its structure, because for all intents and purposes it wasn’t created by humans. In fact, if we take Rani’s latest theories at face value, it wasn’t created at all. It’s a prime mover, an effect without a cause. We can’t influence that. We need to work on the rest, on what’s human, on the web of mathematics surrounding geometry translations. The Starmoth Initiative paved the way for this and I like to think that we reaped a small part of what they so patiently sowed.

There is a way to drastically alter the efficiency of a geometry drive. Instant geometry translations work, but they amount to brute-forcing a problem that can also be bypassed. There is a loophole in the complex corpus of equations use to calculate a translation. Something even Rani’s talent overlooked. A tangent.

We call it the Azur Effect.

A ship submitted to the Azur Effect doesn’t exactly translates to another point; rather, it skims the contact point between dimensions, slipping right under the surface of reality to re-emerge a hundred lightyears a away. Effective range is greatly enhanced at little to no additional computing cost. Sequence interdiction and black hole interference don’t apply any longer. Temporal jumps like the one Zero Fleet suffered from are no longer a danger.It's safe. Efficient. Elegant. 

There is a cost to all of this, however. Where a regular translation is innocuous, the azur effect translation submits the ship to ill-understood forces that manifest themselves under the shape of violent mechanical stress that can and will rip away the hull of any regular starship. The only way to prevent this is to equip FTL-capable vessels with “dimensional sinks” — external appliances that drain the excess force away and out of the ship’s frame. Empirical results show that dimensional sinks work best when they take seemingly aerodynamic shapes, effectively adorning our ships with wings, winglets and streamlined fuselages.

For the first time in the interstellar age, the geometry drive is going to dictate the shape of a spaceship.

NASA/JPL, "Planets of horror" series.