🔭 astrophysics

Cassiopee: Defining the next-generation deformable mirror and high-speed SWIR camera for adaptive optics applications

This paper outlines a systematic approach to defining the specifications for a next-generation high-order deformable mirror and high-speed SWIR camera for adaptive optics, detailing their error budgets, experimental validation in a closed-loop testbed, and a roadmap for on-sky deployment at the Asiago Observatory to support applications ranging from exoplanet imaging to optical communications.

Original authors: B. Neichel, T. Fusco, F. Oyarzun, V. Chambouleyron, J. -F. Sauvage, C. -T. Héritier, J. Charton, C. Pravert, F. Clop, L. Maillet, M. Laslandes, J. -L. Gach, D. Boutolleau, P. Feautrier, P. Bruno, Y. W

Published 2026-08-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: B. Neichel, T. Fusco, F. Oyarzun, V. Chambouleyron, J. -F. Sauvage, C. -T. Héritier, J. Charton, C. Pravert, F. Clop, L. Maillet, M. Laslandes, J. -L. Gach, D. Boutolleau, P. Feautrier, P. Bruno, Y. Wanwanscappel, C. Bremond

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The night sky is not a static backdrop; it is a turbulent ocean of air that constantly distorts the light from distant stars. As starlight passes through Earth's atmosphere, pockets of air at different temperatures and speeds bend the light rays, causing the familiar twinkling of stars. For astronomers, this blurring is a major obstacle, turning sharp points of light into fuzzy blobs and hiding faint objects like planets orbiting other stars. To see clearly, scientists use a technology called adaptive optics. This system acts like a pair of smart glasses for a telescope. It measures the distortion in the incoming light thousands of times per second and then physically reshapes a mirror to cancel out the blurring, restoring the image to its original sharpness. While this technology has been a game-changer for ground-based astronomy, the next generation of telescopes and new applications like high-speed laser communication demand even faster, more precise, and more sensitive tools than currently exist.

A team of researchers in France, working on a project called Cassiopee, has taken a systematic approach to defining exactly what these next-generation tools need to be. Instead of guessing what might be useful, the team started by looking at four very different real-world scenarios where clear vision through turbulence is critical. These scenarios range from trying to photograph Earth-like planets around nearby stars, to sending high-speed data to satellites, to tracking space debris, and finally to focusing powerful laser beams for defense. By analyzing the specific challenges of each of these four areas, the researchers were able to identify the common requirements that would make a single, powerful system capable of handling all of them. They found that while the goals differ, the underlying need for extreme speed and precision is the same, justifying a unified effort to build better components.

The core of their work involved setting strict performance targets for two key pieces of hardware: a special mirror that can change its shape and a camera that can see in the dark and move incredibly fast. For the mirror, the team determined it needs to be large enough to have over sixteen thousand tiny actuators, or motors, that push and pull its surface. These motors must be able to move with a precision finer than a single billionth of a meter, and they must settle into their new shape in less than three hundred microseconds. To achieve this speed and reduce the complexity of wiring, the researchers designed the mirror with its own built-in electronics, placing the control chips directly on the mirror itself rather than in a separate box. This integration allows the mirror to receive commands and react almost instantly, a necessity for keeping up with the rapid changes in the atmosphere.

On the camera side, the requirements are equally demanding. The researchers specified a large sensor capable of capturing images at a rate of three thousand frames per second. To see faint objects without being overwhelmed by the camera's own electronic noise, the device must be able to detect light with a sensitivity better than one electron per pixel. This level of sensitivity is achieved using a specific type of sensor technology that amplifies the signal from incoming light before it is read out, effectively silencing the background noise. The camera also needs to operate at very cold temperatures to prevent heat from creating false signals, all while transferring massive amounts of data instantly to the computer that controls the mirror.

To prove that these new components work together, the team is building a laboratory testbed that simulates the turbulent conditions of the atmosphere. This setup includes a rotating screen that creates artificial turbulence, allowing the researchers to test the mirror and camera in a controlled environment before they ever point a telescope at the sky. They are also planning a final validation step using a new facility at the Asiago Observatory in Italy. There, they will install the system on a real telescope to demonstrate its ability to correct for the actual atmosphere, measure its performance in real-time, and show that it can successfully focus light for fiber-optic communication.

The findings of this paper do not claim that these components are already built and ready for use, but rather that the team has successfully mapped out exactly what is needed and how to build it. By starting with concrete use-cases and working backward to the hardware specifications, they have created a clear roadmap. The result is a set of detailed requirements for a mirror with over sixteen thousand actuators and a camera that can see single electrons at high speed. These specifications represent a significant leap forward from current technology, offering the potential to unlock new capabilities in astronomy, space surveillance, and global communications. The project moves the field from theoretical possibility to a concrete engineering plan, bridging the gap between laboratory components and the future of high-performance optical systems.

Technical Summary: The Cassiopee Project

Problem Statement
Future adaptive optics (AO) systems required for extremely large telescopes (ELTs), free-space optical (FSO) communications, space situational awareness (SSA), and directed-energy defense applications face unprecedented demands for speed, sensitivity, and precision. Current components are insufficient to meet the requirements of these diverse domains, particularly the need for deformable mirrors (DMs) with thousands of actuators and wavefront sensor cameras capable of sub-electron read-out noise at kilohertz frame rates over large formats. The challenge lies in defining a unified set of specifications that satisfies the most stringent requirements across these distinct operational regimes while enabling a single, co-developed technological solution.

Methodology
The Cassiopee project employs a systematic, top-down error budget methodology to derive component specifications. The approach begins with four representative use-cases:

Direct Exoplanet Imaging (XAO for ELT): The most demanding scenario, requiring focal-plane contrasts of $10^{-9}$ .
Free-Space Optical Communications (FSO): Requiring high loop bandwidths to track rapidly moving Low Earth Orbit (LEO) satellites.
Space Situational Awareness (SSA): Demanding high-order correction for non-cooperative targets at low elevation angles.
Directed Energy/Laser Focusing: Emphasizing DM stroke and robustness for long-range atmospheric compensation.

For each use-case, a detailed AO error budget is constructed, decomposing total wavefront error into temporal, noise, fitting, non-common-path, and aliasing terms. The Marechal approximation is used to relate residual variance to Strehl ratio and contrast. By identifying the dominant error terms for each domain, the project establishes quantitative top-level requirements for the DM and camera, ensuring that the final specifications satisfy the most stringent constraints (primarily driven by the XAO/ELT case) while accommodating the specific needs of the other domains.

Key Contributions and Derived Specifications
The paper defines the specifications for two critical next-generation components developed in collaboration with industrial partners Bertin-Alpao and First Light Imaging (FLI):

High-Order Deformable Mirror (DM):
- Actuator Count: $\ge 128 \times 128$ (prototype $\ge 20 \times 20$ ), driven by XAO fitting error requirements.
- Stroke: $\ge 20 \mu m$ peak-to-valley, driven by SSA/FSO requirements at low elevation angles.
- Resolution: Sub-nanometer ( $\le 100$ pm) to suppress stellar speckles for high-contrast imaging.
- Bandwidth: $-3$ dB bandwidth $\ge 1$ kHz with a small-stroke settling time $\le 300 \mu s$ .
- Architecture: A key innovation is the integration of digital-to-analog converters and analog amplifiers directly onto the mirror body. This reduces cabling complexity by a factor of ~20 and enables data throughput of $\ge 10$ Gbit/s, allowing full actuator command updates at loop rates up to 3 kHz.
Large-Format e-APD SWIR Camera:
- Format: $512 \times 512$ pixels, scaling up from the existing $320 \times 256$ C-RED One platform.
- Noise Performance: Sub-electron read-out noise ( $< 1 e^-$ rms) at 3000 fps, achieved via electron-avalanche photodiode (e-APD) technology which amplifies the signal before digitization.
- Frame Rate: $\ge 3000$ fps (single read) to support temporal bandwidths $\ge 2$ kHz.
- Sensitivity: Quantum efficiency $> 70\%$ in J, H, and K bands, with dark current $< 5 e^-/s/pixel$ at 80 K.
- Data Link: High-speed interfaces (2 $\times$ 10 GigE or CXP-12) to handle raw image data transfer without introducing latency.

Experimental Validation Strategy
The project outlines a three-stage validation roadmap:

Standalone Characterization: Initial testing by manufacturers.
Laboratory Integration: A full closed-loop AO testbed at LAM (Laboratoire d'Astrophysique de Marseille) and ONERA. This facility utilizes rotating phase screens to simulate atmospheric turbulence (Paranal median and challenging FSO profiles) and validates the combined performance of the DM and camera. It also investigates next-generation wavefront sensor architectures, such as non-modulated pyramid and Fourier-filtering sensors, to exploit the new camera's capabilities.
On-Sky Validation: Deployment at the EKARUS facility, a new AO bench installed at the Asiago Astrophysical Observatory (1.82 m Copernico telescope). This stage will validate the system under real atmospheric conditions, measuring closed-loop Strehl ratios, actuator dynamics, and fiber injection efficiency as a proxy for FSO coupling.

Significance
The paper asserts that the Cassiopee project represents a "step-change improvement" over the current state of the art. By converging the requirements of astronomy, telecommunications, and defense into a single development effort, the project aims to deliver components that exceed the performance of any currently available device by a factor of at least four. The significance lies not only in the specific performance metrics (e.g., 512 $\times$ 512 format, sub-electron noise, embedded electronics) but in the demonstration that these requirements are achievable through the application of mature technologies (e-APD arrays, electromagnetic voice-coil actuators) pushed to new regimes via architectural innovations in readout electronics and embedded digital control. The successful validation of these components is positioned as a critical bridge between component-level testing and the deployment of next-generation AO systems for ELTs and ground-to-space optical links.

More like this