Comparing the contrast performance of two IFU technologies for exoplanet direct imaging with ELT-PCS

Imagine you are trying to take a photograph of a tiny, glowing firefly sitting on the edge of a massive, blindingly bright spotlight. The firefly is your exoplanet, and the spotlight is its host star. In the world of astronomy, this is the ultimate challenge: direct imaging. The star is so overwhelmingly bright that it washes out the faint light of the planet, making the planet invisible to standard cameras. To solve this, astronomers use special "sunglasses" called coronagraphs to block the star's glare, and super-sharp "glasses" called adaptive optics to fix the blurring caused by Earth's atmosphere. But even with these tools, the planet is still hiding in the shadows. To see it, you need a camera that can split the light into a rainbow (spectroscopy) to analyze the planet's atmosphere, but doing this without losing the delicate contrast needed to spot the planet is like trying to catch a whisper in a hurricane.

This is where the story of the ELT-PCS instrument comes in. The Planetary Camera and Spectrograph (ELT-PCS) is a future super-camera planned for the 39-meter Extremely Large Telescope (ELT). Its mission is to find rocky worlds around nearby stars. To do this, it needs to achieve a contrast of 10⁻⁹ (seeing something one billion times fainter than the star) at certain distances. The big question the scientists faced was: what is the best "lens" to feed this camera? They had two main contenders: Image Slicers (which chop the image into strips and rearrange them like a puzzle) and Lenslet Arrays (which use a grid of tiny lenses to break the image into many small spots). Picking the wrong one could mean the difference between finding a new world and seeing nothing but static noise.

The researchers at the University of Oxford built a clever, modular laboratory test bench to settle this debate. Think of it as a "wind tunnel" for light. Instead of waiting for a real telescope, they created a simulation on their optical bench. They used a special device called a Spatial Light Modulator (SLM) to mimic the messy, wobbly atmosphere and a Lyot coronagraph to block the "star" light, feeding a perfect, diffraction-limited image into their two competing camera technologies. They then measured how well each system could see a faint "planet" next to the blocked "star."

Here is what they found. In the raw data, without any fancy computer tricks, the Lenslet Array was the champion at very close distances to the star (less than 6 times the width of the star's diffraction pattern, or 6λ/D). It managed to dig deeper into the darkness than the Image Slicer in this specific zone. However, the Image Slicer had a secret weapon: it could be paired with a technique called "spectral deconvolution." This is like a post-processing magic trick where the computer uses the fact that the planet's light shifts slightly differently across colors than the star's noise does. When they applied this trick, the Image Slicer soared. It gained a massive advantage, especially at small separations. At a distance of 4λ/D, the Image Slicer with spectral deconvolution was 3.4 times better at finding the planet than the Lenslet Array. In fact, with this processing, the Image Slicer's performance got so good that it hit the "noise floor"—the point where the only thing stopping it from seeing even fainter objects is the electronic noise of the camera itself.

The paper is careful to note that these are initial results from a lab experiment. The Lenslet Array tests were done without the full coronagraph and atmospheric simulation that the Image Slicer had, so a perfectly fair, head-to-head race under identical extreme conditions is still on the to-do list. The authors suggest that while the Lenslet Array is great for wide views, the Image Slicer, when combined with smart computer processing, might be the superior choice for the ELT-PCS's most difficult task: spotting tiny, rocky planets right next to their blinding stars. The team plans to take this test bench to the European Southern Observatory to run it behind a real, two-stage adaptive optics system to confirm these findings.