UKODM 2026 | UK Optical Design Meeting

The Extremely Large Telescope (ELT) is the world’s largest optical/infrared telescope, with a 39 metre primary mirror. HARMONI is its work-horse, near-infrared integral field spectrograph that provides simultaneous spectroscopy of approximately 30,000 spatial pixels in a 150 × 200 element field of view. It is assisted by adaptive optics (AO) using a combination of laser guide stars and faint natural stars as a reference, or, when available, a single bright natural guide star. By combining the immense light-gathering power of the ELT and the exquisite spatial resolution provided by the AO system, HARMONI has the potential to transform the landscape of observational astronomy in the coming decade. Observations of nearby extra-solar planets & extremely distant galaxies, and spatially resolved studies of objects in our own Solar System are some of the key areas where HARMONI will make a huge impact.

In this talk, I will provide a retrospective perspective on the evolution of HARMONI’s design and capabilities over the last 15 years, and provide a detailed walk-through of the changes to the instrument, both to keep up with new technologies and due to external constraints. I will highlight some of the challenges in realising an ELT scale instrument.

The Temperature is in the Blur: A New Use for the Polychromatic PSF

Traditional photometric methods estimate stellar temperatures using sequential observations through multiple filters. These color-based approaches are vulnerable to inter-exposure variability and calibration errors. This talk introduces a novel technique that enables temperature estimation from a single broadband image, by extracting spectral information encoded in the morphology of the polychromatic point spread function (P-PSF).

The method leverages the deterministic relationship between stellar spectra and diffraction-induced PSF broadening. By analyzing intensity ratios at selected radial positions—conceptually analogous to color indices, but extracted from spatial rather than spectral domains—it becomes possible to retrieve temperature with high fidelity. This approach eliminates the need for dispersive elements or filter wheels, making it particularly suitable for resource-limited platforms such as small satellites or rapid-imaging instruments.

Simulation results will be presented to assess the method’s robustness under realistic noise and sampling conditions. Performance trade-offs involving SNR, pixel sampling, and bandwidth will also be discussed. Finally, the potential for retrospective application to archival broadband imagery will be addressed.

Freeform Optics: Engineering the Future While Meeting Today’s Challenges

Freeform optics has redefined the boundaries of optical engineering, enabling unprecedented control of off-axis aberrations, compact architectures, and multifunctional systems. Yet realizing these advantages demands coordinated progress across design, manufacturing, and system integration. Key challenges include embedding manufacturability constraints into optimization, achieving deterministic freeform surface generation, advancing metrology for complex surface geometries, and standardizing the process chain from design to full system assembly. This talk will highlight recent advances in design for manufacture and underscore the challenges we must collectively address as a community. By bridging innovation and practicality, freeform optics is engineering the future while meeting today’s challenges.

Design of a freeform camera for the ELT-HARMONI integral field spectrograph

HARMONI is the first light, adaptive optics-assisted, near-infrared integral field spectrograph for the European Extremely Large Telescope. It covers a spectral range from 811 nm to 2450 nm with resolving powers from 3300 to 8000 and spatial on-sky sampling of 25 mas and 6 mas. After a rescope design phase in 2025, the number of instrument configurations has been reduced, but the requirements to transmission, alignment complexity, volume, mass and cost have become tighter. In order to meet these new requirements, the design of spectrograph sub-system has been revised.

In particular, the camera module represents an unobscured two-mirror system. It works with an offset pupil placed at the disperser plane and anamorphic F-number – F/2.77 in the spatial and F/5.54 in the spectral direction. The image also has an anamorphic format of 60.5 × 124.9 mm² that corresponds to the spatial field of view of 20.26° and spectral dispersion in the angle of 9.83°. Each of the mirrors has a freeform shape described by ZY-symmetric Chebyshev polynomials up to the 5th order. Functionally, the primary mirror has a lower optical power and operates as a corrector placed close to the pupil, while the secondary mirror carries most of the optical power. Thus, at the basic level, this design follows a Schmidt camera architecture. This approach allows to reach RMS wavefront error 36–162 nm across the entire field of view.

Freeform Optics for NASA’s Habitable Worlds Observatory

NASA’s flagship mission for the 2040s, referred to as the Habitable World Observatory (HWO), will search for and characterize habitable planets beyond our solar system, the elusive “Earth 2.0”. The observatory will also simultaneously provide powerful capabilities for transformational astrophysics discoveries from our cosmic backyard of the solar system to the distant universe and everything in between. To do all of this, HWO will require freeform optics to minimize reflections for short-wavelength instruments and maximize throughput. Freeform optics on HWO will also enable packaging very large instruments into restricted volume allocations.

Lens Design for AI and AI for Lens Design: Emerging Synergies and Opportunities

Recent advances underscore an emerging synergy between artificial intelligence (AI) and optical design. On one side, AI is being leveraged to enhance lens design—ranging from the automated generation of starting geometries (such as freeform reflectors and refractive lenses) to simulating ray tracing through complex optical structures like metasurfaces, and to the co-optimization of optical and computational elements for efficient computer vision systems. Conversely, AI itself may benefit from the systematic creation of large, structured datasets produced with commercial lens design software. These datasets enable the training of machine learning models capable of predicting or rendering realistic views of optical systems without the need for physical prototypes. This talk will explore recent progress in both directions, “Lens Design for AI” and “AI for Lens Design”, while highlighting the key challenges and opportunities at their intersection.

The Lost Sixth Primary Aberration

In Seidel’s third-order aberration theory for rotationally symmetric systems, only five primary aberrations are typically mentioned: spherical aberration, coma, astigmatism, field curvature, and distortion. However, there is a sixth rotation-invariant term in the Hamiltonian characteristic function that depends only on the object coordinates. An object-dependent term also exists in the first order, in addition to the defocus and magnification terms. This quadratic phase term is known as the “interfering, object-superimposed” phase term in Fourier optics and disappears in a 4f setup. It also disappears in the incoherent limiting case. This term is often referred to as a “piston” term, meaning it has no influence on the image. We will take a closer look at these object-dependent terms and demonstrate their importance for various optical systems. These terms can be interpreted as “telecentric deviation” and can be seamlessly integrated into Seidel’s aberration theory and a generalized Fourier optics of non-isoplanatic imaging.

Understanding secondary axial colour

The correction of secondary axial colour is a major design driver in broadband optical systems requiring a careful balance between optical performance and the cost of special glasses. Traditional analysis methods often fall short of addressing the true nature of secondary colour aberrations in design. We present a unified framework that combines a data-driven approach to dispersion modelling with a simplified, single-term theory of induced secondary colour aberrations. This approach provides an accessible and practical description of secondary colour effects, enabling deeper understanding and improved material selection for system optimisation.

Super-Athermalized Optical Design Using CODE V

Optical design engineers have for many years struggled to achieve optimal performance across a wide range of operating environmental conditions. Recent advances in optical design software have dramatically changed this situation. Optical design engineers now have unprecedented capabilities to athermalize optical designs. This is an increasingly important capability as more optical systems operate in adverse environments and are expected to deliver peak performance. We will show this capability and the difference it can make in delivering state-of-the-art imaging performance across environmental conditions though an example using Multi-Environment Coupling (MECo) in CODE V.

Image Slicers for Astronomy, from their humble beginnings to current innovation

Since their invention in the late 1980s, integral field spectrographs (IFS) have become mainstream astronomical instruments at modern ground-based and space-borne telescopes because of their versatility and highly efficient way of collecting data-cubes. In particular image slicer based IFSs achieve a very high optical throughput and information density on the detector.

In this talk I will present the evolution of image slicer designs for astronomy, from their “simple” beginning in the MPE3d instrument, to recent “complex” designs, e.g. in HARMONI for the ELT, and how scientific requirements drove and manufacturing techniques enabled this progression, e.g. the prototype slicer for the ELT Planetary Camera and Spectrograph.